Been experienceing some "performance issues"  with the K1.  I have gone in and deleted unused files and the "usual" tweaks.  Currently agent task through put is sitting around .1 to .7.  I have seen it as high as 2.2.  I have the agent interval set at 2hrs with 5 tasks.  When looking on the dashboard, the tasks in progress shows Krash upload around 95%.  I would welcome any advice and or links. I am running 5.4 and have about 3000+ PC's.


0 Comments   [ + ] Show Comments


Please log in to comment

Community Chosen Answer


I actually would increase your 2 hour to at least 3 or 4. How many machiens do you have checking in per minute. You shouldn't have more then 7 or 8. 

What is your scritping update interval. Is it still at one hour? if so you have up to 3000 machiens checking in to see if you have updated any scripts and download the updates. If you don't update scripts that often, I would start my chaning it to a 1 day check and go from there. 

Answered 01/28/2013 by: nshah
Red Belt

  • OK so here are my current settings and it seems to be running pretty good after some tweakage.
    Agent Run = 4
    Inv interval = 0
    Scripting = 12hrs
    Task throughput = 7
    current load avg is .16
    Agent logs retained.
    Looking at all the associated graphs leads me to believe all is well.
Please log in to comment



The Konductor section of this slideshow helped me get started.

I think two hour check ins would be too much for almost anyone. 

Answered 01/28/2013 by: dugullett
Red Belt

  • Also check your scripts. I had to go back through an clean up all my scheduled scripts that were no longer needed, but still running.
    • That document was exactly what I was looking for. It helped clear most everything up. This kind of info is extremely valuable. Thanx
      • There's an updated one that was shown at this years conference. It explained in more detail. I'm not sure if those are available yet. I haven't really looked.
    • I remember hearing about it. But can't remember where. It was a great konference.
      • It was the advanced topics class. I got copies of a lot of the powerpoints. Just couldn't find that one. It when into some more detail. Very useful.
    • I wasn't able to get into that one. Aren't those online somewhere? If not are you willing to share?
    • Found it.
      • Good. I've been too busy to look. I knew they would post them eventually. That class was standing room only. That powerpoint explains more than the first one I sent. Still both good.
  • By still running do you mean enabled?
    • Yes. I had a few quick fixes that were on schedules to make sure they got out. I have to remember to go back in and disable them. It sounds simple, but it'll take up your tasks.
      • Thanx, I'll add that to my todo list for tomorrow.
  • The 2 hr was recommended in training as well as from the on-site engineer. I am looking at that article now.
Please log in to comment

Just to add some info, we have 25,000+ clients and our Load Average is currently 2.65 with a 2 day interval.  2 hours is a very short interval but for the size of your environment I doubt that is your issue.  Not sure if this helps you, but I just wanted to chime in that a 2+ Load Average isn't bad.

Answered 01/28/2013 by: nheyne
Red Belt

  • I had the task throughput set to 7 and the check in interval set to 2hrs when I noticed it was 2.2. 0 to10 is "optimal" so they say. Made some changes and will see how it goes.
Please log in to comment

Well looks as if I was on the right path. Good to hear a confirmation from people who have tested these things "in the wild".

Answered 01/28/2013 by: jdornan
Red Belt

  • Never doubted you for a minute ;)
  • Appreciate it but I do consider the everyday users the real experts.
Please log in to comment
Answer this question or Comment on this question for clarity