I have a four node sql cluster (Server 2008 SP2) that gives 1069 and 1558 errors on the file share witness every fifteen minutes. I have been unable to resolve this issue. I tried changing from node majority/fsw to node majority/disk. However,
I got errors that the disk was not suitable for clustering even though there are 10 other hp eva luns presented without issue. I tried creating an empty resource with physical disk type and mapping it to no avail. I have read where a reform can
be done. However, this is a production sql cluster. There's no way that I can run all resources on one node or take an outage for that matter. I moved the file share witness resource from one server to another as well which didn't help. I
see there is a hotfix that speaks to this issue but is only for 2008 R2. http://support.microsoft.com/kb/2750820 Are there any cluster experts out there have a way to resolve this?
↧