Wireless Access

 View Only
  • 1.  Aruba 8 VMC Cluster Problem

    Posted Apr 21, 2023 05:01 AM

    Hello community,

    I have a customer with an installation of 2 Mobility Conductors and 2 virtuell Controllers, all VMs running in an ESX enviroment with distributed switches.
    There is only one interface in use on the VMC (GE0/0/0 - Network Adapter 2). Promisc. mode, MAC-add changes and forget transmits are enabled.
    On the ports to the VMCs all VLAN are allowed in ESX.
    Aruba OS Version is 8.7.

    I want to form a cluster but the status of both cluster members is still "isolated leader". 
    In CLI the command "show lc-cluster group-membership" says "peer .... DISCONNECTED".
    This is on both controllers.
    I also tried to exclude VLANs, but since all of the configured VLANs are used and working (even VLAN 1), I don't thinks that's the problem.

    I have some other customers with hardware controllers (eg. 7205) and there is no problem with clustering - I did it the same way.

    So my question is: Do I have to consider something with configuring a cluster with VMCs?
    Are there any tips to troubleshoot this problem?

    Thanks in advance,

    Regards, Steffen



  • 2.  RE: Aruba 8 VMC Cluster Problem

    Posted Apr 23, 2023 02:45 PM

    Nothing different other than you called out about promiscuous mode, forged transmit, etc..

    Excluded VLANs are VLANs that are not shared to both controllers. VLAN probes gets sent between the vMCs to check L2 reachability in order to be a L2-Connected Cluster, which offers a faster and more seamless failover when a vMC goes down. Can you share how you have the cluster service set up at the folder level and at the vMC level?



    ------------------------------
    Dustin Burns

    Lead Mobility Engineer @Worldcom Exchange, Inc.

    ACCX 1271| ACMX 509| ACSP | ACDA | MVP Guru 2022-2023
    If my post was useful accept solution and/or give kudos
    ------------------------------



  • 3.  RE: Aruba 8 VMC Cluster Problem

    Posted Apr 24, 2023 05:50 AM

    yes of course:

    Also here is the output of CLI of one of the VMCs. The output from the other one is the same.

    (....<NAME-of-VMC01>....) [MDC] #show lc-cluster group-membership
    
    Cluster Enabled, Profile Name = "Cluster-D..........."
    Redundancy Mode On
    Active Client Rebalance Threshold = 50%
    Standby Client Rebalance Threshold = 75%
    Unbalance Threshold = 5%
    Heartbeat Threshold = 900 msec
    AP Load Balancing: Enabled
    Active AP Rebalance Threshold = 20%
    Active AP Unbalance Threshold = 5%
    Active AP Rebalance AP Count = 50
    Active AP Rebalance Timer = 1 minutes
    Cluster Info Table
    ------------------
    Type IPv4 Address    Priority Connection-Type STATUS
    ---- --------------- -------- --------------- ------
    self   10.35.xxx.xxx      128             N/A ISOLATED (Leader)
    peer   10.35.xxx.yyy      128             N/A DISCONNECTED
    
    
    (....<NAME-of-VMC01>....) [MDC] #show crypto ipsec sa peer 10.35.xxx.yyy
    
    % No active IPSEC SA for 10.35.xxx.yyy







    ------------------------------
    Steffen
    ------------------------------



  • 4.  RE: Aruba 8 VMC Cluster Problem

    Posted Apr 24, 2023 05:54 AM
      |   view attached

    one more annotation: today I updates to 8.10.0.6 but behavior is the same



    ------------------------------
    Steffen
    ------------------------------



  • 5.  RE: Aruba 8 VMC Cluster Problem

    Posted Apr 28, 2023 07:33 PM

    Can you share the output of these commands on each controller?

    show lc-cluster vlan-probe status
    lc-cluster start-vlan-probe



    ------------------------------
    Dustin Burns

    Lead Mobility Engineer @Worldcom Exchange, Inc.

    ACCX 1271| ACMX 509| ACSP | ACDA | MVP Guru 2022-2023
    If my post was useful accept solution and/or give kudos
    ------------------------------



  • 6.  RE: Aruba 8 VMC Cluster Problem

    Posted Apr 24, 2023 01:57 PM

    We had very similar symptoms with one customer. In our case, the cluster of 2 VMCs worked properly for over a year, and suddenly there were problems.

    "show lc-cluster group-membership" said on each VMC self - isolated, peer - disconected. On each VMC in the log we could see that cluster-members are continuously connected and disconnected. There were definitely no configuration changes in the WLAN environment.

    After long interviews, we discovered that new Windows Server VMs were deployed to the ESX host, overbooking the host CPU.

    After moving new VMs to other ESX hosts, everything worked immediately. Cluster was l2 connected, AP and user load balancig was immediately active. Without a configuration change in the VMCs.

    So check RAM and CPU usage on the ESX host.

    I hope it helps. 



    ------------------------------
    Regards,

    Waldemar
    ACCX # 1377, ACEP, ACA - Network Security
    If you find my answer useful, consider giving kudos and/or mark as solution
    ------------------------------