DSC ESENT Database management
In a recent change, the new database for the on-prem Pullserver is now ESENT. Had a question for clarification.
As an old Exchange guy, I have fears that immediately pop into mind. The logs files do not appear to be using circular logging (a way of self cleaning and restricting the amount of log files) – Is this true?
If so, then I’m guess that the log files will continue to grow until the disk is full – which happens to exchange if the log files are not truncated. Is there a way today or in the future we will be able to manage the log files?
Thanks for the help understanding!
Brett Hamilton commented
It looks like this problem has been resolved in the WMF 5.1 update. Read here at section "Enable Circular log on ESENT database instance"
After some deeper debugging I can definitely confirm that the DSC Pullserver does not use the circular log option for the EDB Database. Therefore, the log files get never truncated. Furthermore, there is also no option for doing a correct full backup implemented.
For me this looks like a bug where the circular log setting is simply missing in the DLL of the DSCPullserver.
It is really import that we can fix this issue otherwise the DSC Pullserver is simply no option for use in production! See also Issue on Github: https://github.com/PowerShell/xPSDesiredStateConfiguration/issues/296
Can please someone from MFST investigate or help here?
This is a huge pain. We need really a solution for that. As Jason already wrote it seems that circular logging is not configured for the ESENT database. So in may case I have now over 46GB (!) of log files on my DSC Pullserver and the DSC service responds very slowly. So how can I commit and clean up the log files?
Thank for helping!
Ieuan Jenkins commented
I've yet to try it but there's a potential workaround/hack here that basically proxies the DSC Reporting database to a SQL Server DB via a .MDB file with an ODBC connection:
Luigi Grilli commented
I'm actually experiencing this problem and opened a ticket on github.
In my case the ESENT db of my pull server grows by 300MB a day with only 5 nodes configured. The suggestion that I received is that it's safe to delete the files. So I guess some scheduled task will do.
In my opinion there should be an option to limit the size of the db and after that it should start to roll logs.