Hello Commvault Admin,
The usual configuration would look like this:
Based on the amount of data you plan to manage, you would need to choose the media agent/s (aka dedupe node). Media Agent versions available today are Small, Medium, Large and Extra Large. Hardware specs for a single dedupe node available here: https://documentation.commvault.com/commvault/v11/article?p=system_requirements/commcell_sizing/ma_standard_mode.htm.
You can also combine multiple media agents in a partitioned configuration for resiliency and increased dedupe pool capacity. This link has specs for 2 media agents in a partitioned configuration: https://documentation.commvault.com/commvault/v11/article?p=system_requirements/commcell_sizing/ma_dedup_partitioned_mode.htm.
In v11 we support max of 4 partitions, so with each media agent hosting one partition, you can combine 4 media agents.
From a CV software configuration perspective, you would then create a single Global Dedupe Policy (GDP) that basically ties up the media agents and disk library to a single config. Then create multiple Storage Policies (SP) based on the type of data here - one for File systems, another for Exchange, etc. The primary copies of these individual SP's will point to the common GDP.
Each of these individual SP copies can have independant retention based on your business requirements.
Advantage of a common GDP is that this creates a single common dedupe store which makes all of the data landing into this store dedupe amongst each other for maximum storage savings.
For tape out, follow a similar strategy. Create a Global Secondary copy for tape (to optimize tape use) and each of the individual SP's will now have a secondary copy that points to this common secondary tape copy. The secondary copies can and usually will have different retention than primary copies.
Set up clients to point to each of the individual SP's and then set backup and aux copy schedules.
This is a quick run down on the process. Please ask any questions.