Data de-duplication on disk storage systems is an increasingly available and attractive option. Most data de-dupe systems allow backups to take place like normal but will modify the storage of the backup files to take repeating patterns in the file and de-duplicate them.
For many reasons, most data de-dupe vendors recommend that you not use data compression in conjunction with these systems. While the use of SAP Sybase Adaptive Server Enterprise (ASE) backup compression is not known to break de-duplication systems, it will make the de-duplication process less effective. Why is this?
ASE compression is very similar to the compression employed by the data de-dupe systems, so the data de-duplication won’t actually happen because ASE has already done it. This doesn’t hurt the backup, but it can have the effect of the compressed backups taking more overall space than a non-compressed backup.
When considering whether to use ASE backup compression or disk de-dupe, you must balance the needs of time and space. You likely have limits on the amount of space available on the backup disk, just as you have limits on the maintenance window you’re using for backups.
Many compressed backups will take less time to complete than a non-compressed backup. This is because it typically takes less time to write a smaller compressed file than to write a larger, non-compressed file. So does it work better in your environment to send an uncompressed backup to the de-duplicating backup system and have it compressed there? Or does it make more sense to use ASE compression?
If you’re adding disk de-dupe capabilities to your environment, you must evaluate how to use it effectively with ASE backup compression. The answer, as always, lies in the specific needs of your environment.