Journal Sizing

The RPA is GEN4 with v 3.4.

I do not follow what you said regarding 30% of replica.  Can you elaborate?

I see that you say x3 throughput rate. So if your app was 100MB/s you would need a journal of 300MB.

Just posing the question since the app is not online yet and we are trying to take an initial stab at it.

Hi there,

A common question but realistically there is no such thing as a standard or 'rule of thumb'. There is a specific formula for sizing the journal which is detailed in the RP Admin Guide but this depends on three key factors.

Firstly, the throughput rate in M. Secondly, the protection window (rollback history) and thirdly the number of drives to sustain the distribution rate (x 3 throughput rate in MB/s).

My advice would be to work off a 30% of replica and state these caveats remembering that journal volumes can be added to a CG journal dynamically meaning that that a CG journal can be made larger to compensate for a higher throughut rate, a greater protection window requirement or faster distribution by utilising more disk without disrupting replication or the protection window.

Regards,

Richard

Rule of thumb - 20%

30% of the total amount of disk space allocation to the disk space for the target replicas.

Can you lay it out with an example so I make sure I understand?

Thank you.

Sure.

You have to replicate 100TB at Production which means you have to have 100TB at the target (as replica), therefore journal would be 30TB.