Adding nodes to a cluster

Before you add a node to a cluster, you must make sure that the switch zoning is configured such that the node being added is in the same zone as all other nodes in the cluster.

If you are replacing a node and the switch is zoned by worldwide port name (WWPN) rather than by switch port, make sure that the switch is configured such that the node being added is in the same VSAN/zone.

Special procedures when adding a node to a cluster

If you are adding a node that has been used previously, either within a different I/O group within this cluster or within a different cluster, consider the following situations before adding the node. If you add a node to the cluster without changing its worldwide node name (WWNN), hosts might detect the node and use it as if it were in its old location. This could cause the hosts to access the wrong virtual disks (VDisks).

  • If you are re-adding a node back to the same I/O group after a service action required the node to be deleted from the cluster and the physical node has not changed, then no special procedures are required and the node can be added back to the cluster.
  • In a service situation, a node should normally be added back into a cluster using the original node name. As long as the partner node in the I/O group has not been deleted too, this is the default name used if -name is not specified.
  • If you are replacing a node in a cluster, either because of a node failure or an upgrade, you must change the WWNN of the new node to match that of the original node before you connect the node to the fibre channel network and add the node to the cluster.
  • If you are creating a new I/O group in the cluster and are adding new node, then there are no special procedures since this node has never been added to a cluster before and its WWNN that has not existed before.
  • If you are creating a new I/O group in the cluster and are adding new node, but this node has been added to a cluster before, then host system might still be configured to the node WWPNs and the node might still be zoned in the fabric. Since you cannot change the WWNN for the node, you must ensure other components in your fabric are configured correctly. Verify that any host that was previously configured to use the node has been correctly updated. Also verify the fabric zoning does not currently include this node in any zones.
  • If the node you are adding was previously replaced, either for a node repair or upgrade, you might have used the node's WWNN for the replacement node. Ensure that the WWNN of this node was updated so that you do not have two nodes with the same WWNN attached to your fabric. Also ensure that the WWNN of the node that you are adding is not 00000. If it is 00000, contact your support representative.
Note: Consider the following information when using multipathing device drivers:
  • Applications on the host systems direct I/O operations to file systems or logical volumes that are mapped by the operating system to virtual paths (vpaths), which are pseudo disk objects that are supported by the multipathing device drivers. Multipathing device drivers maintain an association between a vpath and a SAN Volume Controller virtual disk (VDisk). This association uses an identifier (UID) which is unique to the VDisk and is never reused. The UID allows multipathing device drivers to directly associate vpaths with VDisks.
  • Multipathing device drivers operate within a protocol stack that contains disk and fibre channel device drivers that allow it to communicate with the SAN Volume Controller using the SCSI protocol over fibre channel as defined by the ANSI FCS standard. The addressing scheme that is provided by these SCSI and fibre-channel device drivers uses a combination of a SCSI logical unit number (LUN) and the worldwide node name (WWNN) for the fibre-channel node and ports.
  • If an error occurs, the error recovery procedures (ERPs) operate at various tiers in the protocol stack. Some of these ERPs cause I/O to be redriven using the same WWNN and LUN numbers that were previously used.
  • Multipathing device drivers do not check the association of the VDisk with the vpath on every I/O operation that it performs.

Adding nodes to a cluster using the SAN Volume Controller Console

Attention:
  1. If you are re-adding a node to the SAN, ensure that you are adding the node to the same I/O group from which it was removed. Failure to do this can result in data corruption. You must use the information that was recorded when the node was originally added to the cluster. If you do not have access to this information, call the IBM® Support Center to add the node back into the cluster without corrupting the data.
  2. In a service situation, a node should normally be added back into a cluster using the original node name. As long as the partner node in the I/O group has not been deleted too, this is the default name used if -name is not specified.
  3. The LUNs that are presented to the ports on the new node must be the same as the LUNs that are presented to the nodes that currently exist in the cluster. You must ensure that the LUNs are the same before you add the new node to the cluster.
  4. LUN masking for each LUN must be identical on all nodes in a cluster. You must ensure that the LUN masking for each LUN is identical before you add the new node to the cluster.
  5. You must ensure that the model type of the new node is supported by the SAN Volume Controller software level that is currently installed on the cluster. If the model type is not supported by the SAN Volume Controller software level, upgrade the cluster to a software level that supports the model type of the new node. See the following Web site for the latest supported software levels:

    Support for SAN Volume Controller (2145) Web site

Each node in an I/O group must:
  • Be connected to a different uninterruptible power supply.
  • Have a unique name. If you do not provide a name, the cluster assigns a default name to the object.
Note: Whenever possible you must provide a meaningful name for objects to make identifying that object easier in the future.

This task assumes that you have already launched the SAN Volume Controller Console.

Perform the following steps to add a node to a cluster:
  1. Click Work with Nodes > Nodes in the portfolio. The Viewing Nodes panel is displayed.
  2. Select Add a node from the task list and click Go. The Adding a Node to a Cluster panel is displayed.
  3. Select the node that you want to add to the cluster from the Available Candidate Nodes list.
  4. Select the I/O group from the I/O Groups list.
  5. In the Node Name field, type the name that you want to assign to the node.
  6. Click OK.
  7. If you are adding a node into the cluster for the first time, record the following information:
    • Node serial number
    • All WWPNs
    • The I/O group that the node belongs to
    Important: You need this information to avoid possible data corruption if you have to remove and re-add the node to the cluster.
Parent topic: Creating a SAN Volume Controller cluster
Parent topic: Scenario: typical usage for the SAN Volume Controller Console
Parent topic: Nodes
Related concepts
Clusters
Nodes and clusters
Library | Support | Terms of use | Feedback
© Copyright IBM Corporation 2003, 2009. All Rights Reserved.