memory leak issue on exchange catalog

[federiken]

Hi. We currently own BE2010 but are testing BE2016 for exchange. our current environment is

-vmware 5.5 server

-BE2016FU2 running on a windows 2016 VM (also tried on 2012r2) (in spanish)

-echange 2016 CU5 running on a different VM (spanish OS)

-1800 mailboxes, mailstore about 450GB 

-full exchange backup, catalog runs as a separate job, on local disk (no tape)

remsrv64.exe slowly takes all the memory, until it crashes, 24hs later. looks like this one: https://www.veritas.com/support/en_US/article.000127449

job monitor shows strange characters in the progress window (attached), and it takes about 24hs to catalog 1/3rd of the database. I've tried different OS (windows 2012r2, windows 2016), different memory (8gb, 24gb), splitting mailstore database, and it always slows down and fails on the catalog stage (backup runs fine). I have not updated echange 2016 to CU6 yet, buy I don't believe that's the issue. Also tried with BE2016 FU1 and FU2

Any ideas? maybe it's something related to languaje settings?

thanks a lot

[Colin_Weaver]

Please log a formal support case as Backup Exec 16 FP2 (with an additional patch that you have to log a support case to get) should have had a positive effect on your condition

Also even though the article you quoted does not state that we may have a fix against FP2 plus a controlled patch), the bottom of the article does ask you to log a support case if you think you have that condition.

NOTE: this entry has been edited after I discovered more details on this issue via internal sources and it is likely we will eventually make the fix publicly available in due course.