Comments
2 comments
-
Alert data is the smallest catagory - so you will likely increase your database size proportionally (x52 times).
Half a terabyte of historical information is not going to be handled in a very performant manor, unfortunately. However, in an upcoming point release (new features yet unannounced) we may include configurable polling frequencies so you can reduce / turn off metrics you consider less relevant to your environment.
In the mean time, I would set the purge period to one month which although won't be enough for the reports you want should be enough to diagnose problems and spot unusual trends.
Regards, -
Ah OK.
I've only got 70GB left free on the data repository DB box anyway, so I'll probably be looking at 2 months max!
I'll set it at 1 month and look at exporting the data I want/need to keep.
Thanks.
Add comment
Please sign in to leave a comment.
The RedGateMonitor database is currently at 9GB, 74% of which is taken by table 'data.Cluster_SqlServer_SqlProcess_UnstableSamples' - obviously I have no idea what this table actually holds!!
If I increase the retention period to 1 year for 'Machine Data' and 'SQL Server Data' - but keep 'Alert Data' to 1 week - it is possible to state how likely this is to affect database sizes?
I am monitoring 3 servers/instances.