AMCC Computer Hardware 9590SE 4ME User Manual

®
3ware  
Serial ATA RAID Controller  
Command Line Interface  
Supports the 9590SE-4ME for the Power Mac® G5  
PN: 720-0140-00  
September 2006  
Table of Contents  
focus Object . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .19  
update fw=filename_with_path [force] . . . . . . . . . . . . . . . . . . . . . . . . . . . .24  
/cx show . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26  
/cx show attribute [attribute ...] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .27  
/cx show driver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .28  
/cx show model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .28  
/cx show firmware . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .28  
/cx show bios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .28  
/cx show monitor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .28  
/cx show serial . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .29  
/cx show pcb . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .29  
/cx show pchip . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .29  
/cx show achip . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .29  
iii  
/cx show numdrives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .29  
/cx show numports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .30  
/cx show numunits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .30  
/cx show ctlbus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .30  
/cx show exportjbod . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .30  
/cx show spinup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .31  
/cx show stagger . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .31  
/cx show ondegrade . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .31  
/cx show autocarve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .31  
/cx show carvesize . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .32  
/cx show memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .32  
/cx show autorebuild . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .33  
/cx show unitstatus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .33  
/cx show allunitstatus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .34  
/cx show drivestatus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .34  
/cx show all . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .35  
/cx add type=<RaidType> disk=<p:-p> [stripe=Stripe] [noscan]  
[group=<3|4|5|6>] [nocache] [autoverify] [ignoreECC] [name=string]  
[storsave=<protect|balance|perform>] . . . . . . . . . . . . . . . . . . . . . . . . . . .36  
/cx rescan [noscan] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .38  
/cx commit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .39  
/cx flush . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .39  
/cx update fw=filename_with_path [force] . . . . . . . . . . . . . . . . . . . . . . . . . .39  
/cx show alarms [reverse] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .40  
/cx show diag . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .40  
/cx show rebuild . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .41  
/cx show verify . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .43  
/cx show selftest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .44  
/cx add rebuild=ddd:hh:duration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .45  
/cx add verify=ddd:hh:duration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .46  
/cx add selftest=ddd:hh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .47  
/cx del rebuild=slot_id . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .47  
/cx del verify=slot_id . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .48  
/cx del selftest=slot_id . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .48  
/cx set rebuild=enable|disable|1..5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .48  
/cx set verify=enable|disable|1..5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .49  
/cx set selftest=enable|disable [task=UDMA|SMART] . . . . . . . . . . . . . . . . .49  
/cx set exportjbod=on|off . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .49  
/cx set ondegrade=cacheoff|follow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .50  
/cx set spinup=nn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .50  
/cx set stagger=nn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .50  
/cx set autocarve=on|off . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .50  
/cx set carvesize=[1024..2048] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .51  
/cx set autorebuild=on|off . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .51  
/cx start mediascan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .52  
/cx stop mediascan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .52  
/cx/ux show . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .53  
/cx/ux show attribute [attribute ...] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .54  
/cx/ux show status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .54  
/cx/ux show rebuildstatus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .55  
/cx/ux show verifystatus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .55  
/cx/ux show initializestatus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .55  
iv  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
/cx/ux show name . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .55  
/cx/ux show serial . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .55  
/cx/ux show qpolicy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .56  
/cx/ux show storsave . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .56  
/cx/ux show identify . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .56  
/cx/ux show autoverify . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .56  
/cx/ux show cache . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .57  
/cx/ux show ignoreECC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .57  
/cx/ux show volumes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .57  
/cx/ux show all . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .57  
/cx/ux remove [noscan] [quiet] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .58  
/cx/ux del [noscan] [quiet] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .58  
/cx/ux start rebuild disk=p<p:-p...> [ignoreECC] . . . . . . . . . . . . . . . . . . . . .59  
/cx/ux start verify . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .59  
/cx/ux pause rebuild . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .60  
/cx/ux resume rebuild . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .60  
/cx/ux stop verify . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .60  
/cx/ux flush . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .61  
/cx/ux set autoverify=on|off . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .61  
/cx/ux set cache=on|off [quiet] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .61  
/cx/ux set identify=on|off . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .62  
/cx/ux set ignoreECC=on|off . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .62  
/cx/ux set name=string . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .62  
/cx/ux set qpolicy=on|off . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .62  
/cx/ux set storsave=protect|balance|perform [quiet] . . . . . . . . . . . . . . . . . . .63  
/cx/ux migrate type=RaidType [disk=p:-p]  
[group=3|4|5|6] [stripe=Stripe] [noscan] [nocache] [autoverify] . . . . . . . . .64  
/cx/px show . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .69  
/cx/px show attribute [attribute ...] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .69  
/cx/px show status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .70  
/cx/px show model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .70  
/cx/px show serial . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .70  
/cx/px show firmware . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .70  
/cx/px show identify . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .70  
/cx/px show ncq . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .71  
/cx/px show lspeed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .71  
/cx/px show capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .71  
/cx/px show smart . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .71  
/cx/px show all . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .72  
/cx/px remove [noscan] [quiet] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .73  
/cx/px set identify=on|off . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .73  
/cx/bbu show . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .74  
/cx/bbu show attribute [attribute ...] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .75  
/cx/bbu show status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .75  
/cx/bbu show batinst . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .76  
/cx/bbu show lasttest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .76  
/cx/bbu show volt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .76  
/cx/bbu show temp . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .76  
/cx/bbu show cap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .76  
/cx/bbu show serial . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .77  
/cx/bbu show fw . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .77  
v
/cx/bbu show pcb . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .77  
/cx/bbu show bootloader . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .77  
/cx/bbu show all . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .77  
/cx/bbu test [quiet] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .77  
/cx/bbu enable . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .78  
/cx/bbu disable [quiet] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .78  
/ex show . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .79  
/ex show attribute [attribute ...] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .80  
/ex show controllers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .80  
/ex show slots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .80  
/ex show fans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .81  
/ex show temp . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .81  
/ex show all . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .81  
/ex/slotx show . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .81  
/ex/slotx show identify . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .82  
/ex/slotx set identify=on|off . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .82  
/ex/fanx show . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .82  
/ex/tempx show . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .82  
help show . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .85  
help flush . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .85  
help rescan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .86  
help commit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .86  
help focus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .86  
help /cx . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .86  
help /cx/ux . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .86  
help /cx/px . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .86  
help /cx/bbu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .87  
help /ex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .87  
help /ex/slotx . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .87  
help /ex/fanx . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .87  
help /ex/tempx . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .87  
vi  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
About this CLI Guide  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5 provides  
instructions for configuring and maintaining your 3ware controller using  
3ware’s command line interface (CLI).  
Note: This guide assumes that you have already installed your 3ware RAID  
controller in your system and set up your 3ware Sidecar. If you have not yet done  
so, see the installation guide that came with your 3ware RAID controller for  
instructions.  
Note: The 3ware controller used for the Apple® Power Mac® G5 is the 9590SE-  
4ME. AMCC makes a number of other 3ware controller models, including 9000-  
series (9500S, 9550SX, and 9590SE), and earlier 7000/8000 series boards, that  
work on other platforms.  
The Command Syntax described in this document supports all of those boards, and  
some commands indicate the particular boards that support the command.  
For your Power Mac G5, the CLI commands that apply to the 9590SE-4ME are  
those identified as 9000 series and those that are identified as 9590SE.  
Note: You may notice mention of RAID 50 in the Command Syntax sections. RAID  
50 requires six or more drives, and so is not available using the 9590SE-4ME.  
Table 1: Sections in this CLI Guide  
Chapter  
Description  
1. Introduction to 3ware  
Command Line Interface  
Installation, features, concepts  
2. CLI Syntax Reference  
Describes individual commands using the  
primary syntax  
There are often multiple ways to accomplish the same configuration and  
maintenance tasks for your 3ware controller. While this manual includes  
instructions for performing tasks using the command line interface, you can  
®
also use 3DM 2 (3ware Disk Manager) to accomplish these tasks.  
For details, see the user guide or the 3ware HTML Bookshelf.  
1
2
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
1
Introduction to the 3ware  
Command Line Interface  
3ware RAID controllers can be managed using the CLI via a command line or  
script.  
Important!  
For all of the functions of the 3ware CLI to work properly, you must have the proper  
CLI, firmware, and driver versions installed. Check http://www.3ware.com for the  
latest versions and upgrade instructions.  
This chapter includes the following sections:  
Features of the CLI  
3ware CLI is a command line interface for managing 3ware RAID  
Controllers. It provides controller, logical unit, drive, and enclosure  
management. It can be used in both interactive and batch mode, providing  
higher level API (application programming interface) functionalities.  
You can use the CLI to view unit status and version information and perform  
maintenance functions such as adding or removing drives. 3ware CLI also  
includes advanced features for creating and deleting RAID units online.  
For a summary of what you can do using the CLI, see “Common Tasks  
3
   
Chapter 1. Introduction to the 3ware Command Line Interface  
Supported Operating Systems  
The 3ware CLI is supported under the following operating systems:  
Mac OS X, running in a Power Mac® G5 (PowerPC™-based) with PCI  
Express®.  
Windows®. Windows 2000, Windows XP, and Windows Server 2003,  
both 32-bit and 64-bit.  
Linux®. Redhat, SuSE, both 32-bit and 64-bit.  
FreeBSD®, both 32-bit and 64-bit.  
Terminology  
This document (3ware Serial ATA RAID Controller CLI Guide for the Power  
Mac G5) uses the following terminology:  
Logical Units. Usually shortened to “units.” These are block devices  
presented to the operating system. A logical unit can be a one-tier, two-tier, or  
three-tier arrangement. Spare and Single logical units are examples of one-tier  
units. RAID 1 and RAID 5 are examples of two-tier units and as such will  
have sub-units. RAID 10 is an example of a three-tier unit and as such will  
have sub-sub-units.  
Port. A controller has one or many ports. Each port can be attached to a single  
disk drive. On a controller such as the 9590SE-4ME, with a multilane serial  
port connector, one connector supports four ports.  
For additional information about 3ware controller concepts and terminology,  
see the user guide that came with your 3ware RAID controller or the user  
guide portions of the 3ware HTML Bookshelf.  
4
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
   
Installing the 3ware CLI on Mac OS X  
Installing the 3ware CLI on Mac OS X  
3ware CLI can be installed from the 3ware software CD, or the latest version  
can be downloaded from the 3ware web site, http://www.3ware.com/support/  
To install 3ware CLI on Mac OS X  
Run the installer and select CLI as the software to be installed.  
Permissions Required to Run CLI  
To run CLI, you can be logged onto the Power Mac G5 with one of the  
following sets of permissions:  
Administrator  
Root  
You can also use SUDO to run CLI.  
To start CLI  
In a Terminal window, type  
sudo ./tw_cli  
If prompted, enter your password.  
Working with 3ware CLI  
You can work with the 3ware CLI in different ways:  
Interactively, entering commands at the main prompt  
As a series of single commands  
By creating a script—an input file with multiple commands  
The next few topics shows examples of these different methods.  
Examples shown in the CLI Syntax Reference chapter reflects the interactive  
method.  
5
   
Chapter 1. Introduction to the 3ware Command Line Interface  
Using the command interface interactively  
You can use 3ware CLI interactively, entering commands at the main prompt  
and observing the results on the screen.  
To use the CLI interactively  
1
Enter the following command:  
./tw_cli  
The main prompt is displayed, indicating that the program is awaiting a  
command.  
//localhost>  
2
At the CLI prompt, you can enter commands to show or act on 3ware  
controllers, units, and drives.  
For example,  
//localhost> show  
displays all controllers in the system and shows details about them, like  
this:  
Ctl Model Ports Drives Units NotOpt RRate VRate BBU  
------------------------------------------------------------  
c0 9590SE-4 4  
4
1
0
3
5 TESTING  
Using a single command with output  
You can use 3ware CLI with line arguments, processing a single command at  
a time. To do so, simply enter the command and the arguments.  
Single commands can be useful when you want to perform a task such as  
redirecting the output of the command to a file. It also allows you to use the  
command line history to eliminate some typing.  
Syntax  
tw_cli <command line arguments>  
Example  
tw_cli /c0 show diag > /tmp/3w_diag.out  
6
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
   
Working with 3ware CLI  
Using an input file to execute a script  
You can operate 3ware CLI scripts by executing a file. The file is a text file  
containing a list of CLI commands which you have entered in advance. Each  
command must be on a separate line.  
Syntax  
tw_cli -f <filename>  
Where <filename>is the name of the text file you want to execute.  
Example  
tw_cli -f clicommand.txt  
This example executes the file clicommand.txt, and runs the CLI commands  
included in that file.  
Scripting example  
Following is a a scripting example using a text file called config_unit.txt,  
containing three commands. This example sets up a 4-port controller with two  
units, each with 2 drives mirrored. It then prints the configurations for  
verification. The commands included in the script file are:  
/c0 add type=raid1 disk=0-1  
/c0 add type=raid1 disk=0-1  
/c0 show  
Outputting the CLI to a Text File  
You can have the output of the 3ware CLI, including errors, sent to a text file  
by adding 2>&1 to the end of the line. This could be useful, for example, if  
you want to email the output to AMCC Technical Support.  
Examples  
tw_cli /c2/p0 show >> controller2port0info.txt 2>&1  
or  
tw_cli /c0 show diag >> Logfile.txt 2>&1  
7
   
Chapter 1. Introduction to the 3ware Command Line Interface  
Conventions  
The following conventions are used through this guide:  
In text, monospace fontis used for code and for things you type.  
In descriptions and explanations of commands, a bold font indicates the  
name of commands and parameters, for example, /c0/p0 show all.  
In commands, an italic font indicates items that are variable, but that you  
must specify, such as a controller ID, or a unit ID, for example, /c0/p0  
show attribute, and /cx/px show all  
In commands, brackets around an item indicates that it is optional.  
In commands, ellipses (...) indicate that more than one parameter at a time  
can be included, for example, /c0/p0 show attribute [attribute ...], or that  
there is a range between two values from which you can pick a value, for  
example, /cx set carvesize=[1024...2048].  
In commands, a vertical bar (|) indicates an 'or' situation where the user  
has a choice between more than one attribute, but only one can be  
specified.  
Example: In the command to rescan all ports and reconstitute all units, the  
syntax appears as rescan [cid...] [noscan]. The italic cid indicates that you  
need to supply a controller ID. The ellipses (...) indicate that you can specify  
more than one controller ID, separated by spaces. The brackets [ ] indicate  
that you may omit the controller ID, to rescan all controllers, and the noscan  
parameter, so that the operation will be reported to the operating system.  
Understanding RAID Levels and Concepts  
3ware RAID controllers use a Redundant Array of Inexpensive Disks (RAID)  
to increase your storage system’s performance and provide fault tolerance  
(protection against data loss).  
This section organizes information about RAID concepts and configuration  
levels into the following topics:  
“RAID Concepts” on page 9  
“Available RAID Configurations” on page 10  
“Determining What RAID Level to Use” on page 13  
8
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
   
Understanding RAID Levels and Concepts  
RAID Concepts  
The following concepts are important to understand when working with a  
RAID controller:  
Arrays and Units. In the storage industry, the term “array” is used to  
describe two or more disk drives that appear to the operating system as a  
single unit. When you work with 3ware software, “unit” is the term used  
to refer to an array of disks that is configured and managed through the  
3ware software. Single-disk units can also be configured in the 3ware  
software.  
Mirroring. Mirrored arrays write data to paired drives simultaneously. If  
one drive fails, the data is preserved on the paired drive. Mirroring  
provides data protection through redundancy. In addition, mirroring using  
a 3ware RAID controller provides improved performance because  
3ware’s TwinStor technology reads from both drives simultaneously.  
Striping. Striping across disks allows data to be written and accessed on  
more than one drive, at the same time. Striping combines each drive’s  
capacity into one large volume. Striped disk arrays achieve highest  
transfer rates and performance at the expense of fault tolerance.  
Distributed Parity. Parity works in combination with striping on RAID 5.  
Parity information is written to each of the striped drives, in rotation.  
Should a failure occur, the data on the failed drive can be reconstructed  
from the data on the other drives.  
Hot Swap. The process of exchanging a drive without having to shut  
down the system. This is useful when you need to exchange a defective  
drive in a redundant array.  
Array Roaming. The process of removing a unit from a controller and  
putting it back later, either on the same controller, or a different one, and  
having it recognized as a unit. The disks may be attached to different ports  
than they were originally attached to, without harm to the data.  
For definitions of other terms used throughout the documentation, see the  
“Glossary”.  
9
 
Chapter 1. Introduction to the 3ware Command Line Interface  
Available RAID Configurations  
The following RAID levels and configurations are available for drives  
attached to a 9590SE-4ME 3ware RAID controller:  
RAID 0  
RAID 1  
RAID 5  
RAID 10  
Single Disk  
Hot Spare  
Note: You may notice mention of RAID 50 in the Command Syntax sections of the  
CLI Guide. RAID 50 requires six or more drives, and so is not available using the  
9590SE-4ME.  
For how to configure RAID units, see “Configuring a New Unit” on page 34.  
RAID 0  
RAID 0 provides improved performance, but no fault tolerance. Since the  
data is striped across more than one disk, RAID 0 disk arrays achieve high  
transfer rates because they can read and write data on more than one drive  
simultaneously. The stripe size is configurable during unit creation. Requires  
a minimum of two drives.  
When drives are configured in a striped disk array (see Figure 1), large files  
are distributed across the multiple disks using RAID 0 techniques.  
Striped disk arrays give exceptional performance, particularly for data  
intensive applications such as video editing, computer aided design and  
geographical information systems.  
RAID 0 arrays are not fault tolerant. The loss of any drive results in the loss of  
all the data in that array, and can even cause a system hang, depending on  
your operating system. RAID 0 arrays are not recommended for high  
availability systems unless additional precautions are taken to prevent system  
hangs and data loss.  
10  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
 
Understanding RAID Levels and Concepts  
Figure 1. RAID 0 Configuration Example  
RAID 1  
RAID 1 provides fault tolerance and a speed advantage over non-RAID disks.  
Also known as a mirrored array. Mirroring is done on pairs of drives.  
Mirrored disk arrays write the same data to two different drives using RAID 1  
algorithms (see Figure 2). This gives your system fault tolerance by  
preserving the data on one drive if the other drive fails. Fault tolerance is a  
basic requirement for critical systems like web and database servers.  
3ware uses a patented technology, TwinStor®, on RAID 1 arrays for  
improved performance during sequential read operations. With TwinStor  
technology, read performance is twice the speed of a single drive during  
sequential read operation.  
The adaptive algorithms in TwinStor technology boost performance by  
distinguishing between random and sequential read requests. For the  
sequential requests generated when accessing large files, both drives are used,  
with the heads simultaneously reading alternating sections of the file. For the  
smaller random transactions, the data is read from a single optimal drive head.  
Figure 2. RAID 1 Configuration Example  
RAID 5  
RAID 5 provides performance, fault tolerance, high capacity, and storage  
efficiency. Requires a minimum of three drives. Combines striping data with  
parity (exclusive OR) to restore data in case of a drive failure.  
Parity information is distributed across all drives rather than being  
concentrated on a single disk (see Figure 3). This avoids throughput loss due  
to contention for the parity drive.  
11  
Chapter 1. Introduction to the 3ware Command Line Interface  
Figure 3. RAID 5 Configuration Example  
RAID 10  
RAID 10 is a combination of striped and mirrored arrays for fault tolerance  
and high performance.  
When drives are configured as a striped mirrored array, the disks are  
configured using both RAID 0 and RAID 1 techniques, thus the name RAID  
10 (see Figure 4). A minimum of four drives are required to use this  
technique. The first two drives are mirrored as a fault tolerant array using  
RAID 1. The third and fourth drives are mirrored as a second fault tolerant  
array using RAID 1. The two mirrored arrays are then grouped as a striped  
RAID 0 array using a two tier structure. Higher data transfer rates are  
achieved by leveraging TwinStor and striping the arrays.  
In addition, RAID 10 arrays offer a higher degree of fault tolerance than other  
types of RAID levels (such as RAID 1 and RAID 5), since the array can  
sustain multiple drive failures without data loss. Please note that if both halves  
of a mirrored pair in the RAID 10 array fail, then all of the data will be lost.  
Figure 4. RAID 10 Configuration Example  
12  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
Understanding RAID Levels and Concepts  
Single Disk  
A single drive can be configured as a unit through 3ware software. (3DM 2 or  
CLI). Like disks in other RAID configurations, single disks contain 3ware  
Disk Control Block (DCB) information and are seen by the OS as available  
units.  
Single drives are not fault tolerant and therefore not recommended for high  
availability systems unless additional precautions are taken to prevent system  
hangs and data loss.  
Hot Spare  
A hot spare is a single drive, available online, so that a redundant unit can be  
automatically rebuilt in case of drive failure.  
Determining What RAID Level to Use  
Your choice of which type of RAID unit (array) to create will depend on your  
needs. You may wish to maximize speed of access, total amount of storage, or  
redundant protection of data. Each type of RAID unit offers a different blend  
of these characteristics.  
The following table provides a brief summary of RAID type characteristics.  
Table 2: RAID Configuration Types  
RAID Type  
Description  
RAID 0  
RAID 1  
Provides performance, but no fault tolerance.  
Provides fault tolerance and a read speed advantage over non-  
RAID disks.  
RAID 5  
This type of unit provides performance, fault tolerance, and high  
storage efficiency.  
RAID 10  
Single Disk  
A combination of striped and mirrored units for fault tolerance  
and high performance.  
Not a RAID type, but supported as a configuration.  
Provides for maximum disk capacity with no redundancy.  
You can create one or more units, depending on the number of drives you  
have installed.  
13  
 
Chapter 1. Introduction to the 3ware Command Line Interface  
Using Drive Capacity Efficiently  
To make the most efficient use of drive capacity, it is advisable to use drives  
of the same capacity. This is because the capacity of each drive is limited to  
the capacity of the smallest drive in the unit.  
The total array capacity is defined as follows:  
Table 3: Drive Capacity  
RAID Level  
Capacity  
Single Disk  
RAID 0  
capacity of the drive  
(number of drives) X (capacity of the smallest drive)  
capacity of the smallest drive  
RAID 1  
RAID 5  
(number of drives - 1) X (capacity of the smallest drive)  
Storage efficiency increases with the number of disks:  
storage efficiency = (number of drives -1)/(number of drives)  
RAID 10  
(number of drives / 2) X (capacity of smallest drive)  
Through drive coercion, the capacity used for each drive is rounded down so  
that drives from differing manufacturers are more likely to be able to be used  
as spares for each other. The capacity used for each drive is rounded down to  
the nearest GB for drives under 45 GB (45,000,000,000 bytes), and rounded  
down to the nearest 5 GB for drives over 45 GB. For example, a 44.3 GB  
drive will be rounded down to 44 GB, and a 123 GB drive will be rounded  
down to 120 GB. For more information, see the discussion of drive coercion  
under “Creating a Hot Spare” on page 42.  
14  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
2
CLI Syntax Reference  
Note: The 3ware controller used for the Power Mac G5 is the 9590SE-4ME. AMCC  
makes a number of other 3ware controller models, including 9000-series (9500S,  
9550SX, and 9590SE), and earlier 7000/8000 series boards, that work on other  
platforms.  
The command syntax described in this document supports all of those boards,  
however some features are only supported on specific 3ware controller models. We  
have tried to indicate the controller model in the Syntax description. For example, a  
command that is only available for all 9000-series models will specify “9000 series”  
or “9000 only,” while a command that is available for only 9550SX and 9590SE  
models will specify “9550SX and 9590SE”, and a command that is available only on  
7000/8000-series controllers will specify “7000/8000 only.”  
For your Power Mac G5, the CLI commands that apply to the 9590SE-4ME are  
those identified as 9000 series and those that are identified as 9590SE.  
Note: You may notice mention of RAID 50 in the Command Syntax sections. RAID  
50 requires six or more drives, and so is not available using the 9590SE-4ME.  
This chapter provides detailed information about using the command syntax  
for the 3ware CLI.  
Throughout this chapter the examples reflect the interactive method of  
executing 3ware CLI.  
Common Tasks Mapped to CLI Commands  
The table below lists many of the tasks people use to manage their RAID  
controllers and units, and lists the primary CLI command associated with  
those tasks.  
15  
       
Chapter 2. CLI Syntax Reference  
Table 4: Common Tasks Mapped to CLI Commands  
Task  
CLI Command  
Page  
Controller Configuration Tasks  
View information about a controller /cx show  
View controller policies  
/cx show [attribute] [attribute]  
Set policies for a controller  
Export JBODs  
/cx set exportjbod  
Modify staggered spinup  
/cx set stagger and /cx set spinup 50  
Disable write cache on unit  
degrade  
/cx set ondegrade  
Enable/disable autocarving  
Enable/disable autorebuild  
/cx set autocarve  
/cx set autorebuild  
Set the autocarve volume size /cx set carvesize  
Unit Configuration Tasks  
Create a new unit  
/cx add  
Create a hot spare  
/cx add  
Enable/disable unit write cache  
Set the queue policy  
Set the storsave profile  
/cx/ux set cache  
/cx/ux set qpolicy  
/cx/ux set storsave  
Unit Configuration Changes  
Change RAID level  
Change stripe size  
Expand unit capacity  
Delete a unit  
/cx/ux migrate  
/cx/ux migrate  
/cx/ux migrate  
/cx/ux del  
Remove a unit (export)  
Name a unit  
/cx/ux remove  
/cx/ux set name  
Controller Maintenance Tasks  
Update controller with new  
firmware  
/cx update  
Add a time slot to a rebuild  
schedule  
/cx add rebuild  
16  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
Common Tasks Mapped to CLI Commands  
Table 4: Common Tasks Mapped to CLI Commands  
Task  
CLI Command  
Page  
Add a time slot to a verify  
schedule  
/cx add verify  
Add a time slot to a selftest  
schedule  
/cx add selftest  
/cx set rebuild  
/cx set verify  
Enable/disable the rebuild/migrate  
schedule and set the task rate  
Enable/disable the verify schedule  
and set the task rate  
Enable/disable the selftest  
schedule  
/cx set selftest  
/cx show alarms  
View Alarms  
Unit Maintenance Tasks  
Start a rebuild  
/cx/ux start rebuild  
/cx/ux start verify  
Start a verify  
Pause/resume rebuild  
/cx/ux pause rebuild and /cx/ux  
resume rebuild  
Stop verify  
/cx/ux stop verify  
/cx/ux set autoverify  
/cx/ux set identify  
64  
Enable/disable autoverify  
Identify all drives that make up a  
unit by blinking associated LEDs  
Port Tasks  
Locate drive by blinking an LED  
Check if LED is set to on or off  
View information for specific drive  
View the status of specific drive  
/cx/px set identify  
/cx/px show identify  
/cx/px show  
/cx/px show status  
Enclosure Tasks  
View information about an  
enclosure  
/ex show  
Locate a particular drive slot in an  
enclosure by blinking an LED  
/ex/slotx set identify  
17  
Chapter 2. CLI Syntax Reference  
Syntax Overview  
The command syntax uses the general form:  
Object Command Attributes  
Objects are shell commands, controllers, units, ports (drives), BBUs (battery  
backup units), and enclosures.  
Commands can either select (show, get, present, read) attributes or alter (add,  
change, set, write) attributes.  
Attributes are either Boolean Attributes or Name-Value Attributes.  
The value of a boolean attribute is deduced by presence or lack of—that  
is, the attribute is either specified, or not. For example, the command  
show alarms by default lists alarms with the most recent alarm first. If  
you include the attribute reverse, as in the command show alarms  
reverse, alarms are listed in reverse order.  
The value of name-value attributes are expressed in the format  
attribute=value.  
Example: When adding (creating) a unit to the controller with the following  
command string,  
/c1 add type=raid1 disk=0-1  
c1is the object, addis the command, type(for type of array) is an attribute  
with raid1as the value of the attribute, and diskis another attribute with  
0-1as the value (ports 0 through 1).  
Information about commands is organized by the object on which the  
commands act:  
Shell Object Commands. Shell object commands set the focus or provide  
information (such as alarms, diagnostics, rebuild schedules, and so forth)  
about all controllers in the system. For details, see “Shell Object Commands”  
Controller Object Commands. Controller object commands provide  
information and perform actions related to a specific controller. For example,  
you use controller object commands for such tasks as seeing alarms specific  
to a controller, creating schedules during which background tasks are run, and  
setting policies for the controller. You also use the controller object command  
/cx add type to create RAID arrays. For details, see “Controller Object  
Unit Object Commands. Unit object commands provide information and  
perform actions related to a specific unit on a specific controller. For example,  
you use unit object commands for such tasks as seeing the rebuild, verify, or  
initialize status of a unit, starting, stopping, and resuming verifies, starting  
and stopping rebuilds, and setting policies for the unit. You also use the  
controller object command /cx/ux migrate to change the configuration of a  
RAID array. For details, see “Unit Object Commands” on page 53.  
18  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
 
Shell Object Commands  
Port Object Commands. Port object commands provide information and  
perform actions related to a drive on a specific port. For example, you use port  
object commands for such tasks as seeing the status, model, or serial number  
of the drive. For details, see “Port Object Commands” on page 69.  
BBU Object Commands. BBU object commands provide information and  
perform actions related to a Battery Backup Unit on a specific controller. For  
Enclosure Object Commands. Enclosure object commands provide  
information and perform actions related to a particular enclosure. For  
example, you can use enclosure object commands to see information about an  
enclosure and its elements (slots, fan, and temperature sensor elements).  
Help Commands. Help commands allow you to display help information for  
all commands and attributes. For details, see “Help Commands” on page 83.  
Shell Object Commands  
Shell object commands are either applicable to all the controllers in the  
system (such as show, rescan, flush, commit), or redirect the focused object.  
Syntax  
focus object  
show [attribute [modifier]]  
ver  
alarms [reverse]  
diag  
rebuild  
verify  
selftest  
rescan  
flush  
commit  
update fw=filename_with_path [force]  
focus Object  
The focus command is active in interactive mode only and is provided to  
reduce typing.  
The focus command will set the specified object in focus and change the  
prompt to reflect this. This allows you to enter a command that applies to the  
focus, instead of having to type the entire object name each time.  
For example, where normally you might type:  
//hostname/c0/u0 show  
19  
   
Chapter 2. CLI Syntax Reference  
if you set the focus to //hostname/c0/u0, the prompt changes to reflect that,  
and you only have to type show. The concept is similar to being in a particular  
location in a file system and requesting a listing of the current directory.  
object can have the following forms:  
//hostname/cx/uxspecifies the fully qualified URI (Universal Resource  
Identifier) of an object on host hostname, controller cx, unit ux.  
//hostnamespecifies the root of host hostname.  
..specifies one level up (the parent object).  
/ specifies the root at the current focused hostname.  
./object specifies the next level of the object.  
/c0/bbu specifies a relative path with respect to the current focused  
hostname.  
Example:  
//localhost> focus //elvis.amcc.com  
//elvis.amcc.com>  
//elvis.amcc.com> focus /c0/u0  
//elvis.amcc.com/c0/u0>  
//elvis.amcc.com/c0/u0> focus..  
//elvis.amcc.com/c0>  
//elvis.amcc.com/c0/> focus u0  
//elvis.amcc.com/c0/u0>  
//elvis.amcc.com/c0> focus /  
//elvis.amcc.com>  
show  
This command shows a general summary of all detected controllers.  
Note that the device drivers for the appropriate operating system should be  
loaded for the list to show all controllers. The intention is to provide a global  
view of the environment.  
Example:  
Typical output of the Show command looks like the following:  
//localhost> show  
Ctl  
-----------------------------------------------------------------------  
c0 9590SE-4ME  
Model  
Ports  
Drives Units  
NotOpt  
RRate  
VRate  
BBU  
4
4
1
0
2
5
-
The output above indicates that Controller 0 is a 9590SE model with 4 Ports,  
with 4 Drives detected (attached), total of 1 Unit, with no units in a NotOpt  
20  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
 
Shell Object Commands  
(Not Optimal) state, RRate (Rebuild Rate) of 2, VRate (Verify Rate) of 5,  
BBU of '-' (Not Applicable). Not Optimal refers to any state except OK and  
VERIFYING. Other states include VERIFY-PAUSED, INITIALIZING, INIT-  
PAUSED, REBUILDING, REBUILD-PAUSED, DEGRADED,  
MIGRATING, MIGRATE-PAUSED, RECOVERY, INOPERABLE, and  
UNKNOWN. RRate also applies to initializing, migrating, and recovery  
background tasks. (Definitions of the unit statuses are available in the 3ware  
Serial ATA RAID Controller User Guide.)  
For a system with an enclosure unit such as the 3ware Sidecar that includes  
support for an EPCT (Enclosure Port Configuration Table), applicable  
firmware and software, and an appropriate controller (9550SX, 9590SE, or  
9650SE), a global view of the environment also includes summary  
information about detected enclosures.  
Example:  
Typical output of the Show command for a system with an enclosure looks  
like the following:  
//localhost> show  
Ctl  
Model  
Ports  
Drives Units  
NotOpt  
RRate  
VRate  
BBU  
---------------------------------------------------------------------------  
c0  
9590SE-4ME  
4
2
1
0
4
4
-
Encl  
Slots Drives Fans  
TSUnits  
Ctls  
---------------------------------------------------  
e0  
4
2
1
1
1
The output above shows the enclosure summary information with the name of  
the enclosure, the protocol used, the number of drive slots, the number of  
drives, the number of fans, the number of temperature sensors, and the  
number of controllers that are associated with the enclosure.  
show ver  
This command will show the CLI and API version.  
Example:  
//localhost> show ver  
CLI Version = 2.00.00.0xx  
API Version = 2.00.00.0xx  
where xx is the actual version. See the Release Notes for details.  
21  
 
Chapter 2. CLI Syntax Reference  
show alarms [reverse]  
This command shows the alarms or AEN messages of all controllers in the  
system. The default is to display the most recent message first. The reverse  
attribute displays the most recent message last.  
show diag  
This command shows the diagnostic information of all controllers in the  
system.  
show rebuild  
This command displays all rebuild schedules for the 9000 controllers in the  
system.  
The rebuild rate is also applicable for initializing, migrating, and recovery  
background tasks.  
Example:  
//localhost> show rebuild  
Rebuild Schedule for Controller /c0  
========================================================  
Slot  
Day  
Hour  
Duration  
Status  
--------------------------------------------------------  
1
Sun  
Mon  
Tue  
Wed  
Thu  
Fri  
Sat  
12:00am  
12:00am  
12:00am  
12:00am  
12:00am  
12:00am  
12:00am  
24 hr(s)  
24 hr(s)  
24 hr(s)  
24 hr(s)  
24 hr(s)  
24 hr(s)  
24 hr(s)  
disabled  
disabled  
disabled  
disabled  
disabled  
disabled  
disabled  
2
3
4
5
6
7
For additional information about rebuild schedules, see “/cx add  
rebuild=ddd:hh:duration” on page 45, and see the discussion of background  
tasks and schedules in 3ware Serial ATA RAID Controller User Guide.  
22  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
     
Shell Object Commands  
show verify  
This command displays all verify schedules for the 9000 controllers in the  
system.  
Example:  
//localhost> show verify  
Verify Schedule for Controller /c0  
========================================================  
Slot  
Day  
Hour  
Duration  
Status  
--------------------------------------------------------  
1
2
3
4
5
6
7
Sat  
-
-
-
-
11:00pm  
4 hr(s)  
enabled  
enabled  
enabled  
enabled  
enabled  
enabled  
enabled  
-
-
-
-
-
-
-
-
For additional information about verify schedules, see “/cx add  
verify=ddd:hh:duration” on page 46, and see the discussion of background  
tasks and schedules in 3ware Serial ATA RAID Controller User Guide.  
show selftest  
This command displays all selftest schedules for the 9000 controllers in the  
system.  
Example:  
//localhost> show selftest  
Selftest Schedule for Controller /c0  
========================================================  
Slot  
Day  
Hour  
UDMA  
SMART  
--------------------------------------------------------  
1
2
3
4
5
6
7
Sun  
Mon  
Tue  
Wed  
Thu  
Fri  
Sat  
12:00am  
12:00am  
12:00am  
12:00am  
12:00am  
12:00am  
12:00am  
enabled  
enabled  
enabled  
enabled  
enabled  
enabled  
enabled  
enabled  
enabled  
enabled  
enabled  
enabled  
enabled  
enabled  
For additional information about selftest schedules, see “/cx add  
selftest=ddd:hh” on page 47, and see the discussion of background tasks and  
schedules in 3ware Serial ATA RAID Controller User Guide.  
23  
   
Chapter 2. CLI Syntax Reference  
flush  
This command sends a flush command to all 3ware controllers in the system.  
rescan  
commit  
This command sends a rescan command to all 3ware controllers in the system.  
This command sends a commit command to all 3ware controllers in the  
system. For more information, see “/cx commit” on page 39.  
update fw=filename_with_path [force]  
This command downloads the specified firmware image to the controllers that  
are compatible with it and iterates through all the controllers in the system,  
updating the firmware. For more information, see “/cx update  
24  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
       
Controller Object Commands  
Controller Object Commands  
Controller object commands provide information and perform actions related  
to a specific controller, such as /c0. For example, you use controller object  
commands to see alarms specific to a controller, to create schedules for when  
background tasks are run, and to set policies for the controller. You also use  
the controller object command /cx add type to create RAID arrays.  
Syntax  
/cx show  
/cx show attribute [attribute ...] where attributes are:  
achip|alarms|allunitstatus|autocarve (9000 series)|  
autorebuild(9550SX and 9590SE only)|bios|  
carvesize(9000 series)|ctlbus(9550SX and 9590SE only)|  
diag|driver| drivestatus|exportjbod(9000 series)|  
firmware|memory|model|monitor|ondegrade(9500S series)  
|pcb|pchip|numdrives|numports|numunits|  
rebuild(9000 series)|selftest(9000 series)|serial|  
spinup|stagger|unitstatus|verify(9000 series)  
/cx show all (where all means attributes and configurations)  
/cx show diag  
/cx show alarms [reverse]  
/cx show rebuild  
/cx show verify  
/cx show selftest  
(9000 series)  
(9000 series)  
(9000 series)  
/cx add type=<RaidType> disk=<p:-p..> [stripe=<Stripe>]  
[noscan][nocache][group=<3|4|5|6>][autoverfy][ignorECC]  
[name=string](9000 series) RaidType={raid0,raid1,raid5,  
raid10,raid50,single,spare,jbod(7000/8000 only)}  
[storsave=<protect|balance|perform>(9550SX and 9590SE  
only)]  
/cx add rebuild=ddd:hh:duration  
/cx add verify=ddd:hh:duration  
/cx add selftest=ddd:hh  
(9000 only)  
(9000 only)  
(9000 only)  
/cx del rebuild=slot_id  
/cx del verify=slot_id  
/cx del selftest=slot_id  
(9000 only)  
(9000 only)  
(9000 only)  
/cx set exportjbod=on|off  
(9000 only)  
/cx set ondegrade=cacheoff|follow (9500S only)  
/cx set spinup=nn  
/cx set stagger=nn  
/cx set autocarve=on|off  
/cx set carvesize=[1024...2048]  
(9000 only)  
(9000 only)  
(9000 only)  
(9000 only)  
/cx set rebuild=enable|disable|1..5 (9000 only)  
/cx set autorebuild=on|off (9550SX and 9590SE only)  
25  
 
Chapter 2. CLI Syntax Reference  
/cx set verify=enable|disable|1..5 (9000 only)  
/cx set selftest=enable|disable [task=UDMA|SMART](9000 only)  
/cx flush  
/cx update filename_with_path [force] (9000 only)  
/cx commit  
(Windows only. Also known as shutdown)  
/cx start mediascan (7000/8000 only)  
/cx stop mediascan (7000/8000 only)  
/cx rescan [noscan] (Does not import non-JBOD on 7000/8000  
models.  
/cx show  
This command shows summary information on the specified controller /cx.  
This information is organized into a report containing two to three parts:  
A Unit summary listing all present units  
A Port summary section listing of all ports and disks attached to them.  
A BBU summary section listing, if a BBU is installed on the controller.  
The Unit summary section lists all present unit and specifies their unit  
number, unit type (such as RAID 5), unit status (OK, RECOVERY,  
INOPERABLE, UNKNOWN, DEGRADED, INITIALIZING, INIT-  
PAUSED, VERIFYING, VERIFY-PAUSED, REBUILDING, REBUILD-  
PAUSED, MIGRATING, MIGRATE-PAUSED), % Cmpl (of rebuilding,  
verifying, initializing, or migrating), stripe size, size (usable capacity) in  
gigabytes or terabytes, auto verify policy status (on/off), and ignore ECC  
errors policy state (on/off). (Definitions of the unit statuses are available  
in the 3ware Serial ATA RAID Controller User Guide.)  
Note: If an asterisk (*) appears next to the status of a unit, there is an error on one  
of the drives in the unit. This feature provides a diagnostic capability for potential  
problem drives. The error may not be a repeated error, and may be caused by an  
ECC error, SMART failure, or a device error. Rescanning the controller will clear the  
drive error status if the condition no longer exists.  
The Port summary section lists all present ports and specifies the port  
number, disk status, unit affiliation, size (in gigabytes) and blocks (512 bytes),  
and the serial number assigned by the disk vendor.  
The BBU summary lists details about the BBU, if one is installed. It includes  
a few important attributes such as hours left (in which the current BBU can  
backup the controller cache in the event of power loss), temperature, voltage,  
readiness, and so forth.  
Additional attributes about controllers, units, ports and disks can be obtained  
by querying for them explicitly. For details, see the other show subcommands.  
26  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
 
Controller Object Commands  
Typical output looks like:  
//localhost> /c0 show  
Unit  
UnitType Status  
%Cmpl  
Stripe Size(GB) Cache AVerify IgnECC  
--------------------------------------------------------------------------  
u0  
u1  
u2  
RAID-1  
RAID-5  
SPARE  
OK  
OK  
OK  
-
-
-
-
64k  
-
149.05  
298.22  
149.05  
ON  
ON  
ON  
OFF  
OFF  
OFF  
OFF  
OFF  
-
Port  
Status  
Unit  
Size  
Blocks  
Serial  
---------------------------------------------------------------  
p0  
p1  
p2  
p3  
p4  
p5  
p6  
p7  
p8  
p9  
p10  
p11  
OK  
OK  
OK  
OK  
OK  
OK  
OK  
OK  
u0  
u0  
u1  
u1  
u1  
u2  
-
-
-
-
-
149.05 GB  
149.05 GB  
149.05 GB  
149.05 GB  
149.05 GB  
149.05 GB  
149.05 GB  
149.05 GB  
149.05 GB  
-
312581808  
312581808  
312581808  
312581808  
312581808  
312581808  
312581808  
312581808  
312581808  
-
3JS0TF14  
3JS0TETZ  
3JS0VG85  
3JS0VGCY  
3JS0VGGQ  
3JS0VH1P  
3JS0TF0P  
3JS0VF43  
3JS0VG8D  
-
OK  
NOT-PRESENT  
NOT-PRESENT  
NOT-PRESENT  
-
-
-
-
-
-
-
Name OnlineState BBUReady Status  
---------------------------------------------------------------------------  
bbu On Yes OK OK OK 241 22-Jun-2004  
Volt  
Temp  
Hours LastCapTest  
/cx show attribute [attribute ...]  
This command shows the current setting of the specified attributes on the  
specified controller. One or many attributes can be specified. Specifying an  
invalid attribute will terminate the loop. Possible attributes are: achip,  
allunitstatus, autocarve (9000 series), autorebuild (9550SX and 9590SE only),  
bios, carvesize (9000 series), driver, drivestatus, exportjbod (9000 series),  
firmware, memory, model, monitor, numdrives, numports, numunits, ctlbus  
(9550SX and 9590SE only), ondegrade (9500S), pcb, pchip, qpolicy, serial,  
spinup (9000 series), stagger (9000 series), and unitstatus.  
Example: To see the driver and firmware installed on controller 0, enter the  
following:  
//localhost> /c0 show driver firmware  
/c0 Driver Version = 2.x  
/c0 Firmware Version = FE9X 2.x  
Where x=actual version  
27  
 
Chapter 2. CLI Syntax Reference  
/cx show driver  
This command reports the device driver version associated with controller  
/cx.  
Example:  
//localhost> /c0 show driver  
/c0 Driver Version = 2.x  
/cx show model  
This command reports the controller model of controller /cx.  
Example:  
//localhost> /c0 show model  
/c0 Model = 9500-x  
/cx show firmware  
This command reports the firmware version of controller /cx.  
Example:  
//localhost> /c0 show firmware  
/c0 Firmware Version = FGXX 2.x  
/cx show bios  
This command reports the BIOS version of controller /cx.  
Example:  
//localhost> /c0 show bios  
/c0 BIOS Version = BG9X 2.x  
/cx show monitor  
This command reports the monitor (firmware boot-loader) version of  
controller /cx.  
Example:  
//localhost> /c0 show monitor  
/c0 Monitor Version = BLDR 2.x  
28  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
         
Controller Object Commands  
/cx show serial  
This command reports the serial number of the specified controller /cx.  
Example:  
//localhost> /c0 show serial  
/c0 Serial Number = F12705A3240009  
/cx show pcb  
This command reports the PCB (Printed Circuit Board) version of the  
specified controller /cx.  
Example:  
//localhost> /c0 show pcb  
/c0 PCB Version = RevX  
/cx show pchip  
This command reports the PCHIP (PCI Interface Chip) version of the  
specified controller /cx.  
Example:  
//localhost> /c0 show pchip  
/c0 PCHIP Version = 1.x  
/cx show achip  
This command reports the ACHIP (ATA Interface Chip) version of the  
specified controller /cx.  
Example:  
//localhost> /c0 show achip  
/c0 ACHIP Version = 3.x  
/cx show numdrives  
This command reports the number of drives currently managed by the specified  
controller /cx. This report does not include (logically) removed or exported drives.  
Also note that physically removed disk(s) will not be detected unless I/O is  
performed against the disk. See “/cx/px show smart” on page 71 for a workaround.  
Example:  
//localhost> /c0 show numdrives  
/c0 Number of Drives = 5  
29  
         
Chapter 2. CLI Syntax Reference  
/cx show numports  
This command reports the port capacity (number of physical ports) of the  
specified controller /cx.  
Example:  
//localhost> /c0 show numports  
/c0 Number of Ports = 12  
/cx show numunits  
This command reports the number of units currently managed by the specified  
controller /cx. This report does not include off-line units (or removed units).  
Example:  
//localhost> /c0 show numunits  
/c0 Number of Units = 1  
/cx show ctlbus  
This feature only applies to 9550SX and 9590SE controllers.  
This command reports the controller host bus type, bus speed, and bus width.  
Example:  
//localhost> /c0 show ctlbus  
/c0 Controller Bus Type = PCIX  
/c0 Controller Bus Width = 64 bits  
/c0 Controller Bus Speed = 133 Mhz  
/cx show exportjbod  
This feature only applies to 9000 series controllers.  
This command reports the current JBOD Export Policy: on, off, or Not  
Supported.  
Example:  
//localhost> /c0 show exportjbod  
/c0 JBOD Export Policy = Not Supported.  
//localhost> /c1 show exportjbod  
/c1 JBOD Export Policy = on  
30  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
       
Controller Object Commands  
/cx show spinup  
This feature only applies to 9000 series controllers.  
This command reports the number of concurrent disks that will spin up when  
the system is powered on, after waiting for the number of seconds specified  
with the set staggercommand.  
Example:  
//localhost> /c0 show spinup  
/c0 Disk Spinup Policy = 1  
/cx show stagger  
This feature only applies to 9000 series controllers.  
This command reports the time delay between each group of spinups at the  
power on.  
Example:  
//localhost> /c0 show stagger  
/c0 Spinup Stagger Time Policy (sec) = 2  
/cx show ondegrade  
This feature only applies to 9500S controllers.  
This command reports the cache policy for degraded units. If the ondegrade  
policy is “Follow Unit Policy,” a unit cache policy stays the same when the  
unit becomes degraded. If the ondegrade policy is off, a unit cache policy will  
be forced to “off” when the unit becomes degraded.  
Example:  
//localhost> /c0 show ondegrade  
/c0 Cache on Degraded Policy = Follow Unit Policy  
/cx show autocarve  
This feature only applies to 9000 series controllers.  
This command reports the Auto-Carve policy. If the policy is on, all newly  
created or migrated units larger than the carvesize will be automatically  
carved into multiples of carvesize volumes plus one remainder volume. Each  
volume can be treated as an individual drive with its own file system. The  
default carvesize is 2TB. For more information see, “/cx show carvesize”,  
below.  
31  
       
Chapter 2. CLI Syntax Reference  
For operating systems that support units larger than 2TB, there is no need to  
set the policy to on unless you want the operating system to have multiple  
smaller volumes.  
If you use a 32-bit operating system, it is recommended that you keep the  
policy on unless you know that your operating system supports disks that are  
larger than 2 TB.  
When the autocarve policy is off, all newly created units will consist of one  
single volume.  
Example:  
//localhost> /c0 show autocarve  
/c0 Auto-Carving Policy = on  
/cx show carvesize  
This feature only applies to 9000 series controllers.  
This command shows the maximum size of the volumes that will be created if  
the autocarve policy is set to on. The carvesize can be set between 1024 GB  
and 2048 GB. Default carvesize is 2048 GB (2 TB). For more information  
Example:  
//localhost> /c0 show carvesize  
/c0 Auto-Carving Size = 2000 GB  
/cx show memory  
This command reports the size of the memory installed on the controller.  
Note: The 9500S controllers ship with 128 MBytes of cache, yet only 112MB  
shows as memory installed. The other 16 MB is reserved for use by the controller.  
Example:  
//localhost> /c0 show memory  
/c0 Memory Installed = 112MB  
32  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
   
Controller Object Commands  
/cx show autorebuild  
This feature only applies to 9550SX and 9590SE model controllers.  
This command shows the Auto-Rebuild policy. If the policy is enabled, the  
firmware will select drives to use for rebuilding a degraded unit using the  
following priority order. For more information, see “/cx set  
1. Smallest usable spare.  
2. Smallest usable unconfigured (available) drive.  
3. Smallest usable failed drive.  
If the policy is disabled, only spare drives will be used for an automatic  
rebuild operation.  
Example:  
//localhost> /c0 show autorebuild  
/c0 Auto-Rebuild Policy = on  
/cx show unitstatus  
This command presents a list of units currently managed by the specified  
controller /cx, and shows their types, capacity, status, and unit policies.  
Possible statuses include: OK, VERIFYING, VERIFY-PAUSED,  
INITIALIZING, INIT-PAUSED, REBUILDING, REBUILD-PAUSED,  
DEGRADED, MIGRATING, MIGRATE-PAUSED, RECOVERY,  
INOPERABLE, and UNKNOWN. (Definitions of the unit statuses are  
available in the 3ware Serial ATA RAID Controller User Guide.)  
Note: If an asterisk (*) appears next to the status of a unit, there is an error on one  
of the drives in the unit. This feature provides a diagnostic capability for potential  
problem drives. The error may not be a repeated error, and may be caused by an  
ECC error, SMART failure, or a device error. Rescanning the controller will clear the  
drive error status if the condition no longer exists.  
Example:  
//localhost> /c0 show unitstatus  
Unit UnitType Status  
%Cmpl Stripe Size(GB) Cache AVerify IgnECC  
-------------------------------------------------------------------------  
u0  
u1  
u2  
RAID-5  
JBOD  
RAID-0  
OK  
OK  
OK  
-
-
-
64K  
-
64K  
223.485  
76.3352  
148.99  
OFF  
OFF  
ON  
OFF  
OFF  
ON  
ON  
-
-
33  
   
Chapter 2. CLI Syntax Reference  
/cx show allunitstatus  
This command presents a count of total and Not Optimal units managed by  
the specified controller /cx. For more about the meaning of Not Optimal, see  
Example:  
//localhost> /c0 show allunitstatus  
/c0 Total Optimal Units = 2  
/c0 Not Optimal Units = 0  
/cx show drivestatus  
This command reports a list of drives and their port assignment, status, the  
unit with which they are associated, their size in gigabytes and blocks, and the  
serial number assigned by the drive manufacturer. (Definitions of the drive  
statuses are available in the 3ware Serial ATA RAID Controller User Guide.)  
Example:  
//localhost> /c0 show drivestatus  
Port Status  
Unit Size  
Blocks  
Serial  
--------------------------------------------------------------  
p0  
p1  
p2  
p3  
p4  
p5  
p6  
p7  
p8  
p9  
p10  
p11  
OK  
OK  
OK  
OK  
OK  
OK  
OK  
OK  
u0  
u0  
u1  
u1  
u1  
u2  
-
149.05 GB 312581808  
149.05 GB 312581808  
149.05 GB 312581808  
149.05 GB 312581808  
149.05 GB 312581808  
149.05 GB 312581808  
149.05 GB 312581808  
149.05 GB 312581808  
149.05 GB 312581808  
3JS0TF14  
3JS0TETZ  
3JS0VG85  
3JS0VGCY  
3JS0VGGQ  
3JS0VH1P  
3JS0TF0P  
3JS0VF43  
3JS0VG8D  
-
-
-
OK  
NOT-PRESENT  
NOT-PRESENT  
NOT-PRESENT  
-
-
-
-
-
-
-
-
-
-
-
34  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
   
Controller Object Commands  
/cx show all  
This command shows the current setting of all of the following attributes on  
the specified controller: driver, model, firmware, memory, bios, monitor,  
serial, pcb, pchip, achip, numports, numunits, numdrives, unitstatus,  
drivestatus, allunitstatus, exportjbod, ondegrade, spinup, autocarve, and  
stagger.  
Example: (where x represents the actual version number)  
//localhost> /c0 show all  
/c0 Driver Version = 3.x  
/c0 Model = 9550SX-12  
/c0 Memory Installed = 112MB  
/c0 Firmware Version = FE9X 3.x  
/c0 Bios Version = BE9X 3.x  
/c0 Monitor Version = BL9X 3.x  
/c0 Serial Number = xxxxx  
/c0 PCB Version = Rev 0xx  
/c0 PCHIP Version = 1.xx  
/c0 ACHIP Version = 3.xx  
/c0 Number of Ports = 12  
/c0 Number of Units = 2  
/c0 Number of Drives = 12  
/c0 Total Optimal Units = 2  
/c0 NotOptimalUnits = 0  
/c0 Total Units = 2  
/c0 JBOD Export Policy = off  
/c0 Disk Spinup Policy = 7  
/c0 Spinup Stagger Time Policy (sec) = 4  
/c0 Cache on Degrade Policy = Follow Unit Policy  
/c0 Auto-Carving Policy = off  
/c0 Auto-Carving Size = 2047 GB  
/c0 Auto-Rebuild Policy = enable  
Unit UnitType Status  
%Cmpl Stripe Size(GB) Cache AVerify IgnECC  
--------------------------------------------------------------------------  
u0  
u1  
RAID-5  
RAID-5  
OK  
OK  
-
-
256K  
256K  
148.99  
595.961  
ON  
ON  
ON  
OFF  
ON  
ON  
Port  
Status  
Unit  
Size  
Blocks  
Serial  
---------------------------------------------------------------  
p0  
p1  
p2  
p3  
p4  
p5  
p6  
p7  
p8  
p9  
p10  
p11  
OK  
OK  
OK  
OK  
OK  
OK  
OK  
OK  
OK  
OK  
OK  
OK  
u0  
u0  
u0  
u1  
u1  
u1  
u1  
u1  
u1  
u1  
u1  
u1  
74.53 GB  
74.53 GB  
74.53 GB  
74.53 GB  
74.53 GB  
74.53 GB  
74.53 GB  
111.79 GB  
111.79 GB  
111.79 GB  
111.79 GB  
111.79 GB  
156301488  
156301488  
156301488  
156301488  
156301488  
156301488  
156301488  
234441648  
234441648  
234441648  
234441648  
234441648  
3JV3MV1C  
3JV3MK6B  
3JV3LW52  
3JV49S77  
3JV3MVTA  
5JV980Z0  
5JV9820G  
WD-WMAEL10275  
WD-WMAEL10274  
WD-WMAEL10281  
WD-WMAEL10273  
WD-WMAEL10274  
35  
 
Chapter 2. CLI Syntax Reference  
/cx add type=<RaidType> disk=<p:-p>  
[stripe=Stripe] [noscan] [group=<3|4|5|6>]  
[nocache] [autoverify] [ignoreECC] [name=string]  
[storsave=<protect|balance|perform>]  
This command allows you to create a new unit on the specified controller. You  
specify type, disks, and optional stripe size. By default the host operating  
system will be informed of the new block device, write cache will be enabled,  
and a storsave policy of protect will be set. In case of RAID 50, you can also  
specify the layout of the unit by specifying the number of disks per disk group  
with the groupattribute.  
/cx is the controller name, for example /c0, /c1, and so forth.  
type=RaidType specifies the type of RAID unit to be created. Possible unit  
types include raid0, raid1, raid5, raid10, raid50, single, spare, and JBOD  
(7000/8000 only).  
Example: type=raid5  
When a new unit is created, it is automatically assigned a unique serial  
number. In addition, users can assign the unit a name.  
Note: The unit’s serial number cannot be changed.  
disk=p:-p consists of a list of ports (disks) to be used in the construction of  
the specified unit type. One or more ports can be specified. Multiple ports can  
be specified using a colon (:) or a dash (-) as port index separators. A dash  
indicates a range and can be mixed with colons. For example disk=0:1:2-  
5:9:12indicates port 0, 1, 2 through 5 (inclusive), 9 and 12.  
stripe=Stripe consists of the stripe size to be used. The following table  
illustrates the supported and applicable stripes on unit types and controller  
models. Stripe size units are in K (kilobytes). If no stripe size is specified,  
64K is used by default, if applicable. If you need to change the stripe size after  
the unit is created, you can do so my migrating the unit.  
Table 5: Supported Stripe Sizes  
Model R0  
R1  
R5  
R10  
Spare  
Single  
9K  
16  
N/A  
16  
16  
N/A  
N/A  
64  
64  
64  
256  
256  
256  
36  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
 
Controller Object Commands  
group=3|4|5|6 indicates the number of disks per group for a RAID 50 type.  
(This attribute can only be used when type=raid50.) Recall that a RAID 50 is  
a multi-tier array. At the bottom-most layer, N number of disks per group are  
used to form the RAID 5 layer. These RAID 5 arrays are then integrated into a  
RAID 0. This attribute allows you to specify the number of disks in the RAID  
5 level. Valid values are 3, 4, 5 and 6.  
Note that a sufficient number of disks are required for a given pattern or disk  
group. For example, given 6 disks, specifying 3 will create two RAID 5  
arrays. With 12 disks, specifying 3 will create four RAID 5 arrays under the  
RAID 0 level. With only 6 disks a grouping of 6 is not allowed, as you would  
basically be creating a RAID 5.  
The default RAID 50 grouping varies, based on number of disks. For 6 and 9  
disks, default grouping is 3. For 8 disks, the default grouping is 4. For 10  
disks, the default grouping is 5, and for 12 disks, the default grouping is 4. In  
the case of 12, the disks could be grouped into groups of 3, 4, or 6 drives. A  
grouping of 4 is set by default as it provides the best of net capacity and  
performance.  
noscan attribute instructs CLI not to notify the operating system of the  
creation of the new unit. By default CLI will inform the operating system.  
One application of this feature is to prevent the operating system from  
creating block special devices such as /dev/sdb and /dev/sdc as some  
implementations might create naming fragmentation and a moving target.  
nocache attribute instructs CLI to disable the write cache on the newly  
created unit. Enabling write cache increases write performance at the cost of  
potential data loss in case of sudden power loss (unless a BBU or UPS is  
installed). By default the cache is enabled. To avoid the possibility of data loss  
in the event of a sudden power loss, it is recommended not to set nocache  
unless there is a BBU (battery backup unit) or UPS (uninterruptable power  
supply) installed.  
autoverify attribute enables the autoverify attribute on the unit that is to be  
created. For more details on this feature, see “/cx/ux set autoverify=on|off” on  
page 61. This feature is not supported on model 7000/8000. On model 9000,  
the JBOD autoverify attribute is not persistent (does not survive reboots).  
ignoreECC attribute enables the ignoreECC/OverwriteECC attribute on the  
unit that is to be created. For more details on this feature, see “/cx/ux set  
ignoreECC=on|off” on page 62. The following table illustrates the supported  
Model-Unit Types. This table only applies to setting this feature at unit  
creation time. IgnoreECC only applies to redundant units. For the 7/8000  
series, this setting is only applicable during rebuild; it is not applicable  
during creation.  
37  
Chapter 2. CLI Syntax Reference  
Table 6: Supported Model-Unit Types for ignoreECC  
Model  
R-0  
R-1  
R-5  
R-10  
R-50  
Single  
JBOD  
Spare  
7K/8K  
9K  
No  
No  
No  
No  
No  
No  
No  
No  
No  
No  
No  
No  
Yes  
Yes  
Yes  
Yes  
name=string attribute allows you to name the new unit. (This feature is for  
9000 series and above controllers.) The string can be up to 21 characters and  
cannot contain spaces. In order to use reserved characters (‘<‘, ‘>’, ‘!’, ‘&’,  
etc.) put double quotes (" ") around the name string. The name can be changed  
after the unit has been created. For more information, see “/cx/ux set  
storsave=protect|balance|perform attribute allows user to set the storsave  
policy of the new unit. This feature is only for 9550SX and 9590SE  
controllers. For more information, see “/cx/ux set  
/cx rescan [noscan]  
This command instructs the controller to rescan all ports and reconstitute all  
units. The controller will update its list of ports (attached disks), and attempts  
to read every DCB (Disk Configuration Block) in order to re-assemble its  
view and awareness of logical units. Any newly found unit(s) or drive(s) will  
be listed.  
noscan is used to not inform the operating system of the unit discovery. The  
default is to inform the operating system.  
Note: If you are adding new drives, add them physically before issuing the rescan  
commands. Hot swap carriers are required unless you first power-down the system  
to prevent system hangs and electrical damage.  
Example:  
//localhost> /c1 rescan  
Rescanning controller /c1 for units and drives ...Done  
Found following unit(s): [/c1/u3]  
Found following drive(s): [/c1/p7, /c1/p8]  
Note: Rescanning does not import non-JBOD on 7000/8000 models.  
38  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
 
Controller Object Commands  
/cx commit  
/cx flush  
This command only applies to the Windows operating system. It commits all  
changes if a faster shutdown method is needed when running certain database  
applications. Linux and FreeBSD file systems do not require this command  
since they have their own ways of notifying the controller to do clean up for  
shut down.  
This command forces the controller to write all cached data to disk for the  
specified controller.  
/cx update fw=filename_with_path [force]  
This command is only for 9000 series controllers.  
This command allows the downloading of the specified firmware image to the  
corresponding controller.  
Note: Before issuing this command, you must have already obtained the  
firmware image and placed it on your system. You can obtain the firmware  
image from the 3ware website: http:www.3ware.com/downloads.  
fw=filename_with_path attribute allows you to specify the firmware image  
file name along with its absolute path. The new image specified by this  
filename_with_path is checked for compatibility with the current controller,  
current driver, and current application versions. A recommendation is then  
made as to whether an update is needed, and you are asked to confirm whether  
you want to continue. If you confirm that you want to continue, the new  
firmware image is downloaded to the specified controller.  
A reboot is required for the new firmware image to take effect.  
Example:  
//localhost> /c0 update fw=/tmp/prom0006.img  
Warning: We strongly recommend backing up your data before  
updating the firmware. Updating the firmware can render the  
device driver and/or management tools incompatible. It is  
recommended to have a copy of current firmware image for  
rollbacks.  
Examining compatibility data from firmware image and /c0 ...  
Done.  
New-Firmware  
----------------------------------------------------------------------  
FE9X 3.05.00.005 FE9X 3.05.00.005 2.26.04.007 2.01.00.008  
Current-Firmware  
Current-Driver  
Current-API  
Current firmware version is the same as the new firmware.  
39  
     
Chapter 2. CLI Syntax Reference  
Recommendation: No need to update.  
Given the above recommendation...  
Do you want to continue ? Y|N [N]: y  
Downloading the firmware from file /tmp/prom0006.img ... Done.  
The new image will take effect after reboot.  
force attribute is optional. If you include it, the compatibility checks are  
bypassed.  
/cx show alarms [reverse]  
Asynchronous events (also referred to as AENs or alarms) are originated by  
firmware and captured by their respective device drivers. These events reflect  
warning, debugging, and/or informative messages for the end user. These  
events are kept in a finite queue inside the kernel, awaiting extraction by user  
space programs such as CLI and/or 3DM.  
The /cx show alarms command displays all available alarms on a given  
controller. The default is to display the most recent alarm or AEN message  
first. The user can also use the [reverse] attribute to display the most recent  
alarm or AEN message last.  
Alarms generated on 7000/8000 controllers do not have dates, so you will see  
a '-' in the Date column. This means that it is not applicable. In addition,  
alarm messages on 7000/8000 controllers contain the severity in the message  
text, so the Severity column also shows a '-'.  
Typical output looks like:  
tw_cli> /c1 show alarms reverse  
Ctl Date  
--------------------------------------------------------------------------------------------  
Severity Message  
c1 [Fri Nov 28 04:26:31 2003] ERROR  
c1 [Fri Nov 28 06:13:54 2003] INFO  
c1 [Fri Nov 28 06:30:35 2003] INFO  
c1 [Fri Nov 28 06:33:00 2003] ERROR  
c1 [Fri Nov 28 06:33:04 2003] ERROR  
c1 [Fri Nov 28 06:33:46 2003] INFO  
c1 [Fri Nov 28 06:37:58 2003] INFO  
c1 [Fri Nov 28 07:51:34 2003] INFO  
c1 [Fri Nov 28 07:59:43 2003] INFO  
c1 [Mon Dec 1 02:26:12 2003] ERROR  
(0x04:0x0002): Degraded unit detected:unit=0, port=2  
(0x04:0x000B): Rebuild started: unit=0  
(0x04:0x003B): Background rebuild paused:unit=0  
(0x04:0x0002): Degraded unit detected:unit=0, port=0  
(0x04:0x0002): Degraded unit detected:unit=0, port=4  
(0x04:0x000B): Rebuild started: unit=0  
(0x04:0x000B): Rebuild started: unit=0  
(0x04:0x0005): Background rebuild done:unit=0  
(0x04:0x0005): Background rebuild done:unit=0  
(0x04:0x0002): Degraded unit detected:unit=0, port=3  
/cx show diag  
This command extracts controller diagnostics suitable for technical support  
usage. Note that some characters might not be printable or rendered correctly  
(human readable). It is recommended to save the output from this command to  
a file, where it can be communicated to technical support or further studied  
with Linux utilities such as od(1).  
40  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
   
Controller Object Commands  
In order to redirect the output you must run the following command from a  
command line, not from within the tw_cli shell.  
tw_cli /c0 show diag > diag.txt  
/cx show rebuild  
9000 series controllers support background tasks and allow you to schedule a  
regular time when they occur.  
Rebuild is one of the supported background tasks. Migrate and initialize are  
other background tasks that follow the same schedule as rebuild. Other  
background tasks for which there are separate schedules are verify and  
selftest. For each background task, up to 7 time periods can be registered,  
known as slots 1 through 7. Each task schedule can be managed by a set of  
commands including add, del, show and set a task. Background task  
schedules have a slot id, start-day-time, duration, and status attributes.  
For details about setting up a schedule for background rebuild tasks, see  
Rebuild activity attempts to (re)synchronize all members of redundant units  
such as RAID-1, RAID-10, RAID-5 and RAID-50. Rebuild can be started  
manually or automatically if a spare has been defined. Scheduled rebuilds will  
take place during the scheduled time slot, if enabled the schedules are  
enabled. For in depth information about rebuild and other background tasks,  
see “About Background Tasks” in the 3ware Serial ATA RAID Controller  
User Guide.  
The show rebuildcommand displays the current rebuild background task  
schedule as illustrated below.  
//localhost> /c1 show rebuild  
Rebuild Schedule for Controller /c1  
========================================================  
Slot  
Day  
Hour  
Duration  
Status  
--------------------------------------------------------  
1
2
3
4
5
6
7
Mon  
Thu  
-
-
-
2:00pm  
7:00pm  
-
-
10 hr(s)  
18 hr(s)  
-
-
disabled  
disabled  
disabled  
disabled  
disabled  
disabled  
disabled  
-
-
Mon  
Sun  
1:00am  
12:00am  
4 hr(s)  
1 hr(s)  
A status of “disabled” indicates that the task schedule is disabled. In this case,  
the controller will not use the defined schedule timeslots. If the rebuild  
command is entered manually, rebuilding will start within 10 to 15 minutes. It  
will begin automatically if a rebuild is needed and a proper spare drive is set  
up.  
41  
 
Chapter 2. CLI Syntax Reference  
If the rebuild schedule is enabled while a rebuild process is underway, the  
rebuild will pause until a scheduled time slot.  
Example:  
If a unit is in the initialization state at noon on Wednesday and the rebuild  
schedule shown above is in use (with schedules disabled), you would see the  
following status using the show command:  
//localhost>/c1show  
Unit UnitType Status  
%Cmpl Stripe Size(GB) Cache AVerify IgnECC  
--------------------------------------------------------------------------  
u0  
RAID-5  
Status  
INITIALIZING  
Unit  
0
64K  
521.466  
Blocks  
ON  
OFF  
OFF  
Port  
Size  
Serial  
---------------------------------------------------------------  
p0  
p1  
p2  
p3  
p4  
p5  
p6  
p7  
p8  
p9  
p10  
p11  
NOT-PRESENT  
OK  
NOT-PRESENT  
OK  
NOT-PRESENT  
OK  
NOT-PRESENT  
OK  
OK  
OK  
OK  
OK  
-
u0  
-
u0  
-
u0  
-
u0  
u0  
u0  
u0  
u0  
-
-
-
76.33 GB  
-
76.33 GB  
-
76.33 GB  
-
76.33 GB  
74.53 GB  
74.53 GB  
74.53 GB  
74.53 GB  
160086528  
-
160086528  
-
160086528  
-
160086528  
156301488  
156301488  
156301488  
156301488  
Y2NXL7FE  
-
Y2NXLB9E  
-
Y2NXQPZE  
-
Y2NXM4VE  
3JV3WTSE  
3JV3WRHC  
3JV3WQLQ  
3JV3WQLZ  
If you then enable the rebuild schedules, the unit initialization will be paused  
until the next scheduled time slot, as reflected in the examples below:  
//localhost> /c1 set rebuild=enable  
Enabling scheduled rebuilds on controller /c1 ...Done.  
//localhost> /c1 show rebuild  
Rebuild Schedule for Controller /c1  
========================================================  
Slot  
Day  
Hour  
Duration  
Status  
--------------------------------------------------------  
1
2
3
4
5
6
7
Mon  
Thu  
-
-
-
2:00pm  
7:00pm  
-
-
10 hr(s)  
18 hr(s)  
-
-
enabled  
enabled  
-
-
-
-
-
Mon  
Sun  
1:00am  
12:00am  
4 hr(s)  
1 hr(s)  
enabled  
enabled  
//localhost> /c1 show  
Unit UnitType Status  
%Cmpl Stripe Size(GB) Cache AVerify IgnECC  
--------------------------------------------------------------------------  
u0  
RAID-5  
Status  
INIT-PAUSED 0  
Unit  
64K  
521.466  
Blocks  
ON  
OFF  
Serial  
OFF  
Port  
Size  
---------------------------------------------------------------  
p0  
p1  
p2  
p3  
p4  
p5  
p6  
p7  
p8  
NOT-PRESENT  
OK  
NOT-PRESENT  
OK  
NOT-PRESENT  
OK  
NOT-PRESENT  
OK  
OK  
-
u0  
-
u0  
-
u0  
-
-
-
-
76.33 GB  
-
76.33 GB  
-
76.33 GB  
-
76.33 GB  
74.53 GB  
160086528  
-
160086528  
-
160086528  
-
160086528  
156301488  
Y2NXL7FE  
-
Y2NXLB9E  
-
Y2NXQPZE  
-
Y2NXM4VE  
3JV3WTSE  
u0  
u0  
42  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
Controller Object Commands  
p9  
p10  
p11  
OK  
OK  
OK  
u0  
u0  
u0  
74.53 GB  
74.53 GB  
74.53 GB  
156301488  
156301488  
156301488  
3JV3WRHC  
3JV3WQLQ  
3JV3WQLZ  
/cx show verify  
9000 series controllers support background tasks and allow you to schedule a  
regular time when they occur.  
Verify is one of the supported background tasks. Rebuild and selftest are other  
background tasks for which there are separate schedules. Migrate and  
initialize are additional background tasks that follow the same schedule as  
rebuild. For each background task, up to 7 time periods can be registered,  
known as slots 1 through 7. Each task schedule can be managed by a set of  
commands including add, del, show and set a task. Background task  
schedules have a slot id, start-day-time, duration, and status attributes.  
For details about setting up a schedule for background verify tasks, see  
Verify activity verifies all units based on their unit type. Verifying RAID-1  
involves checking that both drives contain the exact data. On RAID-5, the  
parity information is used to verify data integrity. RAID-10 and 50 are  
composite types and follow their respective array types. On 9000 series, non-  
redundant units such as RAID-0, JBOD, single, and spare, are also verified  
(by reading and reporting un-readable sectors). If any parity mismatches are  
found, the array will be automatically background initialized. (For  
information about the initialization process, see the user guide that came with  
your 3ware RAID controller.)  
The show verifycommand displays the current verify background task  
schedule as illustrated below.  
//localhost> /c1 show verify  
Verify Schedule for Controller /c1  
========================================================  
Slot  
Day  
Hour  
Duration  
Status  
--------------------------------------------------------  
1
2
3
4
5
6
7
Mon  
-
2:00am  
-
4 hr(s)  
-
disabled  
disabled  
disabled  
disabled  
disabled  
disabled  
disabled  
Tue  
Wed  
Thu  
Fri  
Sat  
12:00am  
12:00am  
12:00am  
12:00am  
12:00am  
24 hr(s)  
24 hr(s)  
24 hr(s)  
24 hr(s)  
24 hr(s)  
A status of “disabled” indicates that the controller will not use the defined  
schedule timeslots and will start verifying immediately (within 10 to 15  
minutes), if the verify command is entered manually, or it will begin  
automatically if the autoverify option is set. Rebuilds, migrations, and  
initializations will take priority over verifies.  
43  
 
Chapter 2. CLI Syntax Reference  
/cx show selftest  
9000 series controllers support background tasks and allow you to schedule a  
regular time when they occur.  
Selftest is one of the supported background tasks. Rebuild and verify are other  
background tasks for which there are separate schedules. Migrate and  
initialize are additional background tasks that follow the same schedule as  
rebuild. For each background task, up to 7 time periods can be registered,  
known as slots 1 through 7. Each task schedule can be managed by a set of  
commands including add, del, show and set a task. Background task  
schedules have a slot id, start-day-time, duration, and status attributes.  
For details about setting up a schedule for background selftest tasks, see  
Selftest activity provides two types of selftests; UDMA (Ultra Direct  
Memory Access) and SMART (Self Monitoring Analysis and Reporting).  
Both self tests are checked once each day by default.  
UDMA self test entails checking the current ATA bus speed (between  
controller and attached disk), which could have been throttled down during  
previous operations and increase the speed for best performance (usually one  
level higher). Possible speeds include 33, 66, 100 and 133 Mhz (at this  
writing). Note that UDMA selftest is not applicable (or required) with SATA  
drives, but is left enabled by default.  
SMART activity instructs the controller to check certain SMART supported  
thresholds by the disk vendor. An AEN is logged to the alarms page if a drive  
reports a SMART failure.  
The show selftest command displays the current selftest background task  
schedule as illustrated below. Selftests do not have a time duration since they  
are completed momentarily.  
//localhost> /c1 show selftest  
Selftest Schedule for Controller /c1  
========================================================  
Slot  
Day  
Hour  
UDMA  
SMART  
--------------------------------------------------------  
1
2
3
4
5
6
7
Sun  
Mon  
Tue  
Wed  
Thu  
Fri  
Sat  
12:00am  
12:00am  
12:00am  
12:00am  
12:00am  
12:00am  
12:00am  
enabled  
enabled  
enabled  
enabled  
enabled  
enabled  
enabled  
enabled  
enabled  
enabled  
enabled  
enabled  
enabled  
enabled  
44  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
 
Controller Object Commands  
/cx add rebuild=ddd:hh:duration  
This command adds a new background rebuild task to be executed on the day  
ddd (where ddd is Sun, Mon, Tue, Wed, Thu, Fri, and Sat), at the hour hh  
(range 0 .. 23), for a duration of duration (range 1 .. 24) hours. A maximum of  
seven rebuild tasks can be scheduled. This command will fail if no (empty)  
task slot is available.  
Example:  
//localhost> /c1 add rebuild=Sun:16:3  
adds a rebuild background task schedule to be executed on Sundays at 16  
hours (4:00 PM) for a duration of 3 hours.  
Setting Up a Rebuild Schedule  
Setting up a rebuild schedule requires several steps, and several different CLI  
commands in addition to /cx add rebuild.  
To set up the rebuild schedule you want to use, follow this  
process:  
1
2
3
4
Use the /cx show rebuild command to display the current schedule for  
rebuild tasks. (For details, see page 41.)  
If any of the scheduled tasks do not match your desired schedule, use the  
/cx del rebuild command to remove them. (For details, see page 47.)  
Use the /cx add rebuild command to create the rebuild schedule slots you  
want (described above.)  
Use the /cx set rebuild=enable command to enable the schedule (this  
enables all rebuild schedule slots). (For details, see page 48.)  
Warning: If all time slots are removed, be sure to also disable the schedule.  
Otherwise the applicable background task will never occur.  
45  
   
Chapter 2. CLI Syntax Reference  
/cx add verify=ddd:hh:duration  
This command adds a new background verify task to be executed on the day  
ddd (where ddd is Sun, Mon, Tue, Wed, Thu, Fri, and Sat), at hour hh  
(range 0 .. 23), for a duration of duration (range 1 .. 24) hours. A maximum of  
seven verify tasks can be scheduled. This command will fail if no (empty)  
task slot is available.  
Example:  
//localhost> /c1 add verify=Sun:16:3  
adds a verify background task schedule to be executed on Sundays at 16  
hours (4:00 PM) for a duration of 3 hours.  
Setting Up a Verify Schedule  
Setting up a verify schedule requires several steps, and several different CLI  
commands in addition to /cx add verify.  
To set up the verify schedule you want to use, follow this  
process:  
1
2
3
4
5
Use the /cx show verify command to display the current schedule for  
verify tasks. (For details, see page 43.)  
If any of the scheduled tasks do not match your desired schedule, use  
the /cx del verify command to remove them. (For details, see page 48.)  
Use the /cx add verify command to create the verify schedule slots you  
want (described above.)  
Use the /cx set verify=enable command to enable the schedule (this  
enables all rebuild schedule slots). (For details, see page 49.)  
Use the /cx/ux set autoverify=on command to turn on autoverify for each  
unit you want to follow the schedule. (For details, see page 61.)  
Note: If you do not enable autoverify for units or start a verification manually, your  
verify schedule will not run, even if it is enabled with the  
/cx set verify=enable command.  
Warning: If all time slots are removed, be sure to also disable the schedule.  
Otherwise the applicable background task will never occur  
46  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
   
Controller Object Commands  
/cx add selftest=ddd:hh  
This command adds a new background selftest task to be executed on the day  
ddd (where ddd is Sun, Mon, Tue, Wed, Thu, Fri, and Sat), at hour hh  
(range 0 .. 23). Notice that selftest runs to completion and as such no duration  
is provided. A maximum of seven selftest tasks can be scheduled. This  
command will fail if no (empty) task slot is available.  
Example:  
//localhost> /c1 add selftest=Sun:16  
adds a selftest background task schedule to be executed on Sundays at 16  
hours (4:00 PM).  
Setting Up a Selftest Schedule  
Setting up a selftest schedule requires several steps, and several different CLI  
commands in addition to /cx add selftest.  
To set up the selftest schedule you want to use, follow this  
process:  
1
2
3
4
Use the /cx show selftest command to display the current schedule for  
selftest tasks. (For details, see page 44.)  
If any of the scheduled tasks do not match your desired schedule, use  
the /cx del selftest command to remove them. (For details, see page 48.)  
Use the /cx add selftest command to create the selftest schedule slots you  
want (described above.)  
Use the /cx set selftest=enable command to enable the schedule (this  
enables all selftest schedule slots). (For details, see page 49.)  
/cx del rebuild=slot_id  
This command removes the rebuild background task in slot slot_id.  
Example:  
//localhost> /c1 del rebuild=2  
removes the rebuild background task in slot 2.  
Warning: If all time slots are removed, be sure to also disable the schedule.  
Otherwise the applicable background task will never occur  
47  
     
Chapter 2. CLI Syntax Reference  
/cx del verify=slot_id  
This command removes the verify background task in slot slot_id.  
Example:  
//localhost> /c1 del verify=3  
removes verify background task in slot 3.  
Warning: If all time slots are removed, be sure to also disable the schedule.  
Otherwise the applicable background task will never occur  
/cx del selftest=slot_id  
This command removes (or unregisters) the selftest background task in slot  
slot_id.  
Example:  
//localhost> /c1 del selftest=3  
Will remove selftest background task in slot 3.  
Warning: If all time slots are removed, be sure to also disable the schedule.  
Otherwise the selftest background task will never occur.  
/cx set rebuild=enable|disable|1..5  
This command enables or disables all rebuild background task slots on  
controller /cx and sets the priority of rebuild versus I/O operations. When  
enabled, rebuild tasks will only be run during the time slots scheduled for  
rebuilds. If a rebuild is taking place when the schedule is enabled, it will be  
paused until the next scheduled time.  
The priority of rebuild versus I/O operations is specified with 1..5, where 1 is  
more resources and 5 the least. Setting the value to 1 gives maximum  
processing time to rebuilds rather than I/O. Setting the value to 5 gives  
maximum processing time to I/O rather than rebuilds.  
Enabling and disabling rebuild schedules is only for 9000 models, however  
the rebuild rate (1..5) applies to all controllers.  
7000- and 8000-series controllers have only one setting for Task Rate; it  
applies to both rebuild and verify rates. This rate is not persistent following a  
reboot for 7000- and 8000-series controllers.  
48  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
     
Controller Object Commands  
/cx set verify=enable|disable|1..5  
This command enables or disables all verify background task slots on  
controller /cx and (when enabled) sets the priority of verification versus I/O  
operations. When enabled, verify tasks will only be run during the time slots  
scheduled for verifies. If a verify is taking place when the schedule is  
enabloed, it will be paused until the next scheduled time.  
The priority of verify versus I/O operations is specified with 1..5, where 1 is  
more resources and 5 the least. Setting this value to 1 implies fastest verify,  
and 5 implies fastest I/O.  
Enabling and disabling verify schedules is only for 9000 models, however the  
verify rate (1..5) applies to all controllers.  
Note: When enabling the verify schedule you must also remember to enable the  
autoverify setting for the units to be verified. For more information see  
/cx set selftest=enable|disable  
[task=UDMA|SMART]  
This command enables or disables all selftest tasks or a particular  
selftest_task (UDMA or SMART).  
Enabling and disabling selftest is only for 9000 models. 7/8000 models have  
the same internal schedule, but it is not viewable or changeable.  
Example:  
//localhost> /c0 selftest=enable task=UDMA  
enables UDMA selftest on controller c0.  
/cx set exportjbod=on|off  
This command allows you to set the JBOD Export Policy to on or off. By  
default, exportjbod is off.  
If the JBOD export policy is off, CLI will not be able to create JBODs. During  
reboot, firmware will not export JBOD units to the operating system.  
The JBOD Export Policy is only supported on 9000-series controllers.  
Previous models did not have such a policy enforcement feature.  
A JBOD is an unconfigured disk attached to your 3ware RAID controller.  
AMCC recommends that you use Single Disk as a replacement for JBOD, to  
take advantage of features such as RAID level migration.  
49  
     
Chapter 2. CLI Syntax Reference  
/cx set ondegrade=cacheoff|follow  
This command is only for 9500S controllers.  
This command allows you to set a controller-based write cache policy. If the  
policy is set to cacheoff and a unit degrades, the firmware will disable the  
write-cache on the degraded unit, regardless of what the unit-based write  
cache policy is. If the policy is set to follow and a unit degrades, firmware will  
follow whatever cache policy has been set for that unit. (For details about the  
/cx set spinup=nn  
This command is only for 9000 series controllers.  
This command allows you to set a controller-based Disk Spinup Policy that  
specifies how many drives can spin up at one time. The value must be a  
positive integer between 1 and the number of disks/ports supported on the  
controller (4, 8, or 12). The default is 1.  
This policy is used to stagger spinups of disks at boot time in order to spread  
the power consumption on the power supply. For example, given a spinup  
policy of 2, the controller will spin up two disks at a time, pause, and then spin  
up another 2 disks. The amount of time to pause can be specified with the  
Spinup Stagger Time Policy (/cx set stagger).  
Not all drives support staggered spinup. If you enable staggered spinup and  
have drives that do not support it, the setting will be ignored.  
/cx set stagger=nn  
This command is only for 9000 series controllers.  
This command allows you to set a controller-based Disk Spinup Stagger Time  
Policy that specifies the delay between spin-ups. The value must be a positive  
integer between 0 to 60 seconds. This policy, in conjunction with Disk Spinup  
Policy, specifies how the controller should spin up disks at boot time. The  
default is 6 seconds.  
/cx set autocarve=on|off  
This command is only for 9000 series controllers.  
This command allows you to set the auto-carve policy to on or off. By default,  
autocarve is off.  
When the auto-carve policy is set to on, any unit larger than the carvesize is  
created or migrated into one or more carvesize volumes and a remaining  
50  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
       
Controller Object Commands  
volume. Each volume can then be treated as an individual disk with its own  
file system. The default carvesize is 2 TB.  
This feature is useful for operating systems limited to 2TB file systems.  
For example, using the 2 TB default carvesize, a 3 TB unit will be configured  
into one 2 TB volume and one 1 TB volume. A 5 TB unit will be configured  
into two 2 TB volumes and one 1 TB volume.  
When auto-carve policy is set to off, all new units are created as a single large  
volume. If the operating system can only recognize up to 2 TBs, space over 2  
TB will not be available.  
Example:  
//localhost> /c0 set autocarve=on  
Setting Auto-Carving Policy on /c0 to on ... Done.  
/cx set carvesize=[1024..2048]  
This command is only for 9000 series controllers.  
This command allows you to set the carve size in GB. This feature works  
Example:  
//localhost> /c0 set carvesize=2000  
Setting Auto-Carving Size on /c0 to 2000 GB ... Done.  
/cx set autorebuild=on|off  
This command is only for 9550SX and 9590SE controllers.  
This command turns the Auto-Rebuild policy on or off. By default,  
autorebuild is on.  
If the policy is on the firmware will select drives to use for rebuilding a  
degraded unit using the following priority order.  
1. Smallest usable spare.  
2. Smallest usable unconfigured (available) drive.  
3. Smallest usable failed drive.  
Note: Failed drives can be drives that have mechanically failed, or they can be  
drives that have been disconnected from the controller long enough for the  
controller to classify them as failed.  
Enabling Auto-Rebuild allows you to add a drive to the controller and have it  
be available for a rebuild as soon as you tell the controller to rescan, without  
51  
   
Chapter 2. CLI Syntax Reference  
having to specify it as a spare. It also means that if you accidentally  
disconnect a drive (causing the controller to see it as a failed drive) and then  
reconnect it, the controller will automatically try to use it again.  
If the policy is off, spares are the only candidates for rebuild operations.  
Example:  
//localhost> /c0 set autorebuild=enable  
Setting Auto-Rebuild Policy on /c0 to enable ... Done.  
/cx start mediascan  
This command applies only to 7000/8000 controllers. For 9000 series  
controllers, use the verify command.  
This command provides media scrubbing for validating the functionality of a  
disk, including bad block detection, remapping, and so forth. The command  
starts a media scan operation on the specified controller /cx.  
/cx stop mediascan  
This command applies only to 7000/8000 controllers.  
This commands stops a media scan operation on the specified controller /cx.  
(Media scans are started using /cx start mediascan.)  
52  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
   
Unit Object Commands  
Unit Object Commands  
Unit Object commands provide information and perform actions related to a  
specific unit, such as /c0/u1 (unit 1 on controller 0). For example, you use  
logical disk object commands for such tasks as seeing the rebuild, verify, or  
initialize status of a unit, starting, stopping, and resuming rebuilds and  
verifies, and setting policies for the unit.  
Syntax  
/cx/ux show  
/cx/ux show attribute [attribute ...] where attributes are:  
initializestatus|name(9000 series)|  
qpolicy(9550SX and 9590SE only)|rebuildstatus|  
serial(9000 series)|status|verifystatus|  
storsave(9550SX and 9590SE only)|volumes(9000 series)|  
identify (9550SX and 9590SE only)  
/cx/ux show all  
/cx/ux start rebuild disk=<p:-p...> [ignoreECC]  
/cx/ux start verify  
/cx/ux pause rebuild  
/cx/ux resume rebuild  
(7000/8000 only)  
(7000/8000 only)  
/cx/ux stop verify  
/cx/ux flush  
/cx/ux del [noscan] [quiet]  
/cx/ux set autoverify=on|off  
/cx/ux set cache=on|off [quiet]  
/cx/ux set identify=on|off  
/cx/ux set ignoreECC=on|off  
/cx/ux set qpolicy=on|off  
/cx/ux set name=string  
(9550SX and 9590SE only)  
(9550SX and 9590SE only)  
(9000 series)  
/cx/ux set storsave=protect|balance|perform [quiet](9550SX  
and 9590SE only)  
/cx/ux migrate type=RaidType [disk=p:-p] [group=3|4|5|6]  
[stripe=Stripe] [noscan] [nocache] [autoverify]  
(9000 only)  
/cx/ux remove [noscan] [quiet]  
/cx/ux show  
This command shows summary information about the specified unit /cx/ux. If  
the unit consists of sub-units as with the case of RAID-1, RAID-5, RAID-10,  
RAID-50, then each sub-unit is further presented. If the Auto-Carving policy  
was on at the time the unit was created and the unit is over the carve size,  
multiple volumes were created and are displayed at the end of the summary  
information.  
53  
   
Chapter 2. CLI Syntax Reference  
One application of the /cx/ux show command is to see which sub-unit of a  
degraded unit has caused the unit to degrade and which disk within that sub-  
unit is the source of degradation. Another application is to see the source and  
destination units during a migration.  
Example:  
//localhost> /c0/u1 show  
Unit  
UnitType Status  
%Cmpl Port Stripe Size(GB) Blocks  
----------------------------------------------------------------------  
u1  
RAID-0  
DISK  
DISK  
DISK  
DISK  
DISK  
DISK  
DISK  
DISK  
DISK  
DISK  
DISK  
DISK  
Volume  
Volume  
OK  
OK  
OK  
OK  
OK  
OK  
OK  
OK  
OK  
OK  
OK  
OK  
OK  
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
64K  
-
-
-
-
-
-
-
-
-
-
-
-
-
-
3576.06  
298.01  
298.01  
298.01  
298.01  
298.01  
298.01  
298.01  
298.01  
298.01  
298.01  
298.01  
298.01  
2047.00  
1529.06  
7499550720  
624962560  
624962560  
624962560  
624962560  
624962560  
624962560  
624962560  
624962560  
624962560  
624962560  
624962560  
624962560  
4292870144  
3206680576  
u1-0  
u1-1  
u1-2  
u1-3  
u1-4  
u1-5  
u1-6  
u1-7  
u1-8  
u1-9  
u1-10  
u1-11  
u1/v0  
u1/v1  
p0  
p1  
p2  
p3  
p4  
p5  
p6  
p7  
p8  
p9  
p10  
p11  
-
-
-
/cx/ux show attribute [attribute ...]  
This command shows the current setting of the specified attributes. One or  
many attributes can be requested. Specifying an invalid attribute will  
terminate the loop. Possible attributes are: initializestatus, name (9000 series),  
qpolicy (9550SX and 9590SE only), rebuildstatus, serial (9000 series), status,  
storsave (9550SX and 9590SE only), verifystatus, volumes (9000 series),  
autoverify (9000 series), cache, ignoreECC (9000 series), and identify  
(9550SX and 9590SE only).  
/cx/ux show status  
This command reports the status of the specified unit.  
Possible statuses include: OK, VERIFYING, VERIFY-PAUSED,  
INITIALIZING, INIT-PAUSED, REBUILDING, REBUILD-PAUSED,  
DEGRADED, MIGRATING, MIGRATE-PAUSED, RECOVERY,  
INOPERABLE, and UNKNOWN. (Definitions of the unit statuses are  
available in the 3ware Serial ATA RAID Controller User Guide.)  
Example:  
//localhost> /c0/u0 show status  
/c0/u5 status = OK  
54  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
   
Unit Object Commands  
/cx/ux show rebuildstatus  
This command reports the rebuildstatus (if any) of the specified unit.  
Example:  
//localhost> /c0/u5 show rebuildstatus  
/c0/u5 is not rebuilding, its current state is OK  
If the unit is in the process of migrating, the command will return the  
following:  
//localhost> /c0/u5 show rebuildstatus  
/c0/u5 is not rebuilding, its current state is MIGRATING  
/cx/ux show verifystatus  
This command reports the verifystatus (if any) of the specified unit.  
Example:  
//localhost> /c0/u5 show verifystatus  
/c0/u5 is not verifying, its current state is OK  
/cx/ux show initializestatus  
This command reports the initializestatus (if any) of the specified unit.  
Example:  
//localhost> /c0/u5 show initializestatus  
/c0/u5 is not initializing, its current state is OK  
/cx/ux show name  
This feature only applies to 9000 series controllers.  
This command reports the name (if any) of the specified unit.  
Example:  
//localhost> /c0/u5 show name  
/c0/u5 name = Joe  
/cx/ux show serial  
This feature only applies to 9000 series controllers.  
This command reports the unique serial number of the specified unit.  
Example:  
//localhost> /c0/u5 show serial  
/c0/u5 Serial Number = 12345678901234567890  
55  
         
Chapter 2. CLI Syntax Reference  
/cx/ux show qpolicy  
This feature only applies to 9550SX and 9590SE model controllers.  
This command reports the queue policy of the firmware. If the queue policy is  
on, the firmware utilizes the drive queueing policy. If some drives in the unit  
do not support a queueing policy, this policy will have no effect on those  
drives.  
Note that currently only NCQ will be enabled, not tag-queueing.  
Example:  
//localhost> /c0/u5 show qpolicy  
/c0/u5 Command Queuing Policy = on  
/cx/ux show storsave  
This feature only applies to 9550SX and 9590SE model controllers.  
This command reports the storsave policy on the unit.  
Example:  
//localhost> /c0/u5 show storsave  
/c0/u5 Command Storsave Policy = protect  
/cx/ux show identify  
This feature only applies to 9550SX and 9590SE model controllers.  
This command is related to the /cx/ux set identify command. It shows the  
identify status of the specified unit (either on or off).  
Example:  
//localhost> /c0/u0 show identify  
/c0/u0 Identify status = on  
/cx/ux show autoverify  
This feature only applies to 9000 series controllers.  
This command shows the current autoverify setting of the specified unit.  
Example:  
//localhost> /c0/u0 show autoverify  
/c0/u0 Auto Verify Policy = off  
56  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
       
Unit Object Commands  
/cx/ux show cache  
This command shows the current write cache state of the specified unit.  
Example:  
//localhost> /c0/u0 show cache  
/c0/u0 Cache State = on  
/cx/ux show ignoreECC  
This feature only applies to 9000 series controllers.  
This command shows the current setting of the ignoreECC policy for the  
specified unit.  
Example:  
//localhost> /c0/u0 show ignoreECC  
/c0/u0 Ignore ECC policy = off  
/cx/ux show volumes  
This feature only applies to 9000 series controllers.  
This command reports the number of volumes in the specified unit. The  
number of volumes will normally be “1” unless the drive capacity exceeds  
2TB and auto-carving is enabled.  
/cx/ux show all  
This command shows the current setting of all above attributes.  
If the auto-carve policy was on at the time the unit was created and the unit is  
over the carve size, multiple volumes were created and are displayed at the  
end of the summary information.  
Example:  
//localhost> /c0/u1 show all  
/c0/u1 status = OK  
/c0/u1 is not rebuilding, its current state is OK  
/c0/u1 is not verifying, its current state is OK  
/c0/u1 is not initializing, its current state is OK  
/c0/u1 volume(s) = 2  
/c0/u1 name = 1234567  
/c0/u1 serial number = C6CPR7JMF98DA8001DF0  
Unit  
UnitType Status  
%RCmpl %VlmCmpl Port Stripe Size(GB)  
-----------------------------------------------------------------  
u1  
RAID-0  
DISK  
OK  
OK  
OK  
-
-
-
-
-
-
-
64K  
-
3578.40  
298.20  
298.20  
u1-0  
u1-1  
p0  
p1  
DISK  
-
57  
       
Chapter 2. CLI Syntax Reference  
u1-2  
u1-3  
u1-4  
u1-5  
u1-6  
u1-7  
u1-8  
u1-9  
u1-10  
u1-11  
u1/v0  
u1/v1  
DISK  
DISK  
DISK  
DISK  
DISK  
DISK  
DISK  
DISK  
DISK  
DISK  
VOLUME  
VOLUME  
OK  
OK  
OK  
OK  
OK  
OK  
OK  
OK  
OK  
OK  
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
p2  
p3  
p4  
p5  
p6  
p7  
p8  
p9  
p10  
p11  
-
-
-
-
-
-
-
-
-
-
-
-
-
298.20  
298.20  
298.20  
298.20  
298.20  
298.20  
298.20  
298.20  
298.20  
298.20  
500.00  
3078.40  
-
-
/cx/ux remove [noscan] [quiet]  
This command allows you to remove (or export) a unit. Exporting a unit  
instructs the firmware to remove the specified unit from its poll of managed  
units, but retains the DCB (Disk Configuration Block) metadata. As such the  
unit can later be re-imported.  
noscan is used to not inform the operating system of this change. The default  
is to inform the operating system.  
quiet is used for non-interactive mode. No confirmation is given and the  
command is executed immediately. This is useful for scripting purposes.  
Example of interactive mode:  
//localhost> /c0/u0 export  
Are you sure you want to export or remove the unit /c0/u0? Enter  
Y to continue.  
Note: After the unit is removed through the CLI, the unit can be physically  
removed. Hot swap carriers are required to do this while the system is online.  
Otherwise you must power down the system to prevent system hangs and damage.  
/cx/ux del [noscan] [quiet]  
This command allows you to delete a unit. Deleting a unit not only removes  
the specified unit from the controller's list of managed units, but also destroys  
the DCB (Disk Configuration Block) metadata. After deleting a unit, ports (or  
disks) associated with the unit will be part of the free pool of managed disks.  
Warning: This is a destructive command and should be used with care. All data on  
the specified unit will be lost after executing this command.  
58  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
   
Unit Object Commands  
noscan is used to not inform the operating system of this change. The default  
is to inform the operating system.  
quiet is used for non-interactive mode. No confirmation is given and the  
command is executed immediately. This is useful for scripting purposes.  
Example of interactive mode:  
//localhost> /c0/u0 del  
Are you sure you want to delete the unit /c0/u0? Enter Y to  
continue.  
/cx/ux start rebuild disk=p<p:-p...> [ignoreECC]  
This command allows you to rebuild a degraded unit using the specified  
disk=p. Rebuild only applies to redundant arrays such as RAID 1, RAID 5,  
RAID 10, and RAID 50.  
During rebuild, bad sectors on the source disk will cause the rebuild to fail.  
You can allow the operation to continue by using ignoreECC.  
The rebuild process is a background task and will change the state of a unit to  
REBUILDING. Various show commands also show the percent completion as  
rebuilding progresses.  
Note that the disk used to rebuild a unit (specified with disk=p) must be a  
SPARE or a unconfigured disk. You must first remove the degraded drive(s)  
before starting the rebuild. Refer to the command “/cx/px remove [noscan]  
[noscan]” on page 38 to add new drives or to retry the original drive.  
If you are rebuilding a RAID 50 or RAID 10 unit, multiple drives can be  
specified if more than one sub-array is degraded.  
When you issue this command, the specified rebuild will begin if schedules  
are disabled; otherwise it will pause until the next scheduled rebuild. A file  
system check is recommended following rebuild when using the ignoreECC  
option.  
/cx/ux start verify  
This command starts a background verification process on the specified unit  
/cx/ux. The following table shows the supported matrix as a function of the  
controller model and logical unit type.  
N/A (Not Applicable) refers to cases where the given logical unit type is not  
supported on that controller model.  
59  
   
Chapter 2. CLI Syntax Reference  
Table 7: Supported RAID (Logical Unit) Types for Verification  
Model  
R0  
R1  
R5  
R10  
Single  
Spare  
7K/8K  
9K  
No  
Yes  
Yes  
Yes  
Yes  
Yes  
Yes  
N/A  
Yes  
No  
Yes  
Yes  
When you issue this command, the specified verify will begin if schedules are  
disabled; otherwise it will pause until the next scheduled verify. Verify will  
also pause if a rebuild or initialization is currently in progress.  
/cx/ux pause rebuild  
This command allows you to pause the rebuild operation on the specified unit  
/cx/ux.  
This feature is only supported on the 7000/8000 series controllers. 9000 series  
controllers have an on-board scheduler where rebuild operations can be  
scheduled to take place at specified start and stop times. The /cx/ux pause  
rebuild command is provided to enable 7000/8000 users to achieve similar  
functionality with use of Linux-provided schedulers such as cron(8) or at(1),  
or user-supplied programs.  
/cx/ux resume rebuild  
This command allows you to resume the rebuild operation on the specified  
unit /cx/ux.  
This feature is intended only for 7000/8000 series controllers. 9000 series  
controllers have an on-board scheduler where rebuild operations can be  
scheduled to take place at specified start and stop times. The /cx/ux resume  
rebuild function is provided to enable 7000/8000 users to achieve similar  
functionality with use of Linux-provided schedulers such as cron(8) or at(1),  
or user supplied programs.  
/cx/ux stop verify  
This command stops a background verification process on the specified unit  
/cx/ux. Table 7 on page 60 shows the supported matrix as a function of the  
controller model and logical unit type.  
60  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
       
Unit Object Commands  
/cx/ux flush  
This command allows you to flush the write cache on the specified unit /ux  
associated with controller /cx. Note that this command does not apply to spare  
unit types.  
/cx/ux set autoverify=on|off  
This feature only applies to 9000 series controllers.  
This command allows you to turn on and off the autoverify operation on a  
specified unit /cx/ux during allocated schedule windows.  
You can use the show verify command to display the existing schedule  
windows. By default, autoverify is off.  
Auto-verify allows the controller to run the verify function once every 24  
hours. If verify schedule windows are set up and enabled, then the controller  
will ony start an automatic verify task during the schedule time slots. If the  
verify takes longer than the schedule window, the verify process will be  
paused and restarted during the next verify schedule window. For additional  
/cx/ux set cache=on|off [quiet]  
This command allows you to turn on or off the write cache for a specified unit  
/cx/ux. This feature is supported on both 7000/8000 and 9000 models.  
By default, cache is on.  
Write cache includes the disk drive cache and controller cache. Note that for  
some configuration types, there is only disk drive cache and no controller  
cache (for example, JBOD).  
The following table shows the supported RAID types for caching as a  
function of controller model and logical unit type. N/A (Not Applicable)  
refers to cases where the given logical unit type is not supported on a  
particular controller model.  
Table 8: Supported RAID Types for Caching  
Model  
R0  
R1  
R5  
R10  
Single Spare  
7K/8K  
9K  
Yes  
Yes  
Yes  
Yes  
Yes  
Yes  
Yes  
Yes  
N/A  
Yes  
No  
No  
The quiet attribute turns off interactive mode.  
61  
     
Chapter 2. CLI Syntax Reference  
/cx/ux set identify=on|off  
This feature only applies to 9550SX and 9590SE model controllers.  
This command allows you to identify a unit within an enclosure.  
When that the EPCT (Enclosure Port Configuration Table) equates identify  
with blinking, as it does in the 3ware Sidecar Enclosure, the LEDs associated  
with the drive slots of the specified unit blink.  
Example:  
//localhost> /c0/u0 set identify=on  
Sending Identify request for unit /c0/u0 to [on] ... Done.  
/cx/ux set ignoreECC=on|off  
This feature only applies to 9000 series controllers.  
This command allows you to set the ignoreECC policy for a given unit.  
When ignoreECC policy is set to off, if a rebuild process encounters bad  
sectors on the source disk, the rebuild will fail. When ignoreECC is set to on,  
such errors are ignored, and the rebuild will continue. When you use  
ignoreECC, a file system check is recommended following the rebuild, to  
insure data integrity.  
By default, ignoreECC is off.  
/cx/ux set name=string  
This command allows you to name the unit with an arbitrary name. You can  
use this name in conjunction with the unit serial number to cross-reference  
with the unit. The system does not check to ensure uniqueness of names, so be  
careful to assign different names to each unit.  
Note: The unit’s serial number is automatically assigned when the unit is created  
and is not changeable.  
/cx/ux set qpolicy=on|off  
This command applies only to 9550SX and 9590SE controllers.  
This command sets the queue policy of the firmware. If the queue policy is  
on, the firmware utilizes the drive queueing policy. If some of the drives in  
the unit do not support any queueing policy, this policy will have no effect on  
those drives.  
62  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
       
Unit Object Commands  
By default, qpolicy is off.  
Example:  
//localhost> /c0/u5 set qpolicy = on  
Setting Command Queuing Policy for unit /c0/u5 to [on] ... Done.  
/cx/ux set storsave=protect|balance|perform [quiet]  
This command applies only to 9550SX and 9590SE model controllers.  
This command sets the storsave policy to be either protect, balance, or  
perform when the unit write cache is enabled. The default setting is protect.  
The storsave policy adjusts several factors that control the balance between  
protection and performance on a unit. There is a trade-off among the available  
settings. The following description about the settings should help you to  
decide which one is suitable to you and your application. You will find further  
discussion of this setting in the 3ware Serial ATA RAID Controller User  
Guide, under “About StorSave Profile Levels” on page 48.  
protect provides the maximum data protection among the controller settings.  
When storsave is set to protect mode, it means:  
When the unit becomes degraded, the write cache will be disabled.  
Write journaling is enabled. All data flushing from controller cache will  
be flushed to media.  
Incoming FUA (Force Unit Access) host requests will be honored unless  
a BBU is installed and enabled, in which case, they will be ignored.  
perform provides the maximum performance and least data protection of the  
three controller settings. When storsave is set to perform mode, it means:  
When the unit becomes degraded, the write cache will not be disabled.  
Write journaling is disabled. All data flushing from controller cache will  
be flushed to disk. If a BBU is present, this essentially disables the BBU  
for this unit.  
Incoming FUA (Force Unit Access) host requests will be honored.  
If you set the storsave policy to perform, a confirmation message will warn  
you that there could be data loss in the event of a power failure.  
quiet is used for non-interactive mode. No confirmation is given and the  
command is executed immediately. This is useful for scripting purposes.  
balance provides more data protection than perform mode but less data  
protection than protect mode, and provides better performance than protect  
mode but less performance than perform mode. When storsave is set to the  
balance mode, it means:  
When the unit becomes degraded, the write cache will not be disabled.  
63  
 
Chapter 2. CLI Syntax Reference  
Write journaling is disabled, if no BBU is present, and is enabled, if a  
BBU is present. All data flushing from controller cache will be flushed to  
media if a BBU is installed and enabled. Otherwise, data will be flushed  
to disk only.  
Incoming FUA (Force Unit Access) host requests will be honored unless  
a BBU is installed and enabled, in which case, they will be ignored.  
For additional information, see “Setting the StorSave Profile for a Unit” in the  
3ware Serial ATA RAID Controller User Guide.  
Example:  
//localhost> /c0/u5 set storsave=protect  
Setting Command Storsave Policy for unit /c0/u5 to [protect] ...  
Done.  
/cx/ux migrate type=RaidType [disk=p:-p]  
[group=3|4|5|6] [stripe=Stripe] [noscan] [nocache]  
[autoverify]  
This feature only applies to 9000 series controllers.  
This command allows you to change the existing configuration of a unit with  
type=RaidType. You can make three types of changes:  
Increase the capacity  
Change the RAID level (with the same or increased capacity)  
Change the stripe size  
The unit that results from the migration is subject to the same rules and  
policies that apply when creating a new unit with the /cx add command. For  
example, a valid number of disks and parameters must be specified.  
The destination unit must use all source disks and potentially augment the  
number of disks in the disk=p:-p disk list. Unspecified parameters are  
assigned the default values (stripe size of 64K, write cache enabled,  
autoverify disabled, and ignoreECC disabled). Both source name and serial  
number will be carried over to the destination unit.  
A special case of this command is when the source unit has a type of RAID1  
and destination unit has a type of single. In this case, the migrate command  
splits both drives into two identical single disks. The disk name will be  
duplicated on the destination units, but the source unit serial number will not  
be carried over to the new unit. The new destination unit will have its own  
serial number.  
64  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
 
Unit Object Commands  
Warning: Do not split a RAID 1 unit if you are booted from the drives or if the unit  
is mounted. Once the unit is split, any pending writes cannot be written to the  
second drive. In addition, the file system on the drive will not be clean.  
type=RaidType specifies the RAID type of the destination unit. Possible unit  
types include raid0, raid1, raid5, raid10, raid50, or single.  
For example, type=raid5 indicates the destination unit is RAID-5. The  
type=single is a special case of the migrate command. It splits the source unit  
RAID-1 or TWINSTOR into multiple Single units.  
Note: You can only migrate a unit to a RAID level that has the same or more  
capacity as the existing one. A four-drive RAID 5 unit can migrate to a four-drive  
RAID 0, but a four-drive RAID 0 unit cannot migrate to a four-drive RAID 5, without  
adding another drive, due to the need for additional storage capacity for parity bits.  
The following table illustrates valid migration paths:  
Table 9: Valid Migration Paths  
Destination  
Source  
R0  
R0  
R1  
R5  
R10  
Single  
Spare  
Yes  
No  
Yes  
Yes  
No  
No  
Yes  
Yes  
Yes  
Yes  
No  
No  
No  
No  
Yes  
No  
Yes  
Yes  
Yes  
Yes  
No  
Yes  
Yes  
Yes  
No  
No  
No  
No  
No  
No  
No  
No  
No  
R1  
R5  
a
R10  
Yes  
Yes  
Single  
Spare  
No  
a. When migrating a RAID 10 to a RAID 10, the only change you can  
make is the stripe size.  
disk=p:-p.. consists of a list of ports (disks) to be used in addition to the  
source disks in the construction of the destination unit. One or more ports can  
be specified. Multiple ports can be specified using a colon (:) or a dash (-) as  
port index separators. A dash indicates a range and can be mixed with colons.  
For example disk=0:1:2-5:9:12 indicates port 0, 1, 2 through 5 (inclusive), 9  
and 12.  
65  
Chapter 2. CLI Syntax Reference  
group=3|4|5|6 indicates the number of disks per group for a RAID 50 type.  
(This attribute can only be used when type=raid50.) Recall that a RAID 50 is  
a multi-tier array. At the bottom-most layer, N number of disks per group are  
used to form the RAID 5 layer. These RAID 5 arrays are then integrated into a  
RAID 0. This attribute allows you to specify the number of disks in the RAID  
5 level. Valid values are 3, 4, 5 and 6. For example group=3 indicates 3 disks  
of RAID 5 at the bottom layer of RAID 50.  
Note that a sufficient number of disks are required for a given pattern or disk  
group. For example, given 6 disks, specifying 3 will create two RAID 5  
arrays. With 12 disks, specifying 3 will create four RAID 5 arrays under the  
RAID 0 level. With only 6 disks a grouping of 6 is not allowed, as you would  
basically be creating a RAID 5.  
The default RAID 50 grouping varies, based on number of disks. For 6 and 9  
disks, default grouping is 3. For 8 disks, the default grouping is 4. For 10  
disks, the default grouping is 5, and for 12 disks, the default grouping is 4. In  
the case of 12, the disks could be grouped into groups of 3, 4, or 6 drives. A  
grouping of 4 is set by default as it provides the best of net capacity and  
performance.  
Note that RAID-10 always has group=2, so an attribute specifying it’s group  
is not necessary.  
stripe=Stripe consists of the stripe size to be used. The following table  
illustrates the supported and applicable stripes on unit types and controller  
models. Stripe size units are in KB (kilobytes).  
Table 10: Supported Stripe Sizes  
Model  
R0  
R1  
R5  
16  
R10  
16  
Spare  
Single  
9K  
16  
N/A  
N/A  
N/A  
64  
64  
64  
256  
256  
256  
noscan attribute instructs CLI not to notify the operating system of the  
creation of the new unit. By default CLI will inform the operating system.  
One application of this feature is to prevent the operating system from  
creating block special devices such as /dev/sdb and /dev/sdc as some  
implementations might create naming fragmentation and a moving target.  
nocache attribute instructs CLI to disable the write cache on the migrated  
unit. Enabling write cache increases write performance at the cost of potential  
data loss in case of sudden power loss (unless a BBU or UPS is installed). By  
default the cache is enabled. To avoid the possibility of data loss in the event  
of a sudden power loss, it is recommended not to set nocache unless there is a  
BBU (battery backup unit) or UPS (uninterruptable power supply) installed..  
66  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
Unit Object Commands  
autoverify attribute enables the autoverify attribute on the unit that is to be  
migrated. For more details on this feature, see “/cx/ux set autoverify=on|off”  
Migration Process  
In all cases of migration, the background migration process must be  
completed before the newly sized unit is available for use. You can continue  
using the original unit during this time. Once the migration is finished, a  
reboot will be required if you are booted from the unit. For secondary storage,  
depending on your operating system, you may need to first unmount the unit,  
then use CLI commands to ‘remove’ and ‘rescan’ the unit so that the  
operating system can see the new capacity, and then remount the unit. For  
You may also need to resize the file system or add a new partition. For  
instructions, consult the documentation for your operating system.  
Warning: It is important that you allow migration to complete before adding drives  
to the unit. Making physical changes to the unit during migration may cause the  
migration process to stop, and can jeopardize the safety of your data.  
Example of splitting a mirror  
//localhost> /c1/u3 migrate type=single  
Indicates that u3 is a TWINSTOR or RAID-1 and the Migrate command splits  
u3 to u3 and ux with a RAID type of Single.  
Example of capacity expansion  
//localhost> /c0/u0 migrate type=raid10 disk=3-4 stripe=16  
Indicates that the destination unit has a RAID type of raid10 and has the disk  
3 and 4 in addition to all the disks in the existing unit u0.  
Example of migrate output  
The following is an example of how migrating units will be displayed. In this  
example, the report indicates that /c0/u0 is a migrating unit with 67%  
completion. The report also indicate that Source Unit su0 is of type RAID-1  
and Destination Unit du0is of type RAID-10.  
67  
Chapter 2. CLI Syntax Reference  
//localhost> /c0/u0 migrate type=raid10 disk=2-3 stripe=16  
Sending migration message to /c0/u0 ... Done.  
3ware CLI> /c0 show  
Unit UnitType Status  
%Cmpl Stripe Size(GB) Cache AVerify IgnECC  
---------------------------------------------------------------------------  
u0  
Migrator MIGRATING  
Status Unit  
67%  
Size  
-
74.4951  
Blocks  
ON  
OFF  
OFF  
Port  
Serial  
---------------------------------------------------------------  
p0  
p1  
p2  
p3  
p4  
p5  
p6  
p7  
p8  
p9  
p10  
p11  
OK  
OK  
OK  
OK  
u0  
u0  
u0  
u0  
-
-
-
-
-
74.53 GB  
232.83 GB  
232.83 GB  
232.83 GB  
-
-
-
-
-
-
-
-
156301488  
488281250  
488281250  
488281250  
-
-
-
-
-
-
-
-
3JV2Q1VA  
WD-WMAEH15764  
WD-WMAEH17004  
WD-WMAEH17000  
-
-
-
-
-
-
-
-
NOT-PRESENT  
NOT-PRESENT  
NOT-PRESENT  
NOT-PRESENT  
NOT-PRESENT  
NOT-PRESENT  
NOT-PRESENT  
NOT-PRESENT  
-
-
-
3ware CLI> /c0/u0 show  
Unit  
UnitType Status  
%Cmpl Port Stripe Size(GB) Blocks  
-----------------------------------------------------------------------  
u0  
Migrator MIGRATING  
67%  
-
-
-
-
su0  
su0-0  
su0-1  
RAID-1  
DISK  
DISK  
OK  
OK  
OK  
-
-
-
-
p0  
p1  
-
-
-
74.4951  
74.4951  
74.4951  
156227584  
156227584  
156227584  
du0  
du0-0  
du0-0-0 DISK  
du0-0-1 DISK  
du0-1  
RAID-10  
RAID-1  
OK  
OK  
OK  
OK  
OK  
OK  
OK  
-
-
-
-
-
-
-
-
-
16K  
-
-
-
-
148.99  
-
74.4951  
74.4951  
-
312455168  
-
156227584  
156227584  
-
p0  
p1  
-
p2  
p3  
RAID-1  
du0-1-0 DISK  
du0-1-1 DISK  
-
-
74.4951  
74.4951  
156227584  
156227584  
68  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
Port Object Commands  
Port Object Commands  
Port Object Messages are commands that provide information and perform  
actions related to a specific disk, attached to a port, such as /c0/p0. You use  
port object commands for such tasks as seeing the status, model, or serial  
number of the drive.  
Syntax  
/cx/px show  
/cx/px show attribute [attribute ...] where attributes are:  
capacity|firmware|identify (9550SX and 9590SE only)|  
lspeed (9550SX and 9590SE only)|model|  
ncq (9550SX and 9590SE only)|serial|smart|status  
/cx/px show all  
/cx/px remove [noscan][quiet]  
/cx/px set identify=on|off (9550SX and 9590SE only)  
/cx/px show  
This command shows summary information about the specified disk attached  
to port /cx/px. Typical information looks like:  
Example:  
//localhost> /c1/p5 show  
Port Status  
---------------------------------------------------------  
p5 OK u0 149.05 GB 312581808 3JS0L9QW  
Unit Size  
Blocks  
Serial  
The above report indicates that port 5 of controller 1 is attached to one l disk  
with status OK participating in unit 0.  
/cx/px show attribute [attribute ...]  
This command shows the current setting of the given attributes on the  
specified port or drive. One or many attributes can be requested. Specifying  
an invalid attribute will terminate the loop. Possible attributes are: capacity,  
firmware, identify (9550SX and 9590SE only), lspeed (9550SX and 9590SE  
only), model, ncq (9550SX and 9590SE only), serial, smart, and status.  
69  
     
Chapter 2. CLI Syntax Reference  
/cx/px show status  
This command displays the status of the drive attached to the specified port.  
(Definitions of the drive statuses are available in the 3ware Serial ATA RAID  
Controller User Guide.)  
Example:  
//localhost> /c0/p5 show status  
/c0/p5 Status = OK  
/cx/px show model  
This command displays the model of the drive attached to the specified port.  
Example:  
//localhost> /c0/p5 show model  
/c0/p5 Model = WDC WD1600BB-00DAA0  
/cx/px show serial  
This command displays the serial number of the drive attached to the  
specified port.  
Example:  
//localhost> /c0/p5 show serial  
/c0/p5 Serial = WD-WMACK140649  
/cx/px show firmware  
This command displays the firmware version of the drive attached to the  
specified port.  
Example:  
//localhost> /c0/p5 show firmware  
/c0/p5 Firmware Version = 65.13G65  
/cx/px show identify  
This command applies only to 9550SX and 9590SE model controllers that  
have chassis control hardware (enclosure services) attached.  
This command shows whether the LED of the drive attached to the specified  
port is set to on or off. For details, see “/cx/px set identify=on|off” on page 73.  
Example:  
//localhost> /c0/p5 show identify  
/c0/p5 Identify Status = on  
70  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
         
Port Object Commands  
/cx/px show ncq  
This command applies only to 9550SX and 9590SE model controllers.  
This command displays the NCQ (Native Command Queueing) information  
for the drive attached to the specified port, including whether NCQ is  
supported by the drive, and whether it is enabled at the drive.  
For queuing to be used, it must be enabled for the unit and supported by the  
drive.  
Example:  
//localhost> /c0/p5 show ncq  
/c0/p5 NCQ Supported = No  
/c0/p5 NCQ Enabled = No  
/cx/px show lspeed  
This command applies only to 9550SX and 9590SE model controllers.  
This command displays the maximum SATA link speed supported by the  
drive attached to the port and the present SATA link speed setting.  
Example:  
//localhost> /c0/p5 show lspeed  
/c0/p5 SATA Link Speed Supported = 3.0 Gb/s  
/c0/p5 SATA Link Speed = 3.0 Gb/s  
/cx/px show capacity  
This command displays the capacity of the drive attached to the specified port  
in two formats—GB and blocks. Note that of this version, the GB format is  
computed based on division by 1000 (not 1024).  
Example:  
//localhost> /c0/p5 show capacity  
149.05 GB (312581808)  
/cx/px show smart  
This command extracts SMART (Self Monitoring Analysis and Reporting)  
data from the specified disk. Because the data is extracted live from the disk,  
this command can be used to get the most recent data about the presence or  
absence of a disk.  
The SMART data is displayed in hexadecimal form.  
71  
       
Chapter 2. CLI Syntax Reference  
Example:  
//localhost> /c0/p5 show smart  
/c0/p5 Drive SMART Data:  
10 00 01 0B 00 C8 C8 00 00 00 00 00 00 00 03 07  
00 9A 96 BC 14 00 00 00 00 00 04 32 00 64 64 7A  
00 00 00 00 00 00 05 33 00 C8 C8 00 00 00 00 00  
...  
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 2C  
Note: The SMART data is not decoded. If the drive attached to the specified port is  
not present or if there are cabling problems reaching the drive, CLI will return an  
error. This can be one way of detecting whether or not a drive is present.  
/cx/px show all  
This command shows the current setting for all port-related attributes: status,  
model, serial, firmware, capacity, and smart.  
Example:  
//localhost> /c0/p0 show all  
/c0/p0 Status = OK  
/c0/p0 Model = Maxtor 7B300S0  
/c0/p0 Firmware Version = BANC1980  
/c0/p0 Serial = B605X31H  
/c0/p0 Capacity = 279.48 GB (586114704 Blocks)  
/c0/p0 Identify Status - NA  
/c0/p0 SATA Link Speed Supported = 1.5 Gb/s  
/c0/p0 SATA Link Speed = 1.5 Gb/s  
/c0/p0 NCQ Supported = No  
/c0/p0 NCQ Enabled = No  
/c0/p0 Belongs to Unit = u1  
/c0/p0 Drive Smart Data:  
0A 00 01 0F 00 3D 33 25 8C BA 03 00 00 00 03 03  
00 61 60 00 00 00 00 00 00 00 04 32 00 64 64 00  
00 00 00 00 00 00 05 33 00 64 64 00 00 00 00 00  
00 00 07 0F 00 4E 3E 05 13 D8 03 00 00 00 09 32  
...  
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 8D  
72  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
 
Port Object Commands  
/cx/px remove [noscan] [quiet]  
This command allows you to remove (or export) a port (or drive) /cx/px.  
Exporting a port instructs the firmware to remove the specified port from its  
pool of managed ports, but does not retain the DCB (Disk Configuration  
Block) metadata on the attached disk. You can import (or re-introduce) the  
port by rescanning the controller.  
noscan is used to not inform the operating system of this change. The default  
is to inform the operating system.  
quiet is for non-interactive mode.  
Warning: Use caution when using this command as this operation will degrade  
any redundant units. This command will fail if you attempt to remove a drive from a  
non-redundant unit. After the drive is removed in CLI it can be removed physically,  
without powering down the system if a hot swap carrier is available. System hangs  
and damage can occur if a hot swap carrier is not used.  
/cx/px set identify=on|off  
This command applies only to 9550SX and 9590SE model controllers.  
This command sets the LED status of the port to on or off. If identify is set to  
on, the firmware activates the setting of the corresponding LED of the port on  
the controller and causes it to blink.  
Note: This command is equivalent to “/ex/slotx set identify=on|off” on  
Note: This feature requires a chassis with support for Chassis Control Unit (CCU),  
such as the 3ware Sidecar, or a chassis from a third-party vendor. The CCU may  
also support the monitoring of fan speed, temperature sensors, door lock, and so  
forth, depending on the level of chassis vendor support. Check the 3ware support  
web site for future updates on a list of chassis vendors that support the CCU, as  
they become available. For additional information about enclosure-related  
Example:  
//localhost> /c0/p5 set identify=on  
Setting Port Identify on /c0/p5 to [on] ... Done.  
73  
   
Chapter 2. CLI Syntax Reference  
BBU Object Commands  
BBU (Battery Backup Unit) Object Commands are commands that provide  
information and perform actions related to a specific BBU installed on a  
specific controller, such as /c0/bbu.  
This object is only available on 9000 series controllers on which a BBU is  
actually installed. (The BBU is not supported on 9590SE-4ME.)  
Syntax  
/cx/bbu show  
(9000 only)  
/cx/bbu show attribute [attribute ...] where attributes are:  
batinst|bootloader|cap|fw|lasttest|pcb|ready|serial|  
status|temp|volt  
/cx/bbu show all  
(9000 only)  
/cx/bbu test [quiet]  
(9000 only)  
Warning: May take up to 24 hours to complete. Write cache  
will be disabled during the test.  
/cx/bbu enable  
/cx/bbu disable [quiet]  
(9000 only)  
(9000 only)  
/cx/bbu show  
This command presents a summary report on the specified BBU object.  
Example:  
//localhost> /c0/bbu show  
Name OnlineState BBUReady Status  
---------------------------------------------------------------------------  
bbu ON No Testing OK OK 72 01-Jul-2004  
Volt  
Temp  
Hours LastCapTest  
The command output indicates that the battery capacity was last measured on  
01-Jul-2004. The battery is estimated to last for 72 hours from the last tested  
date. In this example, the BBU unit is currently testing the battery. Both  
voltage and temperature are normal. The BBU is not ready to backup the write  
cache on the controller (due to the testing). (For complete information about  
the BBU, see the user guide that came with your 3ware RAID controller).  
Note: If the BBU is either not present or disabled, the following will be displayed  
after the command //localhost> /c0/bbu show.  
Error: (CLI:053) Battery Backup Unit is not present.  
74  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
   
BBU Object Commands  
/cx/bbu show attribute [attribute ...]  
This command shows the current setting of the given attribute(s) on the BBU  
board. One or many attributes can be specified. Specifying an invalid attribute  
will terminate the loop. Possible attributes are: batinst, bootloader, cap, fw,  
lasttest, pcb, ready, status, serial, temp, volt.  
/cx/bbu show status  
This command shows the status of the BBU. Possible values are:  
Testing. A battery test is currently in progress. This test may take up to 24  
hours to complete. During the test, the BBU is not capable of backup  
operation and the write cache of the RAID controller is also disabled. If the  
test is completed with no error and the BBU status changes to WeakBat or  
OK, the write cache will be re-enabled. If a Fault, Failed or Error occurs  
during the test, the write cache remains in the disabled state until the problem  
is fixed.  
Charging. The BBU is currently charging the battery. Charging is started  
automatically by the BBU whenever necessary. During charging, the BBU is  
not capable of backup operation and the write cache is disabled. Once the test  
is completed with no error and the BBU status changes to OK, the write cache  
will be re-enabled. If a FAULT or ERROR occurs during the test, the write  
cache remains in the disabled state until the problem is fixed.  
Fault. A battery fault is detected. The BBU is not capable of backup operation  
and the write cache is disabled. Replace the battery and/or the BBU board as  
soon as possible so that the write cache will be enabled again.  
Error. A BBU error is detected. The BBU is not capable of backup operation  
and the write cache is disabled. Replace the battery and/or the Battery Backup  
Unit as soon as possible so that the write cache will be enabled again.  
Failed. The battery failed a test. In this state, the BBU is not capable of  
backup operation and the write cache is disabled. We recommend you replace  
the battery and/or the Battery Backup Unit as soon as possible so that the  
write cache will be enabled again.  
WeakBat. The BBU is functioning normally and is online and capable of  
backing up the write cache. However, the battery is weak and should be  
replaced.  
OK. The BBU is ready, online and capable of backing up the write cache.  
- (dash) A battery is not present or a Battery Backup Unit is not installed  
75  
   
Chapter 2. CLI Syntax Reference  
/cx/bbu show batinst  
This command shows the date when the current battery was installed.  
/cx/bbu show lasttest  
This command shows the date the battery capacity was last measured. If the  
battery capacity test has never been run, then 'xx-xxx-xxxx' will be displayed.  
Note: The estimated BBU capacity hours displayed is based on the measurement  
taken during the last test. If you have not run the BBU test command for some time,  
this number can be misleading. For information about running a test, see “/cx/bbu  
/cx/bbu show volt  
This command shows the voltage status of the battery. The status can be OK,  
HIGH, LOW, TOO-HIGH, and TOO-LOW. The HIGH and LOW are in  
warning range. TOO-HIGH and TOO-LOW are out of the operating range  
and indicate that it is time to replace the battery. (Contact AMCC to obtain a  
replacement battery.)  
/cx/bbu show temp  
This command shows the temperature status of the battery. The status can be  
OK, HIGH, LOW, TOO-HIGH, and TOO-LOW. The HIGH and LOW are in  
warning range. TOO-HIGH and TOO-LOW are out of the operating range  
and indicate that it may be time to replace the battery. (Contact AMCC to  
obtain a replacement battery.)  
/cx/bbu show cap  
This command shows the battery capacity in hours.  
A value of '0 hours' will be displayed if the battery capacity test has never  
been run.  
Note: The estimated BBU capacity hours displayed is based on the measurement  
taken during the last test. If you have not run the BBU test command for some time,  
this number can be misleading. You can use the command /cx/bbu show lasttest  
to check the date of the last test. For information about running a test, see “/cx/bbu  
76  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
         
BBU Object Commands  
/cx/bbu show serial  
This command shows the BBU serial number.  
/cx/bbu show fw  
This command shows the BBU firmware version number.  
/cx/bbu show pcb  
This command shows the PCB revision number on the BBU.  
/cx/bbu show bootloader  
This command shows the BBU's boot loader version.  
/cx/bbu show all  
This command shows the current settings of all BBU-related attributes: ready,  
status, batinst, lasttest, volt, temp, cap, serial, fw, pcb, bootloader.  
Example:  
//localhost> /c1/bbu show all  
/c1/bbu Firmware Version  
/c1/bbu Serial Number  
/c1/bbu BBU Ready  
/c1/bbu BBU Status  
/c1/bbu Battery Voltage  
/c1/bbu Battery Temperature  
= BBU: 1.04.00.007  
= Engineering Sample.  
= Yes  
= OK  
= OK  
= OK  
/c1/bbu Estimated Backup Capacity = 241 Hours  
/c1/bbu Last Capacity Test = 22-Jun-2004  
/c1/bbu Battery Installation Date = 20-Jun-2004  
/c1/bbu Bootloader Version  
/c1/bbu PCB Revision  
//localhost>  
= BBU 0.02.00.002  
= 65  
/cx/bbu test [quiet]  
This command starts the battery capacity test. The test may take up to 24  
hours to complete. During the test, the BBU is not capable of backup  
operation and the write cache of all units attached to that controller is  
disabled. Once the test is completed with no error and the BBU status returns  
to OK, the write cache will be re-enabled.  
77  
           
Chapter 2. CLI Syntax Reference  
Note: Once started, the test can not be terminated before it completes. Write  
cache cannot be enabled until the test completes.  
AEN (Asynchronous Event Notification) messages are also generated by  
controllers to notify the user of the command status.  
Check for AENs with the alarms command /cx show alarms [reverse]. Using  
the “reverse” attribute displays the most recent AEN message at the bottom of  
the list. (For a list of all AENs, see the user guide that came with your 3ware  
RAID controller.)  
/cx/bbu enable  
This command enables BBU detection on the controller. If the BBU is Ready,  
the controller will utilize BBU functionality in the event of a power failure.  
/cx/bbu disable [quiet]  
This command disables BBU detection on the controller. When disabled, the  
controller ignores the existence of the BBU and will show no BBU is installed  
even if a BBU is physically attached.  
78  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
   
Enclosure Object Commands  
Enclosure Object Commands  
Enclosure object commands provide information and perform actions related  
to a specific enclosure, such as /e0 and its elements, such as /e0/slot0.  
Enclosure object elements include slot, fan, and temperature sensor elements.  
These commands support the 9550SX and 9590SE controllers. If you  
purchased an enclosure directly from a third-party vendor, the appropriate  
EPCT (Enclosure Port Configuration Table) must have been downloaded to  
the controller in order to take advantage of these commands.  
If you have a 3ware Sidecar Kit for use with a Power Mac G5, the required  
EPCT file is already installed.  
Syntax  
/ex show  
(9550SX and 9590SE only)  
/ex show attribute [attribute ...] where attributes are:  
controllers|slots|fans|temp  
/ex show all  
(9550SX and 9590SE only)  
/ex/slotx show  
/ex/slotx show identify  
/ex/slotx set identify=on|off  
(9550SX and 9590SE only)  
(9550SX and 9590SE only)  
(9550SX and 9590SE only)  
/ex/fanx show  
/ex/tempx show  
(9550SX and 9590SE only)  
(9550SX and 9590SE only)  
/ex show  
This command shows summary information on the specified enclosure /ex.  
This report consists of four parts; the Enclosure summary listing the present  
elements, a Fan summary section listing of all present fans, a Temperature  
Sensor summary section listing of all present temperature sensors and a Slot  
summary section listing of slots and associated information for the specified  
enclosure.  
Typical output looks like:  
//localhost> /e0 show  
Encl  
Controllers  
-------------------------  
e0  
/c0  
Fan  
Status  
----------------------  
fan0 OK  
79  
   
Chapter 2. CLI Syntax Reference  
TempSensor  
Temperature  
-------------------------------  
temp0  
Slot  
24~C(75~F)  
Port  
Status  
Identify  
----------------------------------------------------  
slot0  
slot1  
slot2  
slot3  
OK  
OK  
/c0/p0  
/c0/p1  
-
-
No  
Yes  
No  
NO-DEVICE  
NO-DEVICE  
No  
/ex show attribute [attribute ...]  
This command shows the current setting of the given attribute(s). One or  
many attributes can be requested. An invalid attribute will terminate the loop.  
Possible attributes are: controllers, protocol, slots, fans, and temp.  
/ex show controllers  
This command lists the controller associated with enclosure /ex.  
Example:  
//localhost> /e0 show controllers  
/e0 Connected to /c0 controller.  
/ex show slots  
This command reports the slots in enclosure /ex and their associated  
information.  
Example:  
//localhost> /e0 show slots  
Slot  
Status  
Port  
Identify  
----------------------------------------------------  
slot0  
slot1  
slot2  
slot3  
OK  
OK  
/c0/p0  
/c0/p1  
-
-
No  
Yes  
No  
NO-DEVICE  
NO-DEVICE  
No  
80  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
     
Enclosure Object Commands  
/ex show fans  
This command lists the fans in enclosure /ex and shows their status. Possible  
statuses are OK and Unknown.  
Example:  
//localhost> /e0 show fans  
Fan  
----------------------  
fan0 OK  
Status  
/ex show temp  
This command lists the temperature sensors in enclosure /ex and the current  
temperature.  
The maximum temperature for successful use of a drive should be noted in the  
documentation for the drive.  
Example:  
//localhost> /e0 show temp  
TempSensor  
-------------------------------  
temp0 24~C(75~F)  
Temperature  
/ex show all  
This command shows the current settings of all attributes for enclosure /ex.  
/ex/slotx show  
This command shows information about the specified /slotx on the specified  
enclosure /ex. The slot name is followed by its status. If a slot has been  
inserted with a drive and no fault has been detected, the status is OK. If the  
slot is empty the status would indicate NO-DEVICE. The port that is  
correlated to the slot is indicated in the next column. If no device is found in  
the slot, that is indicated with a dash (-) in the Port column. The final column  
shows whether “identify” is currently set for the specified slot.  
Example:  
//localhost> /e0/slot1 show  
Slot  
----------------------------------------------------  
slot1 OK /c0/p1 Yes  
Status  
Port  
Identify  
81  
       
Chapter 2. CLI Syntax Reference  
/ex/slotx show identify  
This command shows the identify status of the specified slot. The status can  
be either on or off.  
Example:  
//localhost> /e0/slot1 show identify  
/e0/slot1 Identify status = on  
/ex/slotx set identify=on|off  
This command causes the slot to be identified by blinking the LED associated  
with it, or turns off identification of the LED for this slot.  
Setting identify to on will cause the LED associated with that slot to blink,  
provided that the EPCT has been set to associate “identify” with “blinking,”  
as is the case in the 3ware Sidecar.  
Note: This command is equivalent to “/cx/px set identify=on|off” on page 73.  
Example:  
//localhost> /e0/slot1 set identify=on  
Sending Identify request to Drive Slot /e0/slot0 to [on] ...  
Done.  
/ex/fanx show  
This command shows the information about the specified fan element /fanx in  
the specified enclosure /ex. The fan name is followed by its status. If a fan is  
on and no fault has been detected, the status would indicate OK.  
Example:  
//localhost> /e0/fan0 show  
Fan  
Status  
-------------------  
fan0  
OK  
/ex/tempx show  
This command shows the information about the specified temperature sensor  
element /fanx in the specified enclosure /ex. The temperature sensor name is  
followed by the temperature sensed in the enclosure unit.  
Example:  
//localhost> /e0/temp0 show  
TempSensor  
Temperature  
-------------------------------  
temp0  
24~C(75~F)  
82  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
       
Help Commands  
Help Commands  
The Help commands provides brief on-line help.  
You can get overview help by typing Help at the top-level prompt. This  
displays a brief definition of commands in both the new syntax and the legacy  
syntax. (For an example, see the discussion of the command “help” on  
page 85.)  
You can also get help with specific commands, by entering help before an  
object name, or by typing a question mark (?) at the point in a command  
where you are uncertain what the attributes are.  
Help with specific commands  
If you enter the help command at the top level, you are considered to be in the  
Shell Object, and the help command will provide help on the Shell commands  
focus, show, flush, rescan, and commit. Using the help command on objects  
(such as /cx, /cx/ux, /cx/px, /cx/bbu, /ex, /ex/slotx, /ex/fanx, and /ex/tempx),  
displays all possible sub-commands associated with the object.  
For example: help on the controller object /cx, will display all the sub-  
commands associated with the controller /cx, like this:  
//localhost> help /cx  
/cx show  
/cx show attribute [attribute ...]  
where attribute is:  
achip|allunitstatus|autocarve|autorebuild(9550SX and 9590SE  
only)|bios|carvesize(9000 series)|ctlbus(9550SX and 9590SE  
only)|driver|drivestatus|exportjbod|firmware|memory|  
model|monitor|numdrives|numports|numunits|pcb|pchip|  
serial|spinup|stagger|unitstatus|ondegrade(9000S only)  
/cx show all where all means attributes and configurations.  
/cx show diag  
/cx show alarms [reverse]  
/cx show rebuild  
/cx show verify  
/cx show selftest  
(9000 only)  
(9000 only)  
(9000 only)  
/cx add type=<RaidType> disk=<p:-p..> [stripe=<Stripe>]  
[noscan] [nocache][group=<3|4|5|6>] [autoverify] [ignoreECC]  
[name=string (9000 only)] [storsave=<protect|balance|  
perform[quiet]>(9550SX and 9590SE only)] RaidType={raid0,  
raid1, raid5, raid10, raid50, single, spare, JBOD  
(7000/8000 only)}  
/cx add rebuild=ddd:hh:duration  
/cx add verify=ddd:hh:duration  
/cx add selftest=ddd:hh  
(9000 only)  
(9000 only)  
(9000 only)  
/cx del rebuild=slot_id  
/cx del verify=slot_id  
(9000 only)  
(9000 only)  
83  
   
Chapter 2. CLI Syntax Reference  
/cx del selftest=slot_id  
(9000 only)  
/cx set exportjbod=on|off  
/cx set ondegrade=cacheoff|follow  
/cx set spinup=nn  
/cx set stagger=nn  
/cx set autocarve=on|off  
(9000 only)  
(9500S only)  
(9000 only)  
(9000 only)  
(9000 only)  
/cx set rebuild=enable|disable|<1..5> enable|disable  
for 9000 only)  
/cx set verify=enable|disable|<1..5> (enable|disable  
for 9000 only)  
/cx set selftest=enable|disable [task=UDMA|SMART](9000 only)  
/cx set autorebuild=on|off  
(9550SX and 9590SE only)  
/cx update fw=filename_with_path [force] (9000 only)  
/cx flush  
/cx commit  
/cx start mediascan  
/cx stop mediascan  
(Windows only) (Also known as shutdown)  
(7000/8000 only)  
(7000/8000 only)  
/cx rescan [noscan] NOTE: Does not import non-JBOD on  
7000/8000 models.  
//localhost>  
Help with attributes  
As you work with specific objects or commands, you can also use ? to get  
help.  
For example: If you enter the command /c0 showand then need help on what  
specific attribute syntax is possible, you can use ? to get help as following:  
//localhost> /c0 show ?  
/cx show  
/cx show attribute [attribute ...]  
where attribute is:  
achip|allunitstatus|autocarve(9000 series)|  
autorebuild(9550SX only)|bios|carvesize(9000series)|driver|  
drivestatus|exportjbod|firmware|memory|model|monitor|  
numdrives|numports|numunits|ctlbus(9550SX and 9590SE only)  
serial|ondegrade (9000S only)|pcb|pchip|spinup|stagger|  
unitstatus|  
/cx show all where all means attributes and configurations.  
/cx show diag  
/cx show alarms [reverse]  
/cx show rebuild  
/cx show verify  
/cx show selftest  
//localhost>  
(9000 only)  
(9000 only)  
(9000 only)  
84  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
 
Help Commands  
help  
This help command provide a table of contents, providing help with the  
overall navigation of the CLI commands. Typical output looks like the  
following.  
//localhost> help  
Copyright(c) 2004, 2005 Applied Micro Circuits Corporation  
(AMCC). All rights reserved.  
AMCC/3ware CLI (version 2.x)  
Commands Description  
--------------------------------------------------------------  
show  
Displays information about controller(s), unit(s) and port(s).  
Flush write cache data to units in the system.  
Rescan all empty ports for new unit(s) and disk(s).  
Update controller firmware from an image file  
Controller specific commands.  
flush  
rescan  
update  
/cx  
/cx/ux  
/cx/px  
/cx/bbu  
/ex  
Unit specific commands.  
Port specific commands.  
BBU specific commands.  
Enclosure specific commands.  
(9000 only)  
(9550SX and 9590SE only)  
/ex/slotx Slot specific commands.  
/ex/fanx Fan specific commands.  
/ex/tempx Enclosure Temperature Sensor specific commands.  
Certain commands are qualified with constraints of controller type/model  
support. Please consult the tw_cli documentation for explanation of the  
controller-qualifiers.  
The controller-qualifiers of the Enclosure commands (/ex) also apply to  
Enclosure Element specific commands (e.g., /ex/elementx).  
Type help <command> to get more details about a particular command. For more  
detail information see tw_cli's documentation.  
help show  
help flush  
This command provides specific show-related help, illustrating various ways  
to use the show command. It provides reports on Controllers, Units and  
information.  
This command provides specific flush-related help, illustrating various ways  
to use the flush command. See the section “Shell Object Commands” on  
85  
     
Chapter 2. CLI Syntax Reference  
help rescan  
This command provides specific rescan related help, illustrating various ways  
to use the rescan command. See the section “Shell Object Commands” on  
help update  
This command provides specific update-related help. See “Shell Object  
Commands” on page 19 for more information.  
help commit  
This command provides specific commit related help, illustrating various  
ways to use the commit command. See the section “Shell Object Commands”  
on page 19 for more information.  
help focus  
help /cx  
This command provides specific focus related help, illustrating various ways  
to use the focus command. See the section “Shell Object Commands” on  
This command provides specific controller /cx related help, illustrating  
various commands associated with the controller /cx. See the section  
help /cx/ux  
help /cx/px  
This command provides specific unit /cx/ux related help, illustrating various  
commands to use on a unit /cx/ux. See the section “Unit Object Commands”  
on page 53 for more information.  
This command provides specific /cx/px related help, illustrating various ways  
to use the /cx/px command. See the section “Port Object Commands” on  
86  
3ware Serial ATA RAID Controller CLI Guide for the Power Mac G5  
             
Return Code  
help /cx/bbu  
help /ex  
This command provides specific /cx/bbu related help, illustrating various  
ways to use the /cx/bbu command. See the section “BBU Object Commands”  
on page 74 for more information.  
This command provides specific enclosure /ex related help, illustrating  
various commands associated with the enclosure /ex. See the section  
help /ex/slotx  
This command provides specific slot /ex/slotx related help, illustrating  
various ways to use /ex/slotx. See the section “Enclosure Object Commands”  
on page 79 for more information.  
help /ex/fanx  
This command provides specific fan /ex/fanx related help, illustrating various  
ways to use the /ex/fanx command. See the section “Enclosure Object  
Commands” on page 79 for more information.  
help /ex/tempx  
This command provides specific temperature sensor /ex/tempx related help,  
illustrating various ways to use the /ex/tempx command. See the section  
Return Code  
While informative messages are written to standard output, error messages are  
written to standard error. On success, 0 is returned. On failure, 1 is returned.  
87  
           

GE General Electric Cooktop ZEU36R User Manual
GE 25942xx2 User Manual
Fisher Paykel Cooktop CG604DX User Manual
Electrolux Thermaline WHWUSAOOOO User Manual
Eizo FLEXSCAN L761T C User Manual
Dell G2410H User Manual
Asus Computer Monitor VH196 User Manual
Asus Computer Monitor 90LMC3101Q01041C User Manual
Aluratek Cyclone ACQES350F4TB User Manual
Acer Computer Monitor G245HQ User Manual