We've been running Kaseya for coming on 8 years now. Have often made adjustments for performance but lately had to strip things down like not collecting event logs on agents which isn't what I like doing.
We have 2 servers, one running the Kaseya Apps which never has high CPU and the other running SQL under Server 2008, SQL 2008, 12 GB RAM, Xeon processor. CPU often pegs at 100% and stays there. Database size is about 8 GB.
A couple of things we do know, KAM sucks SQL dry. While we have KAM service running, some of the associated SQL queries are constantly at the top of the usage list. Don't know why as we only rolled out KAM to about 50 machines.
Following are some of the table contents from System -> Statistics and the two lnks on those pages. Wondering if some people can compare.
Current High CPU Utilization Queries (Past 5 Minutes) | ||||
Query Text | Last Execution Time | Last CPU Time (ms) | Last Logical Reads | Last Logical Writes |
select top 100 dateSubmitted, conduit, plugin, class, executionTimeInSeconds from hermes.vlog where conduit is not null and errorMessage is null and dateSubmitted > dateAdd(d,-30,getdate()) order by executionTimeInSeconds desc | 1:27:34 am 30-Jan-13 | 1,367,078 | 28,158 | 2 |
We often see queries wiht a CPU time of 1,xxx,xxx ms. Something has to be wrong here, either the table info is wrong or we have some off things going on as a query should not be taking 1400 seconds or 23 minutes to run. I'm wondering if other people are seeing queries that are taking this long.
The following should be even more easy for comparison. Not concerned with the items that run once per day, It's more of a question of the ones at the bottom of the table that are running every 1 to 2 minutes. It may be correct, but things like KES Periodic Batch Processing that runs every 2 minutes seems a little more then Periodic.