none
Resource Group Discovery Limit Exceeded RRS feed

  • Question

  • In SCOM we are receiving an alert for the following Rule "Resource Group Discovery Limit Exceeded" and the Alert description is the following

    "The number of resource groups in the cluster exceeds the configured Resource Group Discovery Limit. Please refer to the management pack guide for guidance."

    When looking in the Windows Cluster Management pack documentation it says the following

    "For scalability reasons, discovery and monitoring of resource groups is limited by default to a maximum of 300 resource groups. It is possible but not recommended to override this for specific clusters. An alert (Resource Group Discovery Limit Exceeded) will be generated if a cluster containing more than 300 resource groups is discovered. If you override the limit for specific clusters, CPU utilization for the Operations Manager agent will increase.
    Resource discovery and monitoring is disabled by default because the resource group monitors provide sufficient monitoring. It is not recommended to enable resource discovery and monitoring, but it is possible to do so by using the Discover Cluster Resources override on the Windows Server 2003, 2008, 2008 R2, 2012, and 2012 R2 Cluster Discovery. By default, this has a value of false."

    Can anyone give me any advice how I should proceed with this? It recommends not creating an override for the specific cluster as it will increase the CPU on each agent?

    Any help with this would be greatly appreciated.

    Thanks


    Wednesday, October 30, 2019 1:10 PM

All replies

  • Hi,

    How many resource groups do you have on the cluster that is alerting?

    What kind of workloads are you running on the cluster?

    Are the cluster nodes virtual or physical?

    I would say it depends on the capacity of the cluster nodes, if they are physical you would probably be ”fine” to create an override for this, assuming your cluster nodes have sufficient computing power (i.e. sufficient CPU power). But it also depends on the workloads you’re running on the cluster.

    If the cluster nodes are virtualized, you might want to reconsider, but here again I would say it depends on the workloads running on the cluster, and how CPU allocation you have.


    Best regards,
    Leon




    Blog: https://thesystemcenterblog.com LinkedIn:

    Wednesday, October 30, 2019 3:28 PM
  • Hi,
     
    Agree with Leon, before we override the limit, please make sure the CPU and capability is enough to handle on agent. Meanwhile ,please make sure your SCOM server has the capability to monitor more Resource Group. You can increase the limit via a few times to see if our environment will be affected.
     
    Hope it can help.
     
    Best regards.
    Crystal

    Please remember to mark the replies as answers if they help.
    If you have feedback for TechNet Subscriber Support, contact tnmff@microsoft.com.

    Thursday, October 31, 2019 1:58 AM
  • Morning

    In response to your post the cluster is our production cluster where all our hyper-v virtual machines exist. So this is going to be growing as and when we bring on new services. 

    When it states it will raise the CPU of the agent does this mean it will raise the CPU of each of the Hyper-V hosts or for each virtual machine that exists within the cluster. - Our Hosts are physical Gen 10 HP Blades with plenty of resources (CPU/RAM). - How can we be sure that if we override the limit to something much higher

    I have been told that we shouldn't be monitoring our hyper-v cluster using the Windows Fail-over Cluster Management pack anyway - is this correct? - any help would be appreciated.
    Thursday, October 31, 2019 9:16 AM
  • It will raise some CPU usage on the host running the SCOM agents, not on the virtual machines themselves.

    You can use the following management pack to monitor your Hyper-V environment:

    If you're using SCVMM, you can use these ones:


    Blog: https://thesystemcenterblog.com LinkedIn:

    Thursday, October 31, 2019 9:26 AM
  • Ok - we have recently upgraded SCOM to 2019 and we also have SCVMM on 2019. We have been having issues getting them both integrated however once we get that sorted do we still need to install the management packs you have provided? Being as when the integration takes place it imports the MPs itself. 

    Any clarification on whether we should stop monitoring this cluster via the Windows Cluster MP?

    Thursday, October 31, 2019 11:24 AM
  • If it’s a Hyper-V cluster and you use SCVMM for managing Hyper-V, then simply make the integration and it should be enough, you’ll need to download and import the SCVMM management packs during the integration.

    Integrate VMM with Operations Manager for monitoring and reporting

    I wouldn’t use the normal Windows cluster management for monitoring Hyper-V workloads.


    Blog: https://thesystemcenterblog.com LinkedIn:

    Thursday, October 31, 2019 11:32 AM
  • Hi Leon - sorry for not getting back sooner on this. Is there any guide out there that will outline the steps to get SCOM integrated to SCVMM?

    We did do it previously on our 2012 R2 environment but when we tried it on our 2019 environment it was failing for some reason.

    In regards to management packs doesn't SCVMM import the management packs during the integration process?

    Many thanks

    Tuesday, November 26, 2019 10:51 AM
  • You'll find everything you need here:

    Integrate VMM with Operations Manager for monitoring and reporting
    https://docs.microsoft.com/en-us/system-center/vmm/monitors-ops-manager?view=sc-vmm-2019

    There are some guides out there on the web, here's one for example:
    Configure SCOM and VMM Integration


    Blog: https://thesystemcenterblog.com LinkedIn:



    • Edited by Leon Laude Tuesday, November 26, 2019 10:56 AM
    Tuesday, November 26, 2019 10:54 AM