HBA Powerpath query

[saucyboy]

Hi guys

I have been asked to find issues with a power path down.

I am new to this.

also been asked to look if HBA is down.
i issues a "powermt display dev=all |more" which shows
# powermt display dev=all |more
Pseudo name=hdiskpower0
Symmetrix ID=000190103539
Logical device ID=040A
state=alive; policy=SymmOpt; priority=0; queued-IOs=0
==============================================================================
---------------- Host --------------- - Stor - -- I/O Path - -- Stats ---
### HW Path I/O Paths Interf. Mode State Q-IOs Errors
==============================================================================
0 fscsi0 hdisk3 FA 4aA active alive 0 0

Pseudo name=hdiskpower1
Symmetrix ID=000190103539
Logical device ID=040B
state=alive; policy=SymmOpt; priority=0; queued-IOs=0


so a HBA is not there, also did a cfgmgr -v, did not come back...

any ideas?

also how do i check the power path?

 

[spamly]

Does your system see the other physical adapter?
lsdev -Cc adapter

 

[saucyboy]

yes it see the othe adpater

fcs0 Available 02-08 FC Adapter
fcs1 Available 02-09 FC Adapter
fcs2 Available 03-08 FC Adapter
fcs3 Available 03-09 FC Adapter

 

[spamly]

Four adapters and all "Available"... Okay, do you know which ones have been zoned and have the luns assigned to them?

 

[saucyboy]

Hi

No idea which ones been zoned and luns assigned, i've not worked on SAN config before and new to the game.

I read that if the path is down and adapters are available an ioscan should be run, is this correct, if not please advise.

Thankyou

 

[spamly]

This is a little tricky. As you're probably aware, luns are assigned on the EMC side of the house. The server discovers them through a "cfgmgr" and "powermt config" then creates powerpath devices (hdiskpower*) for any luns visible down multiple paths.

If a lun is only discovered and/or visible down one path, powerpath will still create a powerpath device, but it will only use the single path. I can't say for sure if this is how yours is setup. If this IS the case, the solution of it lies more on the SAN side of the house. Also depending on the disk subsystem, clariion vs. symmetrix, this may vary slightly.

Can you check to see what adapters are zoned and logged into your fabric? Hopefully, a SAN admin can tell you this. Then, if at all possible, find out which adapters have luns assigned to them. Your SAN admin may tell you just the WWNN. If that's the case, let us know and we can help translate that to the specific adapter.

Also, it has been some time since I've worked with powerpath devices. I'm more of a Hitachi guy now. Can somebody out there tell me if we can add a lun to an existing powerpath (hdiskpower) device? I know it is possible to "repair" (see 'man powermt') an existing device, but I can't remember if you can add a path/lun on the fly.

 

[saucyboy]

fsc0 and fsc2 are the cards, all lights flashing on both

 

[saucyboy]

Hey guys,

i sorted the problem i think

so , I got the location codes for both adapters

checked physically on the server, the lights were flashing

advised the SAN ADMIN - he checked and found its a zoneing issue and fixes it.

I re-run cfgmgr -v

run : powermt display dev=all |more

the path is back...

thankyou for your help