Restarting Management agents on ESXi 4.1 / Risk rebooting guest virtual machines?

[WiEmWaerUser]

Unfortunately, we have to restart the management agents on our HP ML350 G6 nearly once a week (can't connect with VSphere client anymore, guests are running fine).

We know this KB-Article: Restarting the Management agents on an ESX or ESXi Server

and this helpful article:

How can we avoid that the ESXi 4.1 command /sbin/services.sh restart will reboot our running virtual machines (guests) as well?

(Our first try indicates: It doens't reboot the guests, but is this for sure or just fortunate coincidence?)

According to the second linked article:

... This will restart the agents that are installed in /etc/init.d/ and with a default install that includes: hostd, ntpd, sfcbd, sfcbd-watchdog, slpd and wsmand. It will aalso restart the VMware HA agent (AAM) if that has been enabled on the host...

Is any of this agents responsible for our running guest virtual machine?

Do we have to ensure "Automatic Startup/Shutdown" of every virtual machine too (as written in "Caution" on the first linked article)?

[marcelo_soares · Accepted solution]

I think this was corected on ESX 3.5 U3. Anyway, this only happened if you have autostart enabled for VMs - the capacity for ESX server to boot VMs when it boots itself. If you don't need it, leave disabled.

Is any of this agents responsible for our running guest virtual machine?

Not at all. VMs run in a very low level of the VMkernel. Even if you loose the COS (which is your main problem here - you should check it...) the VMs will still run.

Marcelo Soares

VMWare Certified Professional 310/410

Virtualization Tech Master

Globant Argentina

Consider awarding points for "helpful" and/or "correct" answers.

[GreatWhiteTec]

Agree. You need to make sure the setting on the host for Startup/Shutdown is disabled. This setting is disabled by default in 4.0 and 4.1. If you have to restart the service every week you may have a bigger issue there.

[PduPreez]

I Agree with the other posters

The Agent restarts will not affect the running VMs

I also agree to solve the root problem, you should not need to restart the agents on a weekly basis.

Do a health check on that ML by checking out the host status through ILO (can't remeber what is called now CIM agents I think)

Or move the VMs off and reinstall the host

I had a similar issue on an IBM blade. It wouls stay connected for 2 days then become unmanageable until agents are restarted.

Turned out to be a hardware issue, replace the blade and been running ever since

Regards

If you find this or any other answer useful please consider awarding points by marking the answer helpful or correct. Thank you.

[WiEmWaerUser]

Thank you all for your answers!

I will open a separate thread to find the reason for the disconnects.