Kaseya Community

High CPU usage by SQL service, Agent goes offline

  • Hi,

    I am experiencing this weird issue from last 3-4 Month, there is something triggers and cause sql service to shoots to 80-90% CPU usage and because of this all the agents goes offline this cycle continues for 1-2 days and then everything comes back to normal and it didn't occur atlease a month , till the kaseya support arrives everything stabilizes. Nothing helped so far- stopped kserver, reapplied schema as somewhere i found it might work, then I restarted server.

    Can somebody help me in this.

  • Is your VSA and SQL on the same server?

    I can't really say if 80-90% CPU is "high" because I don't know how the server is spec'd -- can you include some info about the specs on the server that is running your SQL database?

    Does the usage stay constant at 80-90% for 1-2 days straight?

  • have a look and see what scripts are running, we have a similar problem turns out a pile of scripts running...

    If your up to the task run a sql query to check if you can use SQL manager -

    Disclaimer - use at your own risk, this is only a select statement so will do no harm but still got to do the warning.  Change the date to the date range of your issue, remember scriptlog only stores for a set period.

    Run it after the issue is gone.

    It lists the scripts that were run in a count format, kaseya support gave this to me.

    select a.scriptname, left(a.description,50), count(a.agentguid)

    from scriptlog a, machnametab b

    where a.agentguid = b.agentguid

    and a.eventTime between '2011/10/01 00:00:00' and '2011/10/01 23:59:59'

    group by a.scriptname, left(a.description,50)

    order by count(a.agentguid) desc

  • 1.Yes they both are on same server

    2.80-90% Utilization  constant for 1-2 days this time its 3 days and continuing.

    3. Specs:

    Intel Xeon E5620  @ 2.40GHz

    Total Physical RAM:15.9 GB

    HDD 586GB

  • @martin :Ok I'll try. this

  • If you are into trying things...

    Try this.

    Go to the System Module and click on Statistics under the Server Management heading.

    This will show you a lot of Statistics about your system.  It should show you the Top Scripts running in the last hour.

    I've had problems with errant scripts flogging the processor under SQL.

    But more interestingly.  Try this...

    Point your mouse at the heading "Statistics collected at"

    You will notice your mouse pointer turns to a hand.

    Click on that and it will show you a whole stack of performance indicators that Kaseya use.

    I suggest you skip past the graphs and look at the "Current High CPU Utilization Queries"  This may also give you some idea of where things are going wrong.

    Kaseya have a heap of hidden areas in their system, and some can be very useful for troubleshooting.

    P.S.  Your system may be underspecced for what you want to do with it.  I have no idea on how many endpoints you wish to manage. Or what level of detail you want to collect.