Replication requires resynchronization after every reboot

Question

I have 2 VMs (1 Windows, 1 Linux) set up to replicate between 2 2012 R2 hosts. These are 2012 R2 servers with the Hyper-V role installed. After every reboot/shutdown of the primary server, even if it's completely clean, causes replication to fail and require a "resynchronization." This is a 10 hour process during which my system grinds to a halt as my disks are thrashed around. Even if I pause replication prior to rebooting this happens. Looking at the logs I see the same event logged, Event ID 32526 "The log file associated with one of the VHDs of the virtual machine '<Host Name>' is corrupt. The EOL is not valid." Almost like the system didn't shutdown properly, but I know it did. I've turned off write caching on the disk to no avail. Does anyone have any idea?

Answer 1

Hi,
Normally Resynchronization is required when the following conditions are true:

1. You’re modifying the VHD file using any method other than the methods available with the Hyper-V Manager. In other words, the Hyper-V Replication engine must be aware of any changes being made to the VHD files included in the Hyper-V Replication so that the changes can be recorded in the HRL file.
2. HRL (Hyper-V Replication Log) accumulation is more than 50%.
3. There are issues writing to HRL files.
4. HRL files are not closing gracefully.

I tried in my Windows 2019 Server environment, but I didn’t run into the same problem, please check if the Hyper V host and the VMs are up to date, if not, try update them to the latest version.

Besides, may I know the disk type of the disk you use to store the vhd files and replication files? Is there any advanced settings?

For your reference:
https://learn.microsoft.com/en-us/virtualization/community/team-blog/2013/20130510-resynchronization-of-virtual-machines-in-hyper-v-replica

Thanks for your time.
Best regards,
Danny

-----------------------------

If the Answer is helpful, please click "Accept Answer" and upvote it.

Note: Please follow the steps in our documentation to enable e-mail notifications if you want to receive the related email notification for this thread.

Answer 2

Thanks for the input, there were a couple updates on the primary server so I ran those and rebooted. Now replication has completely failed, won't even start. At least before I could let it run a resync and be back in action. I'm getting "Operation not allowed for virtual machine 'BOINC PXE' because Hyper-V state is yet to be initialized from the virtual machine configuration." in the event logs.

I realize I could probably completely remove replication on these VMs and rebuild it specifying the existing remote data but I'm probably going to bail on replication for the time being. I've spent far more time maintaining this thing than getting value from it. Granted, I know I'm using an old release, so maybe I'll reconsider when I can upgrade to 2019. But the implementation in 2012R2 seems very flakey/beta quality at best.

Thanks again.