Just a general question, how long can you leave a VPG in a non-commit state after live fail-over?. My guess is it all depends on how much space you have for the journal, but I wanted to check. We have a Server that has been acting up ever since it had windows patching done, it takes up to 24hrs for the the server to display the issue, I would like to fail it over prior to patching and leave it in a non-commit state to see if the issue was there prior to patching, if server is OK, I will proceed with the commit.
So is the amount of time you can leave it in a non-commit dependent upon how much space you have for the journal?
This is Vladimir M with Zerto.
Failover test or live process is not intended for long terms because the testing recovery or failed over machine is writing to a scratch volume which has the same length limitations as the journal and it’s not configurable, in order to take a long term test I would recommend an offsite clone procedure. For your convenience, I attached a link in which you’ll find the instructions on how to take an offsite clone: Page 304. http://s3.amazonaws.com/zertodownload_docs/6.5_Latest/Zerto%20Virtual%20Manager%20vSphere%20Administration%20Guide.pdf.
To get around the issue you are currently facing. I recommended stopping our failover test or live and changing the hard limit to “unlimited” and the warning size to a set GB before the datastore fills up. You will have to monitor the datastores to make sure the datastores do not fill up as this will cause more issues with the VPGs