ESX3.5 server showing "not connected" every morning

[njd62495]

I have an ESX 3.5.0 (82663) server the is always showing itself in a Not Connected state when I come in every morning. The host server is pingable as are the VM's that are on the server.

I can reconnect the server by restarting the hostd service (mgmt-vmware) but I cant figure out why it keeps disconnecting. I also checked /var/log and that folder is only utilizing about 10% of it's capacity.

None of my other servers disconnect...just this one.

We're not ready to upgrade to update 2 just yet, most likely that will be done in the next 2-3 weeks.

Any other ideas???

[jayolsen]

You could try removing the VC agents from the host and then adding back to VC which will reinstall the agents.

rpm -e VMware-vpxa-2.0.x.xxxxx(get the full packaged name by typing command "rpm -qa | grep -i vmware-vpxa)

Once complete add it back.

[njd62495]

Thanks...I forgot to post that I tried that already. It didnt update the package as far as version, it installed the same version, but I did remove it and re-add it and I'm still having the same issue...very confused.

[jayolsen]

Other things to check, which you probably have, IP, gateway, DNS settings are all correct. Can you ping by fqdn and verify there isn't a duplicate ip on the network.

[njd62495]

Thanks for the help Jay.

Everything checks out OK with DNS and all that good stuff. I checked it against a couple other host servers I have and everything looks good.

One other thing I forgot to mention is that I just upgrade the server to from 3.0 to ESX 3.5 update 1.

I've also done the update on 1 other host and it doesnt seem to have any problems...I'm trying to exhaust all effort before calling support for something that will most likely be fixed with one little oversight that I'm missing.

[malaysiavm]

You can also refer to this article

http://malaysiavm.com/blog/resolution-esx-hosts-unexpected-disconnected-from-virtual-center-esx-35-u...

Malaysia VMware Communities -

[mike_laspina]

Hello,

Since this is a daily event there is a high probability that one of the cron.daily jobs is related to the management agent issue.

have a look in the /etc/cron.daily and run the jobs one at a time to see if they are causing the issue.

[njd62495]

I'm still getting used to the CLI so how can I force a cron job to run?

[mike_laspina · Accepted solution]

login as root on the console or your user id over ssh then use su - to get root access

cd /etc/cron.daily

./yum.cron

./corewatch

etc ...

[njd62495]

Thanks.