Hi everyone, It it my first time to have discussion in the community. Nice to meet you guys.
I wish I could have awesome and share my experience.
a few weeks ago, I install esxi 5 on Cisco UCS C250 with local storage.(raid 5) there is Meragaid 9261-8i you might know where it is from.
then we added VM which were created on vSphere esxi 4.1 to the inventory. (Filesystem is VMFS5)
2 weeks after from installing, our custromer had problem to control esxi host via vSphere client so that he reboot the machine. after rebooting, esxi boot didn't come up and showed me that "vSphere kernal is broken : file system already exist"
I had to reinstall esxi hyporvisor. it worked well.
However, about 10 days after, similar problem occured again, I couldn't get any properties from vSphere client. (amount of memory, status of vm networking.. etc) but it was possible to access esxi host via vSphere client.
I checked event, there was log that "I/O latency increased" "lost access" "restore access" kind of those thing. I checked on Raid controller. there was critical errors. one of my colleagues told me that "this happes when disk i/o exceed threshold".
VM is Cisco CUCM. I don't believe that this VM requires a lot of I/O
one thing I didn't do was "upgrade vmware hardware version from 7 to 8" at first time I reinstalled esxi hyporvisor. next time I have done after reinstalling esxi hyporvisor. I suspected that not upgrading vmware hardware version was root cause, but not sure.
since 5 days passed, this is working well yet. but I am little bit worried about occuring incident once again..
Is there anyone who experienced this problem?, if it is solved, please let me know KB or your solution.
Thanks to everyone.
Byounghee Kim.