There has been a couple of posts on this and seems most people just say reboot SRM related servers but it has not helped in my case.
SRM 5.0.1 on both sides, Win2K8R2 guest, using vSphere Replication on both sides. The inital replication completes but it starts showing as Not Active afterwards and it will try to sync on it's own but fail after a few seconds.
Manually starting it up shows a popup that says, "Call "HmsGroup.OnlineSync" for object "GID-removed" on Server "ip removed" failed.
vCenter client shows, "vSphere Replication operation error: Virtual machine is in an invalid state."
So far VMware support has not found anything wrong and the guest has had it's VMtools uninstalled and reinstalled. I have rebooted VRMS servers and VR servers on both sides with no results. I've also checked the logs on the VR server to see if there was any login issues but there were none as 4 other VM guests are replicating fine.
Anyone else out there solve this issue that was not from rebooting the appliances?
Well after trying everything with multiple escalation engineers it turns out our VMDK was too big. Reboot, upgrade esxi, firmware, drivers, etc did not help. our VMDKs were 2TB-512B or slightly more, but due to overhead involved with snapshots used by SRM the max size should of been 2032GB in size. Engineer confirm it wasn't in any documentation for SRM and they will work on getting it added, I felt it should of been in big bold letters somewhere stating this requirement.
The snapshot KB article has since been updated to reflect the 2032GB max size at the bottom.