Deduplication scalling

  • 10 August 2021
  • 1 reply

Userlevel 1
Badge +7


We are in CV deployement and Initially we build a single MA with 4 partitioned DDB in the Azure cloud. When data is growing, we moved the two partition into new Media Agent and its running in two MA with two DDB disk each.

Now, both MA reached its bottleneck planning to scalling up further, but management allowed me to add one MA alone.

So, I have one possiblity of running the backup with four-partitioned DDB splited with three MAs as shown below.

  1. MA 1 - Single DDB Disk
  2. MA 2 - Single  DDB Disk
  3. MA3 - Two DDB Disk

I am bit worry about to do that, as I am thinking it make some instablity between MAs. But, I couldn’t find relevent CV documents. 


Can you suggest whether the above design make any sense or not? and will it make any issue in the future?





Best answer by Damian Andre 11 August 2021, 00:13

View original

1 reply

Userlevel 7
Badge +23

In theory this could work, but the media agent with two disks would be handling double the load, so would need to be configured to manage that amount of data. You also need to ensure you don’t exceed the maximum data size per the deduplication building block guide. 

likewise, one drawback is that this one media agent will create a much larger impact if it suffers and outage or requires maintenance.


One other point is that partitions are fluid - you can migrate it to another media agent down the road - so this could be a good interim solution for you.