Take a look at the release notes for the latest release

I think there was an issue that Space Saved wasn’t displayed until the scan was completed.

Whenever we look at the file system propertires, it says:

Status:  Scanning (95% complete)

This has been the case for at least two days, now.  Here is the latest view:

Status: Scanning (95% complete) 

Statistics: As of last file system scan (Saturday, November 5, 2011 8:55:50 AM EDT)

Files scanned: 794,812

Files deduped: 218,281 (62.0% of total files)

Original data size: 569,828 MB (6% of current file system capacity)

Space saved: 0 MB (0% of original data size)

... does the scan ever complete?

check the server_log for error messages regarding dedup.

Often seen, Savvol NoSpace ....

These all came up empty..

server_log server_2 | grep NoSpace

server_log server_2 | grep Savvol

server_log server_2 | grep error

server_log server_2 | grep dedupe

How can we know whether deduplication is working?

Thanks...

Open a service request

If you arent running a recent software release then I would suggest to upgrade

Access Time: 0 Modification Time: 0

I would also suggest to start with the default values there.

Once that works and you understand what they mean you can always change them later.

IMHO setting them to 0 doesnt make sense and causes more work than what you gain in additional space savings.

See the deduplication white paper on what it means where a deduplicated file gets changed and think whether it makes sense to deduplicate files that might be temp. anyway.

Rainer_EMC:  "Open a service request"

I came here after having opened a service request.  The service rep said that nothing would deduplicate until after 15 days.  He couldn't explain his assertion in light of the parameters that were set for Access Time, Modification Time, and Minimum Scan Interval.

I won't rant here about EMC support, but if he was right, at least he could have explained what he meant, or how my understanding was faulty..

Rainer_EMC: "I would also suggest to start with the default values there.   Once that works and you understand what they mean you can always change them later.   IMHO setting them to 0 doesnt make sense and causes more work than what you gain in additional space savings."

Here is my logic.  We have older checkpoints that we would like to migrate off of our production array.  We have created a file system on our secondary Celerra for this data.  If I copy several of the monthly checkpoints to our secondary Celerra without dedupliction, they would take a considerable amount of storage space.  As they should contain a very large percentage of identical files, they should deduplicate well.   As of today, I have copied seven of the checkpoints to the secondary Celerra. 

These files will remain static.  I would like to not wait 15 days for dedupliction, as we want to copy more checkpoints to the file system. 

If my settings are incorrect, what are the proper settings so that the dedupliction process starts immediately? 

Thanks for your assistance.

That depends on whether your method of copying the data preserves the atime / mtime

I would first wait for a successful deduplication run with the default params and then decrease them.

By getting less file deduped in the first run you lower the chance to dedupe stopping because file system or savvol full.

If you know exactly that you want a complete directory deduped (which really means compressed and single instanced) you can just as well use a Windows client and set the compressed attribute – that doesn’t use the scheduled scan and the atime/mtime options.

Rainer

I would insist on having the log files analyzed by support

Deduplication doesn’t just stop without leaving a statement in the logs.

If you are not on the latest release I would strongly suggest to upgrade – no sense in chasing problems that have already been fixed.