ILO100 w/ DL185 G5 becomes unresponsive

[T. Hiukkanen]

The (integrated) ilo card on the server becomes unresponsive after running for some time (24h+) while the server itself seems ok. The card stops answering ping, telnet and http requests. The only thing i found that will make the card respond again is doing 'ipmitool bmc reset cold'.
The server has got the latest firmware (including ILO fw 3.05) installed, and is running in a zero-load installation environment.
Has anyone else witnessed a similar problem?

Thanks for your insights.

[T. Hiukkanen]

It seems there is something fundamentally wrong. Resetting the BMC (ipmitool bmc reset cold) apparently does something to the P400 controller as well, because soon after (30min+) it starts claiming some of the disks have failed. Soon after the server is unusable.

---
May 18 13:06:48 : cciss: cmd ffff880037800000 has CHECK CONDITION sense key = 0x3
May 18 13:06:48 : end_request: I/O error, dev cciss/c0d0, sector 170932640
May 18 13:06:48 : Buffer I/O error on device cciss/c0d0p2, logical block 8257540
May 18 13:06:48 : lost page write due to I/O error on cciss/c0d0p2
---

Event code 5/2/3 with tag 38
with message: Inconsistent stripe, LDrv=1 LBA=0x0741BA800-0x0741BA9FF

Event code 5/1/0 with tag 39
with message: Fatal drive error, Port=1I Box=1 Bay=0

---

Funny thing is that before resetting the bmc everything works smoothly, even bonnie++ runs multiple iterations as expected.

Again, if anyone has had any similar experiences, i would appreciate any speculation very much. Thanks.

[T. Hiukkanen]

Just to share the solution;

LO100 issue fixed by replacing system board. SATA drive firmware update dated 2009-06-01 fixed P400 under-load-inconsistent-stripe issues.

[Jason404]

Updating the firmware to v3.11, on my LO100c, seems to have solved connection problems. I think there are new firmwares for the other models too.

[T. Hiukkanen]

As said, ILO problem solved by firmware update. Strange disk problems continue; more at http://forums13.itrc.hp.com/service/forums/questionanswer.do?threadId=1362671