Optimizing performance for cloud object storage

After you define a storage pool directory, the IBM Spectrum Protect server uses that directory as a temporary landing spot for the data that you are transferring to cloud object storage. The server uses an automated background process to transfer data from local storage in the directory to cloud object storage. You do not need to take any additional steps to start or manage this transfer process. After the server successfully moves the data from local storage to cloud object storage, the server deletes the data from the directory and releases space for more incoming data.

If storage pool directories contain no more free space, backup operations stop prematurely. To avoid this situation, you can allocate more storage pool directories. You can also wait for the data to be automatically removed from the local directories after the data moves to the cloud. The required number of storage pool directories you need to define depends on your disk configuration on the server. When the initial backups occur, the server spreads the data across all the directories you defined.

The amount of space you need for local storage is based on the amount of data you expect to back up each day after data deduplication and compression. If you have a stable network connection to the cloud object storage, the required amount of space is similar to the amount that is required for a daily backup.