Dear VMWare Gurus,
My current Virtual infrastructure is : 6 ESX servers (on Blade server, ESX Server 3.5 update 1 ), 1 Virtual Centre and 1 VCB server connected to tape drive.
I have 2 issues for your advise :
1. one of the ESX servers keeps disconnected intermittently. actually what we did is just manually disconnect and reconnect again. it will solve the problem. however, this case is somehow very disturbing. Network side is checked and nothing went wrong. This server also somehow need the HA to be reconfigured manually. It generates warning for HA randomly within 3-5 hours, so we also need to manually 'reconfigure HA' for this server.
2. the other ESX server (different machine from above case) suddenly have the /var partition full.The culprit is /var/core : there were many files with .core with filedate between 2 weeks ago to yesterday. The size of those files were 80 MB to 120 MB each. I believe this is the core dump file. The actions taken : 1) move those files to somewhere to free up /var partition, restart the vmware management service. Investigation on /var/vmware/hostd.log did not generate any errors. 2) the last attempt is to vmotion all VMs to other esx servers and reboot that esx host, then vmotion all VMs back. After 4 days of monitoring, the coredumps did not appear at all.One of the google search indicates that an ESX .xml configuration corruption can trigger coredumps. But this xml corruption did not appear in this ESX. My question is : why esx generate coredump files ? Is there any logs ( especially in /var/log or other places ) which can explain this anomaly ?
many thanks in advance.
