I am creating new VPG’s in Zerto (Version 6.0 Patch 2) and would like to know some best practices in doing so. More specifically in regards to replication and storage. Do you typically have a dedicated Datastore for the recovered VM/volumes? What is the typical journal history settings and do you use the advanced journal settings? Any information is greatly appreciated!
Hi Jennifer – I’m not sure about Zerto best practices, but I’ve found what works best for me, in my environment, and am happy to share with you.
I always use a dedicated datastore at the remote site, to store both the replicated VM’s and the Journal data. I have many datastores that serve this purpose, but they are always dedicate for Zerto data only, no running VM’s or anything else on these datastores. It just makes for a neat and clean setup.
I store my VM’s and the Journal data for that VM, on the same datastore. Here is an example:
Datastore01: VM “ABC” and “DEF” replicate to this storage. The journal for “ABC” and “DEF” is stored on this same datastore.
I also thick provision my recovery volumes. That way I’m never at risk of over committing via the Zerto settings.
Journal history is all about 2 things, how far you want to be able to recover to, and how much storage you are willing to use. There is a direct correlation. For me, typically every VM gets a 3 day journal by default. I then extend the journal further out, for those VM’s that require it. Most my VM’s are between 3-7 days. I have 1 VM where I do go to the maximum 30 days. You’ll also want to set journal limits (unlimited can only get you into trouble). This plays into the same equation as each VM will have a different IO rate per day. Generally, I start at 50GB journal limit by default, and that covers almost all my VM’s. I have a few chatty ones though where I’ve had to scale that limit up, in order to meet my 3 day retention.
A lot of this is really situation and environment specific. You’ll need to choose what is best for your situation.
Hope that helps.