Client-side dedup / identify duplicates

Arsenal

Newcomer
Joined
Dec 20, 2016
Messages
1
Reaction score
0
Points
0
PREDATAR Control23

Hello,

I would like ask here to avoid of creation new topic.
We use client-side deduplication.
If client nodes send their own deduplicated data whether need to run "IDENTIFY DUPLICATES" command for duplicated storage pool to identify duplicated data within storage pool ?
Assume each client sends its own deduplicated data(chunks) , but within storage pool that chunks might to match with other chunks ( from other clients ). Is it true ?

Thanks!
 
PREDATAR Control23

You got both server side dedup and client side dedup, and add compression to that on client side, and directory pools that dedup inline...and it gets complicated. Struggling myself on what to use. Learned that compression on client and dedup on server is not that efficient. I've had some problems with compression on clients so I've turned that of completely. Compression is more effective than client dedup when I checked on two new servers with only one system disk. Hard to say for servers with different content. Database backups that are compressed won't compress good by backup client. I'm doing client dedup right now and directory pool also dedups.

https://www.ibm.com/support/knowledgecenter/en/SSGSG7_7.1.6/client/c_dedup.html

https://www.ibm.com/support/knowledgecenter/en/SSGSG7_7.1.7/srv.solutions/c_tsm_backup.html
 
PREDATAR Control23

I'm working on this right now. I set up client side dedupe and compress AND server side dedupe for data, that cannot be deduped on the client like Oracle stuff. Right now, the only thing I found is that client dedupe data cannot be compressed on the server. So you have to decide who does the work.
 
Top