We have a copy of our production oracle cluster environment (three node) running in our VMware cluster. We use this for backup purposes. We run into this weird error once in a while:
Servername kernel: I/O error: dev 08:21, sector 45
Servername kernel:scsi1 (0,0,0) : RESERVATION CONFLICT
Servername kernel: I/O error: SCSI disk error : host 1 channel 0 id 0 lun 0 return code = 18
Then sometimes we get:
Servername kernel: ocfs_rw_direct : brw_kiovec() -5
Servername kernel: I/O error: need to fail out
THis causes the OCMSTART service to crash - taking down all the databases with it.
There doesn't seem to be a pattern at all - it worked just fine for weeks and then this happened twice this week. I looked in this form and found this:
http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=103416...
We do not have that turned on currently does anyone have any experience with that? I have seen where people suggest that you move both database servers to the same host to help with the heart beat, also seen were one person suggested changing the disks to "independent" mode.
Before I start taking shots in the dark I was hoping someone might have an idea which direction to start.
Thanks in advance for any ideas.
