For object stores that have per bucket performance limitations CloudBoost does have the ability to do sharding across multiple buckets but there is no way to manually set the number of buckets used for this purpose.

Do you mean CloudBoost will create multiple buckets? If so, how many buckets CloudBoost could create?

It is not something that you need to know about as an administrator. It is just one of the many things the engineers need to deal with when using different object stores. CloudBoost will create however many are needed to hold the data on the object storage being used. Unless you are using an object storage provider that has per bucket limits CloudBoost will only create the two buckets, one for the deduped chunks, and one for the automatic backups it does of itself for DR recoveries.

For example Atmos has a limit of 100,000 blobs per directory (similar to a bucket) and OpenStack Swift starts slowing down at about 1 million blobs per container. The public object stores on the other hand do this load-balancing at a lower layer than the bucket level so CloudBoost does not have to do it itself.

Thanks a lot for your reply. One more question, will CloudBoost create multiple buckets when storing a big object or storing multiple objects simultaneously?

If the object storage provider with S3 API has per bucket limit, CloudBoost will create the third bucket or not?

CloudBoost automatically creates whatever buckets it needs. Although I should add none of the supported object stores that use an S3 API would require it.

So do you mean that CB would create another bucket if the bucket has reached the limit of objects when the object storage is using S3 API?

No I do not mean that. There are no object stores that use the S3 API that have a limit.

If the cloud storage which is using S3 API has a limit of object per bucket and objects has reached the limit, how does CB deal with the situation?