I am new to VMWare and new to Linux as well, so be gentle. I have upgraded to ESX 3.5 and VC 2.5. I am unable to get the VC connected to the host. I was having some trouble with this before the upgrade. When I try to connect, I get a message saying unable to load the vitrual center clent. I have troubleshoot and believe the problem is because the \var partition is full. I see a lot of logs in the log directory, but am not sure what can be deleted safely. Also, why has it filled up in the first place. So perhaps it is a config issue that is causing it? Should I be looking at any other directories within var as well for cleanup? This is not yet in production, but needs to be soon.
Can you post the result of the command vdf -h on your ESX host ?
This will show the usage of all your partition including the VMFS part.
Here is a screen capture bitmap.
can you try and post that again please
Tom Howarth
VMware Communities User Moderator
Filesystem Size Used Avail Use% Mounted on
/dev/sda2 5.0G 5.0G 0 100% /var/log
In fact your /var partition if full
So how do I go about cleaning it up? It looks to me like the log directory is what is full. How much of what is in a log directory can be deleted?
as a rule of thumb and directory ending in a number, these are logs that have rolled over.
Tom Howarth
VMware Communities User Moderator
Hi,
You can do a ls -lh on /var/log in order to view which logs are the biggest.
If they are old enough you can delete them, or move them elsewhere (backup in local via scp for example)
Hope this helps
I got this exact problem after upgrading VC to 2.5. First HA didn't start properly, and when I did "reconfigure for HA", I lost contact with the host server. Ping was ok, but ssh, http and VI-client couldn't connect. VMs were still running.
After some digging, I realized that the /var/log/vmware/aam/<servername>_agent.out was 1.9 GiB, and just repeating the same error messages about starting... error... restarting... over and over.
I moved this file to the home folder, and voila! Everything woke up again.
HA still doesn't work, though. My colleague says he had to delete the cluster, recreate it and re-add all hosts one by one. Anyone else had problems with HA?