Vcenter's shutdown by itself

[iMac001]

Hello everybody,

I have 3vm running in ESXi4 server, these days i face a stronge problem, the server shutdown by itself without any warning, refering to the log file of my server, nothing special except a warning message :

[msg.svgaUI.badLimits] Insufficient video RAM. The maximum resolution of the virtual machine will be limited to 1176x885 at 16 bits per pixel. To use the configured maximum resolution of 2360x1770 at 16 bits per pixel, increase the amount of video RAM allocated to this virtual machine by setting svga.vramSize="16708800" in the virtual machine's configuration file.

can the shutdown be related to this problem of Insufficient video RAM.

Any idea or suggestion it will be welcome.

Many thanks.

iMac

[Troy_Clavell]

Hello and welcome to the Forums.

The message you see should not cause a VM to go down.  Have you looked at the event logs of the OS to maybe get a better understanding?  Also, you have disabled all power options on the guest in question?

[iMac001]

Thank you troy for welcoming and also for you quick response,

Indeed I have checked the event logs of the OS of the three vms, nothing special, it tells that the system shutdown was unexpected, any other information.

About power options in the guest OS, I have the option "Always On".

[Troy_Clavell]

without having any logs in the OS itself, it may be a bit more difficult.  With that said, you can look a the vmware*.logs of the guest(s) in question to see what may be going on.

http://kb.vmware.com/kb/1019064

During this unexpected reboot, your hosts are fine?  There are no HA events that could have triggered the restart of the guest(s)?

[iMac001]

I would just to add a little clarification, it's not the guest who shutdowns but  the Vcenter Server.

[Troy_Clavell]

is the vCenter Server physical or virtual?  Is the database local or remote?  Are there any disk space issues with the Host OS or even your DB Server?  If the server is physical do you have any hardware diagnostics you can run against it?

[iMac001]

yes it's the physical server, and there's no problem with the disk space.

About the diagnostic, that's what I'm planning to do, otherwise it can't be a soft issue related to video RAM or other. Because all the LEDs of the different composante seems be OK.

[RParker]

Video RAM shouldn't be a problem for vCenter, and it has no effect if vCenter is a physical server.  If the service stops for no reason, I would say check the database, there could be a problem with the database (if it's Oracle maybe the database is too small or running out of space).

Where is the database located, on the same server as vCenter or remote?

[Troy_Clavell]

that warning message you posted above is related to your guests. It has nothing to do with vCenter, especially since your vCenter Host OS is physical.

[iMac001]

Please excuse me, but of what DB we're speaken here?

[Troy_Clavell]

vCenter DB.

[iMac001]

It's local DB, installed typically.

[AureusStone]

I need clarification.  You have said that your vCenter server shutdown by itself, but you are also saying your guests were shutdown unexpectively.

vCenter going down will have no effect on your guests.  It is for VM management only.

Is it possible your vSphere server\s went down?

Did your physical vCenter server go down at the same time as your vSphere server?  If that is the case then it would seem to be a power issue.  If your vSphere server went down, but vCenter stayed up then you need to check your hosts logs.  You can generate vSphere logs from within vCenter, but depending on your configuration it may have been wiped when the server went down.  Or you can grab them from syslog if you have one configured.

If vCenter went down, but vSphere stayed up look at your vCenter logs.  Either generate them from vCenter, or grab them from your log directory.

http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=102180...

The log file should tell you if the issue is with your database or something else.  If you can please attach a copy of the log file.  If you are uncertain and you have a support contract you should raise a ticket.

[iMac001]

Hello everybody and thanks a lot for your interesting to help me resolving this issue,

This issue was due to a heat problem of the processor's component rpm, and it was resolved by moving the power supply to the second position.

Thanks again for you all.

[Troy_Clavell]

maybe a good time to think about running vCenter as a VM.