How to 'activate' my Clariion disks ?

Sorry for my noob aix question, brand new in aix-systemadmin

This my config:

root-at-server21:/tmp# lsdev -Ccdisk
hdisk0 Available 06-08-01-3,0 16 Bit LVD SCSI Disk Drive
hdisk1 Available 06-08-01-4,0 16 Bit LVD SCSI Disk Drive
hdisk2 Available 01-08-02 EMC CLARiiON FCP LUNZ Disk
hdisk3 Available 01-08-02 EMC CLARiiON FCP LUNZ Disk
hdisk4 Available 02-08-02 EMC CLARiiON FCP LUNZ Disk
hdisk5 Available 02-08-02 EMC CLARiiON FCP LUNZ Disk
hdisk6 Available 01-08-02 EMC CLARiiON FCP RAID 5 Disk
hdisk7 Available 01-08-02 EMC CLARiiON FCP RAID 5 Disk
hdisk8 Available 02-08-02 EMC CLARiiON FCP RAID 5 Disk
hdisk9 Available 02-08-02 EMC CLARiiON FCP RAID 5 Disk
hdiskpower0 Available 02-08-02 PowerPath Device

root-at-server21:/tmp# lspv
hdisk0 00c78420f33b3cc8 rootvg
active
hdisk1 none None
hdisk2 none None
hdisk3 none None
hdisk4 none None
hdisk5 none None
hdisk6 none None
hdisk7 none None
hdisk8 none None
hdisk9 none None
hdiskpower0 00c78420b82e895a None
root-at-server21:/tmp#

I can't configure my disks with LVM (or from within smit, smit just
doesn't recognize them)
Shouldn't there be a serial number next to the disks ?

On 07/04/2008 01:39 PM, GoldWing wrote:
> Shouldn't there be a serial number next to the disks ?

It's a PVID not a serial number - and the single valid disk in your
output is 'hdiskpower0'. It can be treated like any other disk i.e. you
can add it to a VG.

The underlying paths for the hdiskpower come from the hdisks, and you
can see the active paths for an hdiskpower device by running 'powermt
display dev=hdiskpower0'. You should not use the hdisk (path) in a VG
instead of the hdiskpower device.

At a glance it seems you have hdisk6-9 as paths for hdiskpower0, over 2
separate HBAs (good!)

The hdisk2-5 LUNZ devices would have been created when a external disk
array without any configured Luns was detected by cfgmgr.

They can be removed since they effectively are the remnants of an
incomplete disk array configuration.

If cfgmgr redetects any LUNZ devices you can confidently blame your
storage administrator.

cheers
Niel

On 4 jul, 14:10, Niel Lambrechts wrote:
> On 07/04/2008 01:39 PM, GoldWing wrote:
>
> > Shouldn't there be a serial number next to the disks ?
>
> It's a PVID not a serial number - and the single valid disk in your
> output is 'hdiskpower0'. It can be treated like any other disk i.e. you
> can add it to a VG.
>
> The underlying paths for the hdiskpower come from the hdisks, and you
> can see the active paths for an hdiskpower device by running 'powermt
> display dev=hdiskpower0'. You should not use the hdisk (path) in a VG
> instead of the hdiskpower device.
>
> At a glance it seems you have hdisk6-9 as paths for hdiskpower0, over 2
> separate HBAs (good!)
>
> The hdisk2-5 LUNZ devices would have been created when a external disk
> array without any configured Luns was detected by cfgmgr.
>
> They can be removed since they effectively are the remnants of an
> incomplete disk array configuration.
>
> If cfgmgr redetects any LUNZ devices you can confidently blame your
> storage administrator.
>
> cheers
> Niel

Wow, that seems great, really need to find some books about this,
maybe on safari.
Anyway:

oot-at-server21:/tmp# powermt display dev=hdiskpower0
Pseudo name=hdiskpower0
CLARiiON ID=CK200074000039 [server21]
Logical device ID=60060160D7321E0057B3B356AF21DD11 [LUN 36]
state=alive; policy=CLAROpt; priority=0; queued-IOs=0; write-
throttle=on
Owner: default=SP B, current=SP B Array failover mode: 1
================================================== ============================
---------------- Host --------------- - Stor - -- I/O Path - --
Stats ---
### HW Path I/O Paths Interf. Mode State Q-
IOs Errors
================================================== ============================
0 fscsi0 hdisk6 SP A1 active dead
0 0
0 fscsi0 hdisk7 SP B1 active alive
0 0
1 fscsi1 hdisk8 SP A0 active alive
0 0
1 fscsi1 hdisk9 SP B0 active alive
0 0


Is that normal that hdisk6 is dead ? Ho to 'undead' it ?

>
> oot-at-server21:/tmp# powermt display dev=hdiskpower0
> Pseudo name=hdiskpower0
> CLARiiON ID=CK200074000039 [server21]
> Logical device ID=60060160D7321E0057B3B356AF21DD11 [LUN 36]
> state=alive; policy=CLAROpt; priority=0; queued-IOs=0; write-
> throttle=on
> Owner: default=SP B, current=SP B Array failover mode: 1
> ================================================== ============================
> ---------------- Host --------------- - Stor - -- I/O Path - --
> Stats ---
> ### HW Path I/O Paths Interf. Mode State Q-
> IOs Errors
> ================================================== ============================
> 0 fscsi0 hdisk6 SP A1 active dead
> 0 0
> 0 fscsi0 hdisk7 SP B1 active alive
> 0 0
> 1 fscsi1 hdisk8 SP A0 active alive
> 0 0
> 1 fscsi1 hdisk9 SP B0 active alive
> 0 0
>
>
> Is that normal that hdisk6 is dead ? Ho to 'undead' it ?

No, not normal... Probably a problem at the array, or a zoning issue.

W.B. wrote:
>> oot-at-server21:/tmp# powermt display dev=hdiskpower0
>> Pseudo name=hdiskpower0
>> CLARiiON ID=CK200074000039 [server21]
>> Logical device ID=60060160D7321E0057B3B356AF21DD11 [LUN 36]
>> state=alive; policy=CLAROpt; priority=0; queued-IOs=0; write-
>> throttle=on
>> Owner: default=SP B, current=SP B Array failover mode: 1
>> ================================================== ============================
>> ---------------- Host --------------- - Stor - -- I/O Path - --
>> Stats ---
>> ### HW Path I/O Paths Interf. Mode State Q-
>> IOs Errors
>> ================================================== ============================
>> 0 fscsi0 hdisk6 SP A1 active dead
>> 0 0
>> 0 fscsi0 hdisk7 SP B1 active alive
>> 0 0
>> 1 fscsi1 hdisk8 SP A0 active alive
>> 0 0
>> 1 fscsi1 hdisk9 SP B0 active alive
>> 0 0
>>
>>
>> Is that normal that hdisk6 is dead ? Ho to 'undead' it ?
>
> No, not normal... Probably a problem at the array, or a zoning issue

I think you're actually looking at the "paths" that connect to your
hdiskPOWERX! Have you run powermt config??? We use EMC in my
environment (I'd prefer IBM DS :-( ). Your VGs (if I'm not
misinterpreting this thread/question) are created on HDISKPOWERx NOT on
the "hdiskXXX" (those are your PATHS! to the "hdiskpowerXXX")...then
again I COULD be wrong :-) )

Peter

On 07/05/2008 10:02 PM, Bob Dobbs wrote:
> W.B. wrote:
>>> oot-at-server21:/tmp# powermt display dev=hdiskpower0
>>> Pseudo name=hdiskpower0
>>> CLARiiON ID=CK200074000039 [server21]
>>> Logical device ID=60060160D7321E0057B3B356AF21DD11 [LUN 36]
>>> state=alive; policy=CLAROpt; priority=0; queued-IOs=0; write-
>>> throttle=on
>>> Owner: default=SP B, current=SP B Array failover mode: 1
>>> ================================================== ============================
>>> ---------------- Host --------------- - Stor - -- I/O Path - --
>>> Stats ---
>>> ### HW Path I/O Paths Interf. Mode State Q-
>>> IOs Errors
>>> ================================================== ============================
>>> 0 fscsi0 hdisk6 SP A1 active dead
>>> 0 0
>>> 0 fscsi0 hdisk7 SP B1 active alive
>>> 0 0
>>> 1 fscsi1 hdisk8 SP A0 active alive
>>> 0 0
>>> 1 fscsi1 hdisk9 SP B0 active alive
>>> 0 0
>>>
>>>
>>> Is that normal that hdisk6 is dead ? Ho to 'undead' it ?
>> No, not normal... Probably a problem at the array, or a zoning issue

It sometimes (usually?) is necessary to delete the dead paths with rmdev
once the underlying problem has been addressed. Then run cfgmgr and
powermt display dev=all should show a replacement 'hdiskX' for the dead
path.

> Have you run powermt config??? We use EMC in my
> environment (I'd prefer IBM DS :-( ).

That won't remove/fix a dead path. You can run:
powermt check

to remove all the dead paths, fix the problem and then run cfgmgr. If
the O/S detects the hdisks but they (or the hdiskpower) don't show up in
powermt output, you may need to run 'powermt config'.

These steps should not affect the availability of data on the hdiskpower
device through the other paths.

> Your VGs (if I'm not
> misinterpreting this thread/question) are created on HDISKPOWERx NOT on
> the "hdiskXXX"

Correct.


cheers
Niel