HBA on vpar

[Ronelle A van Niekerk]

I have a vpar with two HBA's assigned to it.
One is working perfectly and is happily talking to the SAN.
The other I have only just plugged into a switch, but it came up as AWAITING_LINK_UP.
So after a bit of playing around on my part I am now stuck with an fc card that doesn't work.
I tried ioscans and insf -e's till my fingers were numb. Then I removed the special file for that card, but an insf -e doesn't restore it - even if I use the -H option.

What am I missing?

# ioscan -funC fc
Class I H/W Path Driver S/W State H/W Type Description
===================================================================
fc 2 1/0/12/1/0 td CLAIMED INTERFACE HP Tachyon XL2 Fibre Channel Mass Storage Adapter
fc 3 1/0/14/1/0 td CLAIMED INTERFACE HP Tachyon XL2 Fibre Channel Mass Storage Adapter
/dev/td3

[Torsten.]

"AWAITING_LINK_UP" that is what the HBA is doing - just waiting.

3 possibilities:

- bad HBA (no light)
- bad cable
- bad switch port

I would check this first.

[Sandilyan S]

hi,

Please see the SAN switch

>show port

and colect the logs from SAN switch

>support show

to check the port u can see the WWN or not

OR Check Fiber Cable

[Torsten.]

What WWN do you expect to see without a link?

Carefully check if there is some light on the HBA, cable, switchport ...

[Sandilyan S]

hi,

please check the Fiber cable only

[Ronelle A van Niekerk]

OK, before I messed around with the car:
I checked the cable, no light coming through when plugged into the server.
Light soming through when plugged into the switch.
So it's not the cable.

I'm pretty sure it's the card.

I'm not sure if you read the whole thread, but my problem now is NOT that it is AWAITING_LINK_UP, but that there is no longer a driver associated with it and I can't get it to associate one...

I removed the special file and the insf -e won't recreate it.

[Torsten.]

Did you try this?

# insf -e -H 1/0/12/1/0

This should do the job. How did you remove the device file?

[YAQUB_1]

Hi,

1. U change the fiber cable
2. "ioscan" & "ioscan -fn" then "insf -ev"
3. send us MP event log (MP:VW>sel---press D)
4. check the HBA card LED status (AMBER or Green)

put the MP logs here...

Thanks-Yaqub
I am a Customer Advocate!!!

[Torsten.]

>> "I'm pretty sure it's the card. "


Me too.


No idea what HBA related information he want to read from the MP logs.

[Ronelle A van Niekerk]

"but an insf -e doesn't restore it - even if I use the -H option."

I removed the file with: rmsf /dev/td2

I have no idea what is going on here.
The card is assigned to the vpar - otherwise I wouldn't even see it.

Maybe it's faulty.

[Torsten.]

Is the result different for

# ioscan -fnCfc

I never used the -u option, but according to the man page:

"-u
Scan and list usable I/O system devices instead of the actual hardware. Usable I/O devices are those having a driver in the kernel and an assigned instance number."


Well, it does not scan and list the actual H/W...

[YAQUB_1]

Hi,

>>> Maybe it's faulty.
==> Without MP log we can not confirm Ur HBA card faulty.

Please send Ur system MP logs....

Thanks--Yaqub
I am a Customer Advocate!!!

[Torsten.]

My conclusion at this point:

Because of the "no light" fact I would assume a bad laser or some other failure directly on the card. As long as the card is CLAIMED, I would not assume to see something in the MP logs, but maybe some diag/ems messages (syslog/ root mailbox). I guess the reboot may solve the device file problem.

Last thing to try could be a power cycle or re-seat of the card (rad/olrad), but finally it will end with a HBA replacement.

[Ronelle A van Niekerk]

" Is the result different for

# ioscan -fnCfc "

I've run both, same result.

I don't know how to get teh MP logs - never heard of them.

[YAQUB_1]

Hi,

check the below URL site:-
http://docs.hp.com/en/AB464-9003C/ch03s07.html

Thanks--Yaqub
I am a Customer Advocate!!!

[Ronelle A van Niekerk]

I don't think this is very helpful:
[rx7640-ux01] MP:VWR (,,+,-,?,F,L,J,D,K,R,T,A,C,U,^B) > [rx7640-ux01] +
1629 SFW 0,1,2 2 448001cf12e0021e 00ffff01ff1bff74 MEM_SBE_IN_RANK
1629 02/09/2008 14:53:13
1630 SFW 0,1,2 2 418001cf12e00220 0000000037fb4180 MEM_SBE_IN_RANK
1630 02/09/2008 14:53:13
1631 SFW 0,1,2 2 478001cf12e00222 37fb102000005908 MEM_SBE_IN_RANK
1631 02/09/2008 14:53:13
1632 SFW 0,1,2 2 478001cf12e00224 0000000008d8ed40 MEM_SBE_IN_RANK
1632 02/09/2008 14:53:13
1633 2147b055c3020226 ff0f066f001f0300 IPMI Type-02 Event
1633 02/11/2008 14:03:47
1634 2147b05b70020227 ff0f066f001f0300 IPMI Type-02 Event
1634 02/11/2008 14:28:00
1635 2147b05ee1020228 ff0f066f001f0300 IPMI Type-02 Event
1635 02/11/2008 14:42:41
1636 SFW 0,1,2 2 448001cf12e00229 00ffff01ff1bff74 MEM_SBE_IN_RANK
1636 02/12/2008 12:38:32
1637 SFW 0,1,2 2 418001cf12e0022b 0000000037fb4180 MEM_SBE_IN_RANK
1637 02/12/2008 12:38:32
1638 SFW 0,1,2 2 478001cf12e0022d 37fb102000005908 MEM_SBE_IN_RANK
1638 02/12/2008 12:38:32
1639 SFW 0,1,2 2 478001cf12e0022f 0000000008d8ed40 MEM_SBE_IN_RANK
1639 02/12/2008 12:38:32

[rx7640-ux01] MP:VWR (,,+,-,?,F,L,J,D,K,R,T,A,C,U,^B) > [rx7640-ux01] +
1639 SFW 0,1,2 2 478001cf12e0022f 0000000008d8ed40 MEM_SBE_IN_RANK
1639 02/12/2008 12:38:32
1640 SFW 0,1,0 2 448001cf10e00231 00ffff01ff1bff74 MEM_SBE_IN_RANK
1640 03/01/2008 16:30:38
1641 SFW 0,1,0 2 418001cf10e00233 0000000037fb4180 MEM_SBE_IN_RANK
1641 03/01/2008 16:30:38
1642 SFW 0,1,0 2 478001cf10e00235 37fb102000005908 MEM_SBE_IN_RANK
1642 03/01/2008 16:30:38
1643 SFW 0,1,0 2 478001cf10e00237 0000000008d8ed40 MEM_SBE_IN_RANK
1643 03/01/2008 16:30:38
1644 SFW 0,1,2 2 448001cf12e00239 00ffff01ff1bff74 MEM_SBE_IN_RANK
1644 03/05/2008 19:41:03
1645 SFW 0,1,2 2 418001cf12e0023b 0000000037fb4180 MEM_SBE_IN_RANK
1645 03/05/2008 19:41:03
1646 SFW 0,1,2 2 478001cf12e0023d 37fb102000005908 MEM_SBE_IN_RANK
1646 03/05/2008 19:41:03
1647 SFW 0,1,2 2 478001cf12e0023f 0000000008d8ed40 MEM_SBE_IN_RANK
1647 03/05/2008 19:41:03
*********************** End of log ***********************

[Torsten.]

You have some single bit errors, but this is a different story ...

[Ronelle A van Niekerk]

I'm giving you zero points just for that!

[Torsten.]

Why?

the system has logged single bit errors in certain RAM modules. This is not related to your current problem.


If you want to dig into this, we have to take a closer look on this.

But as mentioned above, this is a different problem, not related to your FC problem.

[Ronelle A van Niekerk]

I thought you were being sarcastic.

Apologies.

[Torsten.]

Not at all ... still trying to help you.

[Ronelle A van Niekerk]

There are HBA cards assigned to the other vpar on this server that I could possibly reassign to this one and see if they work better.
I just can't reboot the server at the moment - it's being used and this "faulty" HBA isn't actually causing any problems, except for me.

[Torsten.]

You may consider to replace the card online (see manual for rad/olrad), but this is up to you.

Next time apologize with points - LOL!

Good luck!