Commvault definitely does not see ‘do not consume’ as the same as free space. You could do the same on a disk path on D:\ with 200 GB of free space but limit the usage to 100 GB - the actual capacity is still 200 GB but you’ve capped how much is usable. If you reach the cap, you could increase it by a few GB temporarily, knowing how much you have in reserve from the free space metric.

I see your point, though, and it would be good to set some sort of manual capacity value so it could be monitored in the software, but that option does not exist.

Hi @Kiwi_Dave 

Since this is a cloud library, generally it is considered infinite storage so Commvault does not try to calculate it. May I ask what library type you added during setup? If it’s one of the known on-prem S3 compatible types, I believe we still do the space calculations however if it’s generic, we treat as actual in cloud S3.

Thanks

Device is Cloudian Hyperstore (that is the API I am using)

Many cloud providers recommend that you limit the number of objects in your buckets. As this option is not available in CommVault, I have picked a bucket size as my limiter. The challenge I have is that I want to only add buckets as I am approaching my max limit so I do not spread data over too many buckets.

In this scenario my physical Library capacity is irrelevant, it is nearing the max capacity limit (number of buckets x max bucket size) that I’m interested in.

Hi @Kiwi_Dave 

What I meant was, during when you added library, did you select the option in Commvault as Cloudian or as S3 compatible?

@Kiwi_Dave yes, there are a few vendors in the market who recommend to limit the amount of objects in a bucket because of performance and stability reasons and one of them is Cloudian. in their latest release you can specify a workload during the creation of a bucket to optimize the way how it stores the data.

anyhow what you can do is monitor the bucket size and once you reach the soft limit you just freeze the bucket for new writes and add another one. by default most of them already recommend the creation of multiple buckets to spread the data across the buckets. 

please also check my post from a few days ago in regards to the "size on disk” value