First of all thank you for your reply. I really appreciate this.
Yes, the HP switch is integrated in the blade chassis (virtual connect). We already called them and even they couldn't help. I am not getting any help from either HP support or VMware Tech support.
It has been more than 4 months now. I did a fresh install of ESXi 5.1 on these blade servers. However I used the same database which was being used in earlier version 4.1. I migrated the SQL database and upgraded it suring vcenter install. I installed vcenter fresh on a VM. Vcenter, SSO and DB are running on separate virtual servers.
Since we migrated to 5.1, initially it worked fine for couple of weeks and then started getting this alarm "host connection failure". However it was not affecting the VMs or may be it was but it used to last for a few seconds only.
Lately it started to become more frequent and now even hosts were not pinging each other. After removing LACP, hosts are pinging each other but this alarm comes every now and then.
About storage remaining connected or not? I would say i don't get enough time to check it as the issue is intermittent and last only few secons. However twice, the host remained in disconnected state for some time and when i checked it, Local disk of the esxi server was greyed out (disconnected state) but shared datastores were connected. This happened once. Second time, Only shared datastores were in disconnected state (greyed put). However both times, I was able to connect to the esxi hosts directly through VI client and host was working normally while as showing down in vcenter.
Kind Regards,
AG