vmwaretool in centos keep randomly crashing

[bellocarico]

Hi all,

I have two different centos 6.5 (x64) on two different hosts where I installed vmwaretool.

To cut a long story short occasionally I need to restart/halt the VM and sometime I find the options "shut down guest" "restart guest" greyed out.

When I go under the vSphere client VM tab the "VMware tool running status" says: no

If I reinstall the tools everything goes back to normal... until the next crash 😕

I have reinstalled the vmwaretools few times now but still occasionally experience this problem which is annoying...

Is anybody else experiencing this very same problem?

Thanks

[zXi_Gamer]

It does seem serious issue. VMware tools being crashed also brings down the performance of the VM. If you have valid support contract, I would suggest to contact tech support @VMware.

In meantime, you can figure out what caused the tools to crash by vmware.log or by increasing the troubleshooting level of the vmware tools logs by:

http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=100787...

If you can post the vmware.log when the tools crash happened, we can try to figure out what is happening in the system.

Thanks,

zXi

[MKguy]

What ESXi version/build are you using? WHat VMware Tools version? (/usr/bin/vmware-toolbox-cmd -v)

Are you perhaps having some OpenVM Tools components installed? (rpm -qa | grep -i vm)

Try using a more recent VMware Tools package like the one included in the latest 5.5 U1 patches:

http://kb.vmware.com/kb/2077405

Running newer VMware Tools versions on older hosts including Tools of newer major builds is fully supported by VMware:

VMware Product Interoperability Matrixes

[JarryG]

Well, this is not solution to your problem, but just workaround I use for similar cases:

I have a small tool called Monit which periodically checks all critical services and (re)starts them if they crash or stop responding. It does a lot of other usefull things too, i.e. checking of disk free space/inodes, cpu-load, memory usage, log-files, monitoring access to some files/directories, checking suspicious activities, and sending notifications (to log or email) if trigger is reached. Monit itself is another service daemon, but it can be also started from inittab so that it is auto-restarted whenever it dies.

Once in the past my ssh-server died and with no remote KVM-console I had to go to colo-center personally to check what was going on with my server. Since then I use this great tool. And I'm sure it could be used for monitoring of vmware-tools, and restart them if they crash...