random reboot

[Martha Mueller]

I found this thread on the server forum. Does anyone have a comment?


http://forums.itrc.hp.com/cm/QuestionAnswer/1,,0xe05d0fe6d0f7d61190050090279cd0f9,00.html

[John Poff]

Hi,

I saw that thread. We have three rp8400's and an rp7400 (N4000) running 11.11 now. We haven't seen any reboots. I have seen the issue that Prashant mentioned about geting more EMS messages, and more complaints about the GSP. When I check the GSP it is always fine.

I mentioned the 'hold' to our local CE Friday afternoon. He just raised an eyebrow and said he hadn't heard anything. I guess it could be an internal problem HP is working through, or it could be that Justin's salesman ran into trouble and needed an excuse to buy some more time. ;)

I'm watching that thread and the server forum, but I'm not real worried about the issue at this point.

JP

[Dave Elliott]

Guys
I have had a major problem with our n4000,
it would just reboot out of the blue.
we are using 11.00 not 11i though, but after alot of crash file analysys we replaced three CPU's only to find that it turned out to be a mainboard fault that caused it to report in the crash files that cpu's were failing.
might i suggest that if people have had this problem that they interigate there crash files directory in /car/adm/crash/crash?,
also there is a peaice of software that you can get from hp called Crashinfo, this will unpack the files and make then more easily readable.
hope this helps
since changing the mainboard no more issues.

[Pete Randall]

Martha,

I've had an N4000 for 3 years now with no problems. I would comment though, that HP is usually more forthcoming and proactive (at least the "old" HP was). We also have an FC60 disk array and, even though we had experienced no difficulty at all, HP came out and replaced the firmware, the cables, and the SCSI bus controllers because others had reported problems.

I'm hoping that the "new" HP will be as open and honest with us.

Pete

[harry d brown jr · Accepted solution]

Martha and others,

This is a KNOWN issue with HP, yet they have NO clue, or they haven't told anyone what the cure is.

It is NOT a CRASH, it is NOT a REBOOT, it is a POWER FAILURE, therefore there will be NO crash logs.

I also have N-class servers that do NOT experience this issue, yet I have 4 out of 20 N-class servers FAILING, which is 20%. HP had better fix this fast, or we will be requiring them to replace them with rp8410's at their expense!

As for the POWER FAILURE, it is NOT caused by any EXTERNAL FAILURES, as we have had our electricians set up monitoring on the power before the rails, and after the rails, showing no indication that the POWER SOURCE was an issue.


live free or die
harry

[Scott Donaldson]

D Class machines suffer from a similar problem. I've seen it in my shop over the last year (d260) and it was discussed in the forums about 2 years ago where faulty DC Power switches were implicated. See:

http://forums.itrc.hp.com/cm/QuestionAnswer/1,,0xb7e66c96588ad4118fef0090279cd0f9,00.html

[Ian Cameron]

Hi Everyone,

There was an issue with some D class boxes and faulty power supplies/switches a while ago. I am unfamiliar with the makeup of the N and RP class boxes but the symptoms sound the same.

I had a d370 running 11i earlier this year that was driving us nuts with its desire to reboot and not give us any feedback. A very helpful and competant HP FE worked on it off and on for over three months. He eventually replaced everything but the cpu board, case, drives and RAM.Like Harry we had all power checked from breaker to box and nothing found there. After two power supplies and several boards the processor board finally failed in September(tombstone file with data this time) and we haven't had a problem since.

Don't know if any of this helps; seems you have to roll the component replacement dice to get the right ones sometimes...

[Todd Larchuk]

I have had many spontaneous reboots on K580 running 11i. Usually have 2 or 3 reboots in a period of a day or so, then the system quites down. Usually have no syslog or crash info, but this latest time got a memory failing warning. Haven't yet had any reboots on L, but I am responsible for only a few of them.

[harry d brown jr]

Ian,

We actually had the system board replaced, and it still power cycles. I really think it's a power supply issue.

live free or die
harry

[Ian Cameron]

Martha, Harry, et al.

Went back through the emails from the FE and there was a point where he changed out all the fans on the D box as well. Seems there were some instances of case fans (FE rumour/field experience ;o))that were sending out 'noise' on the status monitoring lines that caused the power supply to think that both had failed and initiate a shutdown. The box was OK after that until the processor board (which had already been changed twice in the course of the 'rebooting' fixes) actually went down with errors at the end of August.

Ian

[harry d brown jr]

IAN,

Thanks for the INFO. I'll have our SA's check with our HP dudes.

thanks again

live free or die
harry