4 Replies Latest reply: Jan 9, 2013 12:26 PM by wangjundragon RSS

two nodes cluster

wangjundragon
Currently Being Moderated

In a two-node cluster, epsilon, the value that indicates a node as the master and a requirement in clusters with more than two nodes, is

not held by either node. Instead, both nodes are continuously polled to ensure that if takeover occurs,the node that is still up and running has full read-write access to data as well as access to logical  interfaces and management functions. This continuous polling function is referred to as cluster high

availability or cluster HA.

 

Continuously polled who can be explained in more detail

  • Re: two nodes cluster
    parisi
    Currently Being Moderated

    "Continuous polling" means that the system will always check that the nodes are up and running. If one node goes down, that node is then considered as the "epsilon" node (though it won't show in cluster show).

     

    The polling takes place on the mailbox disks.

    • Re: two nodes cluster
      wangjundragon
      Currently Being Moderated

      I want to know if the 4-node cluster, a node has gone down, followed by the worst happens,

      Another node goes down, It's epsilon node, This time cluster what state is, if node failure recovery, How do I restore the entire state of the cluster normal

      • Re: two nodes cluster
        Mrinal Devadas
        Currently Being Moderated

        If two nodes in a four node cluster are down and one of the nodes that is down owned epsilon then the next step is to call up NetApp Support with a priority 1 support case. This is because the Vservers in the cluster might not be able to consistently serve data to clients.

    • Re: two nodes cluster
      wangjundragon
      Currently Being Moderated

      I understand that the HA configuration, each node has its own status information in real time is written to the mailbox disk, partner node continues to check whether the mailbox disk information updated mailbox disks of a node is not updated when a certain period of time,that this node has failed, right?

      If so, how long is not updated, you will think that this node failure?

More Like This

  • Retrieving data ...