We ran into this issue when rebuilding some hosts. When trying to vMotion the guests to the newly rebuilt hosts, the vMotion would fail at 82% with the error ”Source detected that destination failed to resume.”
I came across the following article that gave me an idea: KB: 1006052
The article mentions the following:
Adding a host by these three different methods produces three different UUIDs:
- hostname
- hostname.domainname
- IP.address.of.host
It appears that the hostname or hostname.domain is also case sensitive. Mounting a datastore by “DATASTORE.domain.com” will generate a different UUID than “datastore.domain.com”.
Running esxcfg-nas –l on a host will show you the name of the NFS mounts. ls -l from the /vmfs/volumes directory will show you the UUID of the datastores.
Compare these two hosts:
HOST1
~ # esxcfg-nas -l
<datastorename1> is <path1> from <nas.domain.com> mounted
<datastorename2> is <path2> from <Nas.domain.com> mounted
<datastorename3> is <path3> from <Nas.domain.com> mounted
<datastorename4> is <path4> from <Nas.domain.com> mounted
<datastorename5> is <path5> from <Nas.domain.com> mounted
UUID: datastorename3 -> 3ba0562f-6ca841f9
HOST2
[root@tevp03187 root]# esxcfg-nas -l
<datastorename1> is <path1> from <nas.domain.com> mounted
<datastorename2> is <path2> from <nas.domain.com> mounted
<datastorename3> is <path3> from <nas.domain.com> mounted
<datastorename4> is <path4> from <nas.domain.com> mounted
<datastorename5> is <path5> from <nas.domain.com> mounted
UUID: datastorename3 -> 961816a8-7064b2f7
TLDR Version: Ensure that when mounting datastores you use the same case for the paths across all hosts.
{ 0 comments }
