Forum Stats

  • 3,814,069 Users
  • 2,258,814 Discussions
  • 7,892,523 Comments

Discussions

SB2500 silver, shuts it self down

1027031
1027031 Member Posts: 2

Hi,

I have several sunblade 2500's which seem to shut them selves down quite regular with false temp readings. I don't think its the hardware as it is happening on pretty much all of them. The error reported in the messages file is

SUNW_piclenvd: 'int-amb0' sensor temperature -127 outside safe operating limits (0...70).

Now there is no way its -127 here so I believe this is either a fw issue or a patch (os is sol 10 08/07)

The workstation is running obp 4.17.3, I have checked the latest obp and I can't find a bug fix which fits this issue.

I would appreciate any help

Many thanks

Tagged:

Answers

  • I was hoping for others to respond but a couple of days have passed so I'll give it a try.   There are a couple of Oracle employees that occasionally post here as individuals, not as spokespeople of the company, but they may not have seen your post yet.     (It's been more than eight years since I had any detailed access as a former Sun employee, so my thoughts may not quite be accurate.)

    .

    .

     SUNW_piclenvd: 'int-amb0' sensor temperature -127 outside safe operating limits (0...70).
    

    Yes, a negative 127C is a bit doubtful for the internal chassis ambient temperature.

    You could have a faulty sensor, but that is ruled out because multiple boxes are behaving the same.    That will then  point to a firmware or software issue.    I suspect both.

     The workstation is running obp 4.17.3, I have checked the latest obp and I can't find a bug fix which fits this issue.
    

    You state you've looked at the README for applicable OBP bug fixes, so that suggests you have MOS credentials.  Good.   You might consider getting the latest OBP patch #140689 and at least keep it archived locally should you decide to update to version 4.30.x

    I think the real culprit may be your patch level of Solaris.

    Historically, since PICL is an OS daemon its viability is handled or "broken" by the OS you run.

    I can remember far back in an earlier incarnation helping customers with early Sunfire V240 systems that had nasty peculiar PICL errors until they patched their OS.

    Suggestion?  Log on to MOS, get the latest Solaris 10 full patch bundle and apply it to the most troublesome of those workstations.   It would need to be the full bundle because of inter-dependencies.  See what happens.

  • 1027031
    1027031 Member Posts: 2

    Many thanks for your input, I also believe its OS related. I am ex Sun/Oracle also

    I will keep this thread updated with my findings

    Cheers

This discussion has been closed.