VMware Cloud Community
rodah
Contributor
Contributor

Error when resetting guest

Hi

Today I tried to reset a guest that had freezed totally in Virtual Center. It freezed during filecopy. When I did the reset, the resetprocess hung at 95 % for almost an hour. I couldnt do anything with the server in virtual center so I stopped the guestprocess on the host locally using the kill command.

The guest had several snapshots and after I used the kill-command the disk-settings on the guest was all screwed up. It wanted to use a file named Servername-000002.vmdk as disk. But it couldnt start with this file as it complained about swap-files missing. I changed back the vmdk-file to Servername.vmdk on the guest instead and everything seemed to work.

I couldnt revert to one of the snapshots but another one worked. I reverted to that and deleted all the other snapshots. After that the guest complained about corrupt files and I ran chkdsk. After the chkdsk the server started bluescreening. Probably because one of the files that was corrupt was a systemfile.

Anyone know why this happend from the start. Why did the guest hang and why was it not possible to do a reset? Why did the filesystem get corrupt?

Best regards,

Robert

Reply
0 Kudos
4 Replies
danam
Contributor
Contributor

See the thread I just reposted. Maybe the solution is there. In a nutshell, the crashed machine was still running somewhere, one of the cluster hosts. After all processes were killed I could restart the VM

Reply
0 Kudos
danam
Contributor
Contributor

Reply
0 Kudos
rodah
Contributor
Contributor

Thanks for the thread, but I dont think this is the same error.. Yesterday I restored the VM from backup and everything ran smoothly until this morning when it hanged again. This time I looked in the vmware.log -file and saw these errors:

Jun 13 10:32:53.824: mks| SOCKET 11 recv error 5: Input/output error

Jun 13 10:32:53.824: mks| SOCKET 11 destroying VNC backend on socket error: 5

These correspond with the time the server hanged.

Ive read some forumthreads about this and it seems that vmware doesnt have a solution to the problem other than killing the process for the guest.

One thread stated that this was more common on systems running w2003 and SQL2005 or IIS. This system we having problem with run all these programs. And the last time the problem occured we were restarting the IIS-service.

One thread talked about mounted ISO-images this server has this also, could this cause problems?

Anyway, when I now tried to shutdown the server it stops at 95%, this log entry is seen in vmware.log:

Jun 13 10:33:05.381: vmx| TOOLS sending 'OS_Halt' (1) state change request

Jun 13 10:35:35.376: vmx| Stopping VCPU threads...

When I killed the process yesterday, I couldnt startup the server. It complained about a swapfile missing. Is it possible to recreate it somehow. I suspect the same problem will occur if I kill the process today too..

Two snapshots is present on the machine, and I dont want to loose them..

Is it any use opening a supportcase with vmware?

Reply
0 Kudos
rodah
Contributor
Contributor

I tried kill command now and in virtual center it nicely stated that the process of shutting down was completed on the server. It didnt do that yesterday.. Then it just was a matter of starting up the server again. No swap file missing messages this time..

The differences from yesterdays shutdown was that I unmounted the iso-file before the kill command, and that I used shutdown instead..

Could this be the reason why the server didnt broke during restart?

Reply
0 Kudos