Kaseya Community

CPU/ Memory alarms

  • What is your opinion about these alarms

    80 % Committed Bytes In Use
    Alarm Duration: 2 Hour(s)

    In my mind I am thinking. Ok the computer is doing something. IF its early am of late night in might be updates. Should we be proactive about these or delete them?

  • We generally do not apply performance ALERTS anywhere, and never apply performance monitors to workstations. Client systems are rarely performance-optimized, so applying these alerts is a sure way to overload your ticketing system. We provide our MSP clients with performance MONITORS that can be used to evaluate performance and provide insight to optimize the platform. This requires customer cooperation - optimization might require changes to RAM or CPU counts, which could affect costs. Once the optimizations are done, performance alerts begin to make some sense.

    We actually filter the performance alerts so they only turn them into tickets when the MSP helpdesk is staffed, otherwise you have a ticket for an event with little evidence of what caused the issue (unless it's still happening, which usually isn't the case!)

    Next thing to consider - the 80% committed bytes alert arrives.. what do you do? Unless you have a specific process to follow for an alert, what value does monitoring it provide? If I checked our clients - some 65,000 agents in total at this point, I doubt I'd find more than 10 systems where performance alerting was enabled, and just a handful more where performance monitoring was in place. Our guidance is to have a procedure to deploy perfmon when a client complains about performance and get 24-48 hours (minimum) of logs to analyze the environment and identify the problem - and a solution.

    Glenn

  • We have a script that runs when the alarm is initiated that queries the top ten running processes using CPU and or memory and save that list in a text file, then save it in the get file location for that computer in the VSA.   When an alarm comes in, we look at the text file and it tells us what at that time what was using the resources to pinpoint that process for further review.

    Process Name            CPU (%) Memory (MB)

    ------------            ------- -----------

    _Total                      390     1941.36

    Idle                        356        0.01

    HPCollaborationKeyboard      11        6.03

    Teams#3                      11      251.45

    WmiPrvSE#3                    5         6.6

    Teams                         5       65.15

    svchost#30                    0        0.71

    svchost#21                    0        0.09

    svchost#29                    0        0.56

    svchost#26                    0       10.85

  • That sounds cool. It could come in handy.