cancel
Showing results for 
Search instead for 
Did you mean: 

Dedupe broken and Veritas official response is "we cant fix it, sorry"?

JohnSit
Level 3

I have a support call logged with Veritas.

3 volumes on a Netapp filer, 1 dedupes perfectly with a typical ratio of 19:1.

2 volumes have no dedupe at all yet I have solid proof in screenshots of it working on the two volumes, two months ago (yes, it has taken 2 months to get this answer from Veritas!!) with the ratio of 95% (475:1).

So the Veritas engineer firstly said that all my data was changing on a weekly basis (poppycock!, anlysis shows over 70% is over 3 years old!).

I explained this, then his second statement was "in that case, if it is working on one volume but not another then we have no way of telling why its gone wrong".

So basically, they are saying its broken and we have no way of fixing it. REALLY? Since the two volumes house 90% of all of our data is the only recourse to get a refund and move to software that can be debugged and actually perform a deduplicated disk based backup?

Such terrible support...

2 REPLIES 2

Gurvinder
Moderator
Moderator
Employee Accredited Certified

do you have an earlier job log showing a better dedupe ratio ? when did the dedupe ratio start to lower for the 2 netapp volume ? can you share the latest Job log which shows less dedupe ratio. When you backup these 2 Netapp volumes, do you get any errors in the event viewer on BE server. can you also provide the support case number.

Thanks for the response!

The problems started occuring when an engineer advised us to change a registry key on the server as part of case 180919-000754 (totally unrelated to this). Since then, even after a full rebuild, the volumes have not deduped. Obviously the registry key has been reverted.

We get no errors at all when the job completes, all successful.

Attached is the log showing virtually no dedupe.

Here is an image which shows the exact same volume getting a 95% dedupe ratio last year.

ScreenshotScreenshot

The current case is 181121-000843

Thanks again.