The work took too long.
The vm was too slow after living failovers.
The vpg state was a propotion, and it took about eight to nine hours to complete.
Even considering that the storage is sata, it is incomprehensible.
At that time, latency (highest on VMware) was about 40 ms.
We’ve run into long promotion times when the source disk had a snapshot that had not been committed in sometime. The snapshot would need to be committed before the promotion could completed. However, in those cases, the operations would time out after a few hours and we would have to perform manually…. I do believe there is IO throttling enabled by default that is 40ms, sounds like you might have been hitting that threshold. It would be best to collect the logs right away to cover the failover period for Zerto support to analyze.
The source VM had not snapshot.
The VM is SQL server.