VMware Cloud Community
proussel
Contributor
Contributor

Unexpected shutdown of a VM

Hello all,

I am not sure if it is the good place to post this, if it's not, sorry for that.

Since the last 3 days, I had an issue every nights: my ESX 3.5 shutdown at least one machine around 2am. I have no ideas of why! When I go to the ESX logs, I have nothing until:

2:00:20 User root@127.0.0.1 logged in

2:00:31: Ma_VM does not exist

2:00:31 User root@127.0.0.1 logged in

So I looked on this VM logs and I have:

2:00:31: Virtual Machine does not exist on 192.168.xx.xx

2:00:31: Alarm of Ma_VM change from yellow to Red

...Performing actions (like sending mail and automatically powering on)...

2:00:31 Task: power on Virtual machine

2:04:51 Host is connected

2:04:51 Ressource allocation changed

2:04:51 Registered with Virtual Center

2:04:51 Alarm of Ma_VM changed from gray to red

...same actions as before....

2:04:51 Task: power on Virtual machine

2:04:51 Migration from 192.168.xx.xx to 192.168.xx.xx (same IP, for my ESX) : Device "USB Controller" is not supported. This is a general limitation of the host

2:04:51 Virtual Machine is starting

2:04:56 Resource allocation changed

2:04:56 Virtual machine on 192.168.xx.xx is powered on

2:04:56 Alarm of Ma_VM changed form red to yellow

3:57:08 Alarms for CPU and memory changed from gray to green

And everything is going on... Also this removed my VM from its resource pool... I have no idea of why this happened, the windows (Ma_VM's os) log doesn't give me any usefull information

I have vRanger running backup on the ESX server, with the user root, but I don't think it's related since it also run at 11am without any impact... Also I have another ESX waiting in case of disaster and my vcenter is on another physical machine... (the same as vranger)

This issue is becoming critical so if anyonehad informations or ideas about that, I would be thanksfull forever Smiley Wink

Regards,

0 Kudos
2 Replies
ShanVMLand
Expert
Expert

As you mentioned that anyone VM shutdown at 2:00~AM, what were CPU/Memory/Network performance of the host?

Can you please confirm me that when vRanger backup kicks off either 11:00AM or 11:00PM?

Any information from /var/log/vmkernel and vmwarning

Thanks,

Shan

If you found this information useful, please consider awarding points for "Correct" or "Helpful". Thanks!!!
proussel
Contributor
Contributor

Hum... I think I found something thank to you, CPU and memory were high before that because of the back, then felt when the VM failed, same for the network... But I discovered that network become high around 3am, I wanted to restore backed up vm on my other ESX but I made a mistake and I did it on this ESX, it may be the cause of the fail... I'll see if this happen again...

anyway thank you Smiley Wink

0 Kudos