cluster command

Purpose

Creates a cluster that uses the Virtual I/O Server (VIOS) partitions and performs cluster-related operations.

Syntax

To create a cluster:

cluster -create -clustername ClusterName -repopvs PhysicalVolume ... -sp StoragePool -sppvs PhysicalVolume ... [-hostname HostName]

cluster -create -clustername ClusterName -repopvs PhysicalVolume ... -sp StoragePool -file -sppvs FileName [-hostname HostName]

cluster -create -clustername ClusterName -repopvs PhysicalVolume ... -sp StoragePool [-systier [SysTierName:] PhysicalVolume ...] -usrtier UsrTierName: PhysicalVolume ... [-hostname HostName]

cluster -create -clustername ClusterName -repopvs PhysicalVolume ... -sp StoragePool -file [-systier [SysTierName:] FileName] -usrtier UsrTierName: FileName1 [-hostname HostName]

To add a VIOS partition node to the cluster:

cluster -addnode [-clustername ClusterName] -hostname HostName

To remove a VIOS partition node from the cluster:

cluster -rmnode [-f] [-clustername ClusterName] {-hostname HostName | -uuid HostUuid}

To remove a cluster from the system:

cluster -remove [-clustername ClusterName]

To display the health status of a cluster:

cluster -status [-clustername ClusterName] [-field FieldName... ] [-fmt Delimiter] [-verbose ]

To synchronize the cluster and Object Data Manager (ODM):

cluster -sync [-clustername ClusterName]

To list all clusters:

cluster -list [-field Fieldname ... ] [-fmt Delimiter ]

Note: If IPv6 is configured, then for setting up a VIOS node as a member of a Shared Storage Pool (SSP) cluster, IPv6 stateless auto configuration is suggested.

Description

The cluster command is used to create and remove the cluster.

You can create a cluster with a separate restricted system tier (holds metadata only) and user tier (holds user data only) by using the -systier option and the -usrtier option. By default, a cluster is created with the single co-mingled tier (holds metadata and user data) using the -sppvs option or only the -usrtier option.

The status subcommand displays a global cluster health status and the status for each node in the cluster. The cluster status can be OK, DEGRADED, or DOWN. If the cluster status is OK, all the nodes of the cluster are working. If the cluster health status is DEGRADED, cluster services may not be active on some of the nodes. If the cluster status is DOWN, the cluster services are not active on all the nodes. The node status is either OK or DOWN. A node is in the OK state, if the cluster services are active, and a node is in the DOWN state, if the cluster services are not active on that node. A pool status is provided for each node that is not DOWN. The pool status is OK for a node if the node can service the shared storage pool (SSP) virtualized storage to the client logical partitions (LPARs). The pool status is DOWN if it is not usable from the node.

If the -verbose option is requested, additional status details such as the number of nodes in each state is provided. For each node that is not DOWN, it displays the node roles and the storage pool status from the point of view of the node.

If the -fmt option is requested, status output is displayed in a format of delimiter-separated value report.

With the -field option, you can specify which fields to display. If no pool information is requested, the command does not retrieve pool status on each node, thereby accelerating the command response time.

Flags

Flag name Description
-addnode Adds a member node to the cluster.
-clustername Specifies the name of the cluster.
-create Creates a cluster.
-remove Removes a cluster.
-f Forces the removal of the VIOS partitions from the cluster.
-field
Supports the following fields if the -list flag is specified:
cluster_name, cluster_id
Supports the following fields if the -status flag is specified:
cluster_name, cluster_state, node_name, node_mtm, node_partition_num, 
node_state, pool_state
Supports the following fields if the -status flag is specified with the -verbose option:
cluster_name, cluster_id, cluster_state, repos_mode, number_of_nodes, nodes_ok, 
nodes_down, pool_name, pool_id, pool_state, node_name, node_id, node_mtm,
node_partition_num, node_state, node_repos_state, node_upgrade_status, 
node_roles
-file Specifies that you must specify a file name with the -sppvs option. The file must contain physical volume names separated by a space.
-fmt Divides output by a user-specified delimiter.
-hostname Specifies the host name or IP address of the VIOS partition.
-list Lists all clusters which have membership of the current node.
-repopvs Lists physical volumes on which a repository can be created.
-rmnode Removes the member node from the cluster.
-sp Specifies the name of the storage pool.
-sppvs Lists the physical volumes on which a shared storage pool can be created. If you specify the -file option, this flag lists the file name with the list of physical volume names.
-systier

Specifies the optional name of the system tier. A list of physical volumes or file names follows, which is used to create a system tier.

-usrtier

Specifies the name of the user tier. A list of physical volumes or file names follows, which is used to create a user tier. Without the -systier option, the cluster command creates a co-mingled type tier.

-status Displays the health status of the cluster.
-sync Synchronizes the cluster and the ODM.
-uuid Specifies the UUID of a node.
-verbose Displays additional detailed information.

Examples

  1. To create a cluster with a co-mingled tier, type the following command:
    cluster -create -clustername testcluster -repopvs hdisk1 -sp testpool -sppvs hdisk2
    The system displays the output as follows:

    Cluster testcluster has been created successfully.

  2. To create a cluster with a co-mingled tier by using the -file flag, type the following command:
    cluster -create -clustername testcluster -repopvs hdisk1 -sp testpool -file -sppvs pvlist.txt
    The system displays the output as follows:

    Cluster testcluster has been created successfully.

  3. To create a cluster with a restricted system tier and a user tier, type the following command:
    cluster -create
                -clustername testcluster -repopvs repo_hdisk -sp testsp -systier metadatatier: hdisk2
                -usrtier datatier: hdisk3
    The system displays the output as follows:

    Cluster testcluster has been created successfully.

  4. To create a cluster with a co-mingled tier, type the following command:
    cluster -create -clustername testcluster -repopvs hdisk1 -sp testsp -usrtier mixedtier: hdisk2
    The system displays the output as follows:

    Cluster testcluster has been created successfully.

  5. To remove a cluster from the system, type the following command:
    cluster -remove
    The system displays the output as follows:

    Cluster testcluster has been removed successfully.

  6. To add a member node in a cluster, type the following command:
    cluster -addnode -hostname testhost2
    The system displays the output as follows:

    testhost2 has been added to the testcluster cluster.

  7. To remove a member node from a cluster, type the following command:
    cluster -rmnode -hostname testhost2
    The system displays the output as follows:

    testhost2 has been removed from the testcluster cluster.

  8. To display the health status summary, type the following command:
    cluster -status
    The system displays the output as follows:
    Cluster Name State
    testcluster OK
    Node Name Pool State MTM Partition #State
    testhost1 OK 9115-505031006A2A 1 OK
    testhost2 DOWN 9115-5050310069FA 1 OK
  9. To display a detailed health status report, type the following command:
    cluster -status -verbose
    The system displays the output as follows:
    Cluster Name: testcluster
    Cluster Id: d23624fe335f11e0bd510011257e1447
    Cluster State: OK
    Repository Mode: EVENT
    Number of Nodes: 2
    Nodes OK: 2
    Nodes DOWN: 0
    Pool Name: testpool
    Pool Id: b118d73c42b322d0a5850011257e1447
    Pool Mirror State: NOT_MIRRORED
    Node Name: testhost2
    Node Id: 4d5012722f9f11e0813c0011257e1447
    Node MTM: 8233-E8B020687AER
    Node Partition Num: 2
    Node State: OK
    Node Repos State: OK
    Node Upgrade Status: 2.2.4.0 ON_LEVEL
    Node Roles:
    Pool Name: testpool
    Pool Id: b118d73c42b322d0a5850011257e1447
    Pool State: OK
    Node Name: testhost1
    Node Id: 3d5012722f9f11e0813c0011257e13444
    Node MTM: 8233-E8B020687AER
    Node Partition Num: 6
    Node State: OK
    Node Repos State: OK
    Node Upgrade Status: 2.2.4.0 ON_LEVEL
    Node Roles: DBN
    Pool Name: testpool
    Pool Id: b118d73c42b322d0a5850011257e1447
    Pool State: OK
  10. To display a formatted health status report, type the following command:
    cluster -status -fmt ’,’
    The system displays the output as follows:
    testcluster,OK,testhost1,9115-505031006A2A,4,OK,OK
    testcluster,OK,testhost2,9115-5050310069FA,9,OK,OK
  11. To display a formatted and detailed health status report, type the following command:
    cluster -status -fmt ’,’ -verbose
    The system displays the output as follows:
    testcluster,d23624fe335f11e0bd510011257e1447,OK,ASSERT,2,2,0,
    testhost1,da18d52a32b211e0a5850011257e1447,9115-505031006A2A,4,OK,OK,ON_LEVEL,DBN,testpool, 
    b118d73c42b322d0a5850011257e1447,OK
    testcluster,d23624fe335f11e0bd510011257e1447,OK,ASSERT,2,2,0,
    testhost2,4d5012722f9f11e0813c0011257e1447,9115-5050310069FA,9,OK,OK,ON_LEVEL,,testpool,
    b118d73c42b322d0a5850011257e1447,OK
  12. To display a formatted health status report with some specific fields, type the following command:
    cluster -status -fmt ’:’ -field cluster_name node_name pool_label pool_state -verbose
    The system displays the output as follows:
    testcluster:testhost1:testpool:OK
    testcluster:testhost2:testpool:OK
  13. To display a specific health status report with specific fields, type the following command:
    cluster -status -field node_name node_state cluster_name cluster_state
    The system displays the output as follows:
    Node Name: testhost1
    Node State: OK
    Cluster Name: testcluster
    Cluster State: OK
    Node Name: testhost2
    Node State: OK
    Cluster Name: testcluster
    Cluster State: OK
  14. To display the cluster information on the current node, type the following command:
    cluster -list
    The system displays the output as follows:
    CLUSTER_NAME: testcluster
    CLUSTER_ID: a64c8c725bfc11e1993500215e188da0
  15. To display the cluster information on the current node in a formatted output by using the -fmt flag, type the following command:
    cluster -list -fmt ,
    The system displays the output as follows:
    testcluster,a64c8c725bfc11e1993500215e188da0



Last updated: Wed, November 18, 2020