EPiServer Maintain Database Job Not Rebuilding CMS Indexes

Kevin Larsen

Vote:

Hello,

Sorry if this is in the wrong forum as I know the EPiServer Maintain Database job is a Commerce job, but the issue is occuring with the CMS database.

When running the job on a development site, I see that the job completes sucessfully, stating that it has Processed all indexes in the CMS database and Commerce database. When looking at the indexes for the Commerce database before and after, it does rebuild/reorganize the database indexes that were above 50% fragmented; however, when looking at the CMS database indexes there is no change in index fragmentation percentages.

I see no errors in the logs and the status is set to Succeeded in the job History. Is there something that I'm misunderstanding about the job and what it is supposed to do for the CMS database?

We are currently running EPiServer CMS 10.10.4 and Commerce 10.8.0 if for some reason there was an issue with the job in those versions that I was not able to find.

Thank you,

Kevin Larsen

#187423

Jan 22, 2018 21:42

Quan Mai

Vote:

Hi,

We are aware of this problem and it was fixed in a later version - 11.4 IIRC. We of course recommend you to upgrade if possible, we introduced a lot of new features and improvements over the versions.

If upgrading is not an option for you then one possible workaround is to take the SP and run it in the context of the CMS database directly. You can even write a scheduled job for that, but the effort might be as much as upgrading to 11.4 :)

#187432

Jan 23, 2018 0:27

Kevin Larsen

Vote:

Hi,

Thanks for the quick response! Are there any issues with still running the job to maintain the Commerce database until we can update to 11?

#187500

Jan 24, 2018 15:33

Quan Mai

Vote:

Hi,

There is another known issue that if you database is highly fragmented, then the SP might take longer than 30s to run, resulting in a timeout error. We fixed this issue in another version (11.2.1 IIRC) by introducing a setting to adjust the timeout value for calling that SP, but if you are using an older version, then the obvious workaround is to run the SP manually in SQL Server Management Studio which is not limited by the timeout to rebuild the most fragmented indexes. After that set up the job to run daily. Because it only needs to reorganize some indexes now, it should be fairly quick and complete within 30 seconds.

Of course I would recommend to upgrade as soon as possible :)

#187501

Jan 24, 2018 15:40

Michael Nylund

Vote:

Quan Mai - Regarding this job, what setting was introduced and how is it configured (non DXC setup)?

It's referenced here: https://world.episerver.com/documentation/Release-Notes/ReleaseNote/?releaseNoteId=COM-5226

#190529

Edited, Apr 11, 2018 16:38

Quan Mai

Vote:

I explained it here: https://vimvq1987.com/maintaining-your-indexes/

#190530

Apr 11, 2018 16:43

Michael Nylund

Vote:

Was just about to read it :) But you can only controll the threshold (with episerver::LowFragmentationThreshold and episerver::HighFragmentationThreshold)? How about the time-out?

#190534

Edited, Apr 11, 2018 16:57

Quan Mai

Vote:

Ah - the other one is episerver::DataBaseIndicesJobCommandTimeOut

#190539

Apr 11, 2018 17:12

Michael Nylund

Vote:

Thanks a lot!

#190540

Edited, Apr 11, 2018 17:23