06-30-2017 06:07 AM
Just as an update on this for anyone interested. I've finally got to the bottom of the disk space usage - which I think someone did point me at during my upgrade but for reasons I'll explain in a minute it didn't resolve the issues straight away.
Basically the disk was being used by the Performance data collection - which, currently, we don't use as we have other tools that do this. So I've disabled all the historical data collection and run the clear-performance-data script to delete all the old data.
Now, the first time I ran this - during my migration - it cleared out 10GB, not a lot - but it did error. At the time I didn't think much of it and just assumed it had worked on the whole.
I have since run some scripts on the SQL through PGAdmin to report on the table sizes and all the main culprits were the 'time series data' tables - which if I'm honest I didn't know what they were for.
anyway, in a last ditch effort I ran the clear-performance-data script again this morning and it cleared out 80GB (and errored again), bringing my DB down to 100GB. However, looking at the table sizes again there were still a lot of 'time series data' tables with a lot of info in them. So I've run the script again and now my DB is 4.25GB! much more managable!!
So now I can increase my event rentention and once I've got my head around what - if any - performance data we require we can switch it on and tune it to what we need.
mystery solved, in the end by a lot of poking around and a bit of guess work.