has your local clariion array lost a drive?

[Manuel Contreras]

Other then the following HP command...is anyone familiar w/another way to determine if your local clariion array LOST a drive ?

/opt/Navisphere/bin/navicli -h SP? getlun | grep Removed | sort -u


looking for options before I create something for cron.
thanks

[Pedro Cirne]

Hi,

If your array looses a drive, then when you do ioscan -fnCdisk it will become NO_HW.

Probably you'll root will also receive a mail from EMS with a critical event.

Enjoy :)

Pedro

[Uwe Zessin]

No, Manuel talks about the loss of a physical disk drive in a Fibre Channel RAID array. The servers will most likely only see RAID sets mapped in the LUN address space, but not individual disk drives. So the loss of a single disk will not be noticed by the host.

[Pedro Cirne]

Hi,

You're right Uwe, I completely agree with you, my interpretation was not correct.

Manuel, If you have NaviAgent installed on hp-ux, it can be configured to report a loss of a phisical disk in the storage!

Enjoy :)

Pedro

[Manuel Contreras]

you are correct sir!

# ioscan -fnC disk | grep NO_HW

will not return anything because of raidX.

when I say lost drive, I am referring to physical disks (spindles).
external arrays normally utilize RAID...because of this your box will not report NO_NW.

raidX is your friend = no lost data

dmesg might complain some...but that's about it.

[Manuel Contreras]

/opt/Navisphere/bin/navicli -h SP? getdisk | grep Removed

will also work...

It is looking like the navicli command may be the only option, besides monitoring through a navisphere agent.

thank you all for your input.

[Florian Heigl (new acc)]

Only a small addition:

rather run navicli -getall > /tmp/getall.spN and filter this output, it's quite a bit faster and You'll always have all data available.

Also You can use navicli getlog -100 for the last 100 messages from the event log inside the box.

[Manuel Contreras]

thanks for the log idea, don't know If I should give you any points (just found out when it actually failed) :)

aWHILEago
(2580)Storage Array Faulted Bus 0 Enclosure 2 : Faulted Bus 0 Enclosure 2 Disk 5 : Removed

the getdisk w/DID looks promissing as well:

# /opt/Navisphere/bin/navicli -h 192.25.235.18 getdisk 0_2_5
Bus 0 Enclosure 2 Disk 5
Vendor Id: SEAGATE
Product Id: ST373207 CLAR72
Product Revision: 6A0C
Lun: 80 81 82 83 84 85 86 183 184 185 186 187 188 189 190 191 277 278 279 280 281 282 283 284 285 286 287 288 289 290 337
391 392 393 394 395 396 397 398
Type: 80: RAID5 81: RAID5 82: RAID5 83: RAID5 84: RAID5 85: RAID5 86: RAID5 183: RAID5 184: RAID5 185: RAID5 186: RAID5 187
90: RAID5 191: RAID5 277: RAID5 278: RAID5 279: RAID5 280: RAID5 281: RAID5 282: RAID5 283: RAID5 284: RAID5 285: RAID5 286: RAID5 287: RAID5
5 337: RAID5 338: RAID5 384: RAID5 385: RAID5 386: RAID5 387: RAID5 388: RAID5 389: RAID5 390: RAID5 391: RAID5 392: RAID5 393: RAID5 394: RAI
AID5 398: RAID5
State: Equalizing



drive has been replaced, by yours truly...