Cluster Server 7.4 Agent for EMC SRDF Configuration Guide - Windows

Last Published:
Product(s): InfoScale & Storage Foundation (7.4.1)
Platform: Windows
  1. Introducing the agent for EMC SRDF
    1.  
      About the agent for EMC SRDF
    2.  
      Supported hardware for EMC SRDF
    3.  
      Supported software
    4.  
      Typical EMC SRDF setup in a VCS cluster
    5. EMC SRDF agent functions
      1.  
        About the EMC SRDF agent's online function
      2.  
        About dynamic swap support for the EMC SRDF agent
    6.  
      Installing the agent for EMC SRDF
  2. Configuring the agent for EMC SRDF
    1. Configuration concepts for the EMC SRDF agent
      1.  
        Resource type definition for the EMC SRDF agent
      2. Attribute definitions for the SRDF agent
        1.  
          Required attributes
        2.  
          Optional attributes
        3.  
          Internal attributes
      3.  
        Sample configuration for the EMC SRDF agent
    2. Before you configure the agent for EMC SRDF
      1.  
        About cluster heartbeats
      2.  
        About configuring system zones in replicated data clusters
      3.  
        About preventing split-brain
    3. Configuring the agent for EMC SRDF
      1. Configuring the agent manually in a global cluster
        1.  
          Configuring the Symm heartbeat on each cluster
      2.  
        Configuring the agent manually in a replicated data cluster
      3.  
        Setting the OnlineTimeout attribute for the SRDF resource
      4.  
        Additional configuration considerations for the SRDF agent
  3. Testing VCS disaster recovery support with EMC SRDF
    1. How VCS recovers from various disasters in an HA/DR setup with EMC SRDF
      1.  
        Failure scenarios in global clusters
      2.  
        Failure scenarios in replicated data clusters
    2.  
      Testing the global service group migration
    3.  
      Testing disaster recovery after host failure
    4.  
      Testing disaster recovery after site failure
    5.  
      Performing failback after a node failure or an application failure
    6.  
      Performing failback after a site failure
  4. Setting up fire drill
    1.  
      About fire drills
    2.  
      Fire drill configurations
    3. About the SRDFSnap agent
      1.  
        SRDFSnap agent functions
      2.  
        Resource type definition for the SRDFSnap agent
      3.  
        Attribute definitions for the SRDFSnap agent
      4.  
        About the Snapshot attributes
      5.  
        Sample configuration for a fire drill service group
    4.  
      Additional considerations for running a fire drill
    5.  
      Before you configure the fire drill service group
    6. Configuring the fire drill service group
      1.  
        About the Fire Drill wizard
    7.  
      Verifying a successful fire drill

About cluster heartbeats

In a replicated data cluster, ensure robust heartbeating by using dual, dedicated networks over which the Low Latency Transport (LLT) runs. Additionally, you can configure a low-priority heartbeat across public networks.

In a global cluster, VCS sends ICMP pings over the public network between the two sites for network heartbeating. To minimize the risk of split-brain, VCS sends ICMP pings to highly available IP addresses. VCS global clusters also notify the administrators when the sites cannot communicate.

In global clusters, the VCS Heartbeat agent sends heartbeats directly between the Symmetrix arrays if the Symmetrix ID of each array is known. This heartbeat offers the following advantages:

  • The Symmetrix heartbeat shows that the arrays are alive even if the ICMP heartbeats over the public network are lost. So, VCS does not mistakenly interpret this loss of heartbeats as a site failure.

  • Heartbeat loss may occur due to the failure of all hosts in the primary cluster. In such a scenario, a failover may be required even if the array is alive. In any case, a host-only crash and a complete site failure must be distinguished. In a host-only crash, only the ICMP heartbeat signals a failure by an SNMP trap. No cluster failure notification occurs because a surviving heartbeat exists. This trap is the only notification to fail over an application.

  • The heartbeat is then managed completely by VCS. VCS reports that the site is down only when the remote array is not visible by the symrdf ping command.