Questions

Weekly Server Reboot Crash

+
0 Votes
Locked

Weekly Server Reboot Crash

daring
We have a mission-critical PowerEdge 2950 server with Windows Server 2003 running SQL2000 and Terminal Services. To keep it running optimally, we have a weekly Scheduled Task reboot. This has been working fine for months, but recently about 50% of the reboots result in a strange crash.

It seems like the OS is almost done shutting down but not completely. All services (HTTP, RDP, etc.) are shut down properly however the server still pings. (This may only be a result of the NIC still holding the IP?) The screen is completely black and it does not respond to any keyboard commands.

Luckily we have a DRAC installed so I am able to reboot it remotely. However, the 3AM wake up calls every other week are not preferred.

Anybody have any ideas on how to diagnose the issue, or force a power cycle upon failure? Thanks in advance.
  • +
    0 Votes
    netwrk_admn

    Are there any services not responding timely to stop requests?

    I would find which service is hanging on restart, and before the restart command (is it a batch script?) issue a net stop to the particular service.

    Sounds to me it's outside the OS level though, and a DELL tech should help you pinpoint if it's a hardware issue.

    I actually had the same issue with a Poweredge 1750 a couple years ago, I would restart it from home after WinUpdates and have to come in because it wouldn't come back up periodically. Black screen and all, but you could ping it. It was out of warranty but we had no need for it - we had already planned on buying a new server to replace it.

    +
    0 Votes
    daring

    Thanks for the response.

    From what I can tell, the OS is either completely shut down or almost completely. If it's almost completely, then all services including event viewer, server, workstation, etc. are already shut down.

    What I do know is that during the procedure (yes, a batch file), the server does not get to the point of POST.

    +
    0 Votes
    netwrk_admn

    I didn't go further on troubleshooting with my DELL server (warranty expired, etc), but I suggest, if you can, work with DELL.

    Also, before that route, they will want all the drivers updated, I would start there and see if that corrects the issue. Of course you may not know until the next weekly reboot...

  • +
    0 Votes
    netwrk_admn

    Are there any services not responding timely to stop requests?

    I would find which service is hanging on restart, and before the restart command (is it a batch script?) issue a net stop to the particular service.

    Sounds to me it's outside the OS level though, and a DELL tech should help you pinpoint if it's a hardware issue.

    I actually had the same issue with a Poweredge 1750 a couple years ago, I would restart it from home after WinUpdates and have to come in because it wouldn't come back up periodically. Black screen and all, but you could ping it. It was out of warranty but we had no need for it - we had already planned on buying a new server to replace it.

    +
    0 Votes
    daring

    Thanks for the response.

    From what I can tell, the OS is either completely shut down or almost completely. If it's almost completely, then all services including event viewer, server, workstation, etc. are already shut down.

    What I do know is that during the procedure (yes, a batch file), the server does not get to the point of POST.

    +
    0 Votes
    netwrk_admn

    I didn't go further on troubleshooting with my DELL server (warranty expired, etc), but I suggest, if you can, work with DELL.

    Also, before that route, they will want all the drivers updated, I would start there and see if that corrects the issue. Of course you may not know until the next weekly reboot...