config always stale on a new node - Veritas Cluster Server

This is a discussion on config always stale on a new node - Veritas Cluster Server ; Hi Folks, Have a wierd problem, recently added a fourth node to the 3 node cluster that we had. The only difference between the existing nodes and the current node is that the new node uses /dev/eri0 and /dev/qfe0 for ...

+ Reply to Thread
Results 1 to 2 of 2

Thread: config always stale on a new node

  1. config always stale on a new node


    Hi Folks,
    Have a wierd problem, recently added a fourth node to the 3 node cluster
    that we had. The only difference between the existing nodes and the current
    node is that the new node uses /dev/eri0 and /dev/qfe0 for heartbeat, while
    all previous nodes use /dev/qfe0 and /dev/qfe1.

    The problem I am facing is that the new node would always say that its configuration
    is stale. And when I do a hasys -display I notice that the configchecksum
    is different on this node.

    Any ideas on why this is happening?

    Otherwise the cluster is functioning well. Gab works both ports a and h are
    up.
    lltstat -n doesn't report any errors. the failover groups work well. Except
    that the configuration state on the new node always says stale. Even after
    dumping the conf several times. There is no .stale file in the config directory
    either.

  2. Re: config always stale on a new node


    have you tried to run an hasys -force on the new node?


    "Arun" wrote:
    >
    >Hi Folks,
    >Have a wierd problem, recently added a fourth node to the 3 node cluster
    >that we had. The only difference between the existing nodes and the current
    >node is that the new node uses /dev/eri0 and /dev/qfe0 for heartbeat, while
    >all previous nodes use /dev/qfe0 and /dev/qfe1.
    >
    >The problem I am facing is that the new node would always say that its configuration
    >is stale. And when I do a hasys -display I notice that the configchecksum
    >is different on this node.
    >
    >Any ideas on why this is happening?
    >
    >Otherwise the cluster is functioning well. Gab works both ports a and h

    are
    >up.
    >lltstat -n doesn't report any errors. the failover groups work well. Except
    >that the configuration state on the new node always says stale. Even after
    >dumping the conf several times. There is no .stale file in the config directory
    >either.



+ Reply to Thread