VMware Cloud Community
andromeda7383
Contributor
Contributor

Hosts disconnect from the VC frequently

Hi!

I have an ESX server 3.5 and virtual centre 2.5 patched with the latest update 3. The patching took place a couple of weeks ago.

Since today morning, while I am working on the VC - suddenly the connection to the host lost and the status says "not rsponding". The status of all the VM's is "disconnected".

I checked the vpxa on the esx host and tried to restart it - but I get error message saying - its already running.

So, I right click on th esx host and still find the option to disconnect - after disconnecting - I reconnect it back and it works normal for a while only to get disconnected again later.

I doubt its the issue with the "Update 3" as I did not face this immediately after applying it.

Any other thoughts would be appreciated.

Thank you

Hugo

0 Kudos
5 Replies
Rob_Bohmann1
Expert
Expert

I would next try and disconnect the server again, then remove the server from virtual center(vm's will remain running). Wait a few minutes, then add it back into virtual center.

Also check your service console network connections, and the logs in Virtual Center and on your host for more info.

/var/log/vmkernel

/var/log/vmware/hostd.logs' to start.

0 Kudos
Karunakar
Hot Shot
Hot Shot

This problem might occur when the VPX agent running in the ESX is not handling the connection properly with VC.

Can you try to restart the mgmt-vmware service in the ESX, and check if this is the same.

-Karunakar

0 Kudos
andromeda7383
Contributor
Contributor

Thank you guys for the replies. Iam monitoring the VC and the host. Will give further feedback soon.

Appreciate your help.

Hugo

0 Kudos
mtidgewell
Contributor
Contributor

Don't be so sure it isn't u3. I'm having this and many other strange communication problems on 3 VC 2.5 U3 servers. I'm hoping by bumping your thread up more people with this problem will pipe in. Is it posible that there are some new ports used for communication to VC? I have an ASA between the hosts and VCs. Two of these VCs were upgrades that did not have any of these problems before the upgrade. The third one was a new installation, but I opened the same ports in the ASA.

0 Kudos
KristianW
Contributor
Contributor

I am seeing the same thing with some esx-servers. However I am not using U3 yet. Restarting the VPXA-service (and/or the mgmt-vmware service) usually helps, but this is not a good "solution".

I have tried reinstaling the servers. I have tried breaking the DRS cluster. I have checked the database for inconsistency. I have checked the network. I have reinstalled VC (!!). The only thing I haven't done is purge the database

I am also seeing alot of strange messages in vpxa.log (see attachment)

VC 2.5 U3

ESX 3.5 U2

Regards,

Kristian

0 Kudos