Low deduplication ratio

[FelixZhou]

Looks we are getting low deduplication ratio without snapshots. it is about 1.3 to 1.5. We have 70% SAN and 30% of CIFS/NFS data. We are using AFF300 on 9.7P6, with cross volume deduplication and compression enabled. We also have aggregate level encryption enabled. Not sure if it is normal or not. Please share your experience and ways to improvements.

thanks.

[paul_stejskal · Accepted solution]

What is the workload? Some workloads just aren't as compressible/dedupable, especially if a high change rate. If you have things like tempDB, swap, or other files, it's better to keep in a separate FlexVol that doesn't get snapshots/dedupe.

More details will help us understand, but segmenting workloads into smaller volumes may help too.