raffie613
asked on
Windows 2008 server unexpected shutdowns
We have a 2008 server that is shutting down and I have no idea why. It seems to happen Friday-Sunday and in middle of the night. I have a ups backup battery attached with the parachute program and it has no data as ever switching over to the battery for power. When I check the event viewer logs, there is absolutely nothing. Seems like an application log from MSSQL that the backup is running during that time but that is because it run every single day in middle of the night. It never shuts down during the other times it runs through. I can only see from the event logs when the server gets turned back on and the recorded event ID 6008 that " the previous system shutdown was unexpected t such and such time." I have no other Event logs showing anything happening at all during that time. How can I determine what is happening here. The server has to be manually turned back on when this happens.
Hi,
if the server is set to turn itself back on after power loss, yet it's requiring a manual start after this issue occurs, then that means it probably is a heat issue. If it's possible one thing you can try is taking the side or top panel off the server case and blowing a fan on the main board to see if the issue stops.
- gurutc
if the server is set to turn itself back on after power loss, yet it's requiring a manual start after this issue occurs, then that means it probably is a heat issue. If it's possible one thing you can try is taking the side or top panel off the server case and blowing a fan on the main board to see if the issue stops.
- gurutc
Whenever this happens next, immediately try to turn it on. Go in the BIOS where you get Hardware Monitoring tab which shows CPU Temperature which will give you some idea about the heating. If it is really heating then check if any BIOS settings been done to turn off the Server if CPU or Panel heat reach particular temperature.
Or above option to use Fan after opening up the Front or Top Panel to cool down the server and see if it makes any difference.
Or above option to use Fan after opening up the Front or Top Panel to cool down the server and see if it makes any difference.
ASKER
If the issue is the Heat, then why does it always seem to happen over the weekend when no one is there accessing or using it all? Wouldn't it be more prone to shutting down in middle of the day when CPU usage would be much higher?
That is a a good question, but first see if that is the real case.
Possibility could be as below (just getting these in my mind)
-- Could be any job running that time which putting more pressure on the box. Like Backup or any other Job if this is SQL one.
-- Cooling system (AC or Ventilation Machine in Server Room) is not working or someone may be turning it off by chance due to weekend.
Possibility could be as below (just getting these in my mind)
-- Could be any job running that time which putting more pressure on the box. Like Backup or any other Job if this is SQL one.
-- Cooling system (AC or Ventilation Machine in Server Room) is not working or someone may be turning it off by chance due to weekend.
ASKER
I know the AC unit is staying on the entire time. We have other servers in the room not having this issue? Could it be a firmware issue?
We also have SQL backup running every day and it doesn't turn off most of the time.
We also have SQL backup running every day and it doesn't turn off most of the time.
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
It is unlikely to be a firmware issue if it's a recent(ish) problem. AND just because you have other servers in the room, doesn't mean that there's not a hardware issue. I really would take a look for any raised capacitors - it's not an uncommon problem. Make sure the server is clean (internally) - carefully vacuum it, or if necessary, clean the fans / heatsinks. Lots of computers (servers included) will shut down if they overheat....
Glad you're working.
- gurutc
- gurutc
ASKER
this worked. I was right.
How old is the server? I would suggest (if you can arrange the downtime) to check the heatsinks / fans for dust and dirt - cleaning them if necessary, and perhaps replacing the thermal paste between the heatsink and CPU.
Also check for any blown (raised) capacitors on the motherboard just in case it is a hardware fault.
Also, you could perhaps run diagnostics on the memory just in case it's a RAM issue.