Sun Microsystems Logo
Products and Services
 
Support and Training
 
 

A  B  C  D  E  F  G  H  I  J  K  L  M  N  O  P  Q  R  S  T  U  V  W  X  Y  Z  
 
Maintenance Commandsnhcmmstat(1M)


NAME

 nhcmmstat - display information about peer nodes, trigger a switchover, or force the qualification of a master-eligible node

SYNOPSIS

 nhcmmstat [-h] [-c command [-t] ] [-n nodeid]

DESCRIPTION

The nhcmmstat tool displays information about a peer node or a group of peer nodes, displays notifications sent by the nhcmmd daemon, and performs operations on the cluster. Use this tool at regular intervals when you are performing tasks that might change the status of a node.

The nhcmmstat tool provides the following information:

  • The node ID of a node

  • That the cluster configuration files contain coherent information

  • The role of a node

  • The attributes of a node

  • That the master and vice-master disks contain the same shared information

You can use the nhcmmstat tool to modify the state of the cluster as shown in the examples section below.

For information about the role and attributes of a node, see the Netra High Availability Suite Foundation Services 2.1 6/03 CMM Programming Guide.

OPTIONS

The following options are supported by nhcmmstat:

-c

You can specify the nhcmmstat command to be executed. The specified command is executed and nhcmmstat exits. If this option is not used, you use the nhcmmstat command in an interactive mode. In this case, you must exit using the exit or quit command to return to the cursor.

-h

Displays help.

-n

You can specify the nodeid of the node on which you want to run nhcmmstat. This option is obligatory when using the info or potential commands.

-t

Shows the start and end times.

EXTENDED DESCRIPTION

The following commands can be used with the nhcmmstat tool to get information about a single node:

info

Get information about a node in the cluster. You must provide the nodeid of the node.

local

Get the nodeid of the current node.

master

Get information about the master node.

mynode

Get information about the current node.

potential

Get information about a node that is in the cluster node table but has the CMM_OUT_OF_CLUSTER role. You must provide the nodeid of the node.

When a node has the CMM_OUT_OF_CLUSTER role, the nhcmmstat tool gives meaningless values for the following administrative attributes: CMM_ELIGIBLE_MEMBER, CMM_FLAG_DISQUALIFIED, and CMM_FLAG_SYNCHRO_NEEDED.

vice

Get information about the vice-master node.

The following commands can be used with the nhcmmstat tool to get information about all peer nodes.

all

Get information about all peer nodes except those with the role CMM_OUT_OF_CLUSTER.

count

Get a count of the nodes in the cluster.

The following commands can be used with the nhcmmstat tool to modify the cluster. You must log in as superuser to use these commands.

reload

Force a reload of the cluster_nodes_table configuration. This command can be run from the master node only. The supported operations are add and remove a node in the cluster_nodes_table file, with the node powered off.

To add nodes that are not included in the original cluster definition, you must consider how the cluster was installed. For information about how to add new diskless node or dataless nodes to a cluster, see the Netra High Availability Suite Foundation Services 2.1 6/03 Cluster Administration Guide. You cannot add a master-eligible node to the cluster because there is a limit of two master-eligible nodes per cluster.

so

Force mastership change to the vice-master node. This command can be used on the master node only, and when the vice-master is present and able to take the master role.

squalif

Force the requalification for the current node to make it start an election. This command can be used on a master-eligible node when no peer node is qualified to be master. This command displays the following warning message:

Warning! This asynchronous command might take up to 300 s to succeed!

Note that an unsynchronized master-eligible node can only be elected as the master node if its former role was master.

The following commands can be used to exit from or get help with nhcmmstat:

exit

Exit.

help

Display help information.

quit

Exit.

If an nhcmmstat command fails, an error is displayed.

Notifications

When nhcmmstat is used in interactive mode, the following notifications are emitted by the nhcmmd daemon and displayed.

CMM_INVALID_CLUSTER

"[USER CB] INVALID CLUSTER" is displayed. The cluster is not in a coherent state. No information can be returned by the CMM API until the cluster has been returned to a coherent state.

CMM_VALID_CLUSTER

"[USER CB] VALID CLUSTER" is displayed. The cluster is in a coherent state.

CMM_STALE_CLUSTER

"[USER CB] STALE CLUSTER" is displayed. The cluster has not had a master node for the preceding ten seconds.

CMM_MASTER_DEMOTED

"[USER CB] master demoted = %d" is displayed. %d is replaced by the nodeid of the demoted master.

CMM_MASTER_ELECTED

"[USER CB] master elected = %d" is displayed. %d is replaced by the nodeid of the new master. When this notification is received, all available information about the master is displayed.

CMM_MEMBER_LEFT

"[USER CB] member left cluster = %d" is displayed. %d is replaced by the nodeid of the node that has left the cluster.

CMM_MEMBER_JOIN

"[USER CB] new node in cluster = %d" is displayed. %d is replaced by the nodeid of the new node.

CMM_VICEMASTER_DEMOTED

"[USER CB] vice-master demoted = %d" is displayed. %d is replaced by the nodeid of the demoted vice-master node.

CMM_VICEMASTER_ELECTED

"[USER CB] vice-master elected = %d" is displayed. %d is replaced by the nodeid of the new vice-master. When this notification is received, all available information about the vice-master is displayed.

When nhcmmstat is used in command line mode the notifications are not displayed.

Node Information Displayed

When information is requested for a node, or when notifications of cluster changes are received, the following information is displayed in the order shown:

nodeid

The nodeid of the current node, followed by "[This is the current node]" when the displayed information concerns the current node.

domain_id

The domainid of the cluster of the current node.

name

The name of the node as specified in the /SUNWcgha/remote/etc/cluster_nodes_table file.

role

The role of the node in the cluster: master, vice-master, in, or out. The master, vice-master and out roles correspond to CMM_MASTER, CMM_VICEMASTER, and CMM_OUT_OF_CLUSTER. A node that does not have the CMM_OUT_OF_CLUSTER role, is in.

qualified

YES or NO is displayed. If YES, the node is qualified to be master. If NO, the node is not qualified to be master. This information is relevant for master-eligible nodes only.

synchro

NEEDED or READY is displayed. If NEEDED, the master and vice-master shared file systems do not contain the same information. If READY, the master and vice-master node file systems contain the same information. This information is relevant for master-eligible nodes only.

frozen

YES or NO is displayed. If YES, the node is frozen. When a node is frozen, the master cannot change the role of this node even if events require it. If NO, the node is not frozen.

excluded

YES or NO is displayed. If YES, the node is excluded from the cluster. An excluded node acts as if it has the CMM_OUT_OF_CLUSTER role. If NO, the node is not excluded.

eligible

YES or NO is displayed. If YES, this node can participate in an election and be elected master if it is sufficiently qualified. If NO, this node cannot participate in an election.

incarn

The incarnation number of the time that the node was last booted. The value is an integer (number of seconds since 00:00 universal coordinated time Jan 1 1970 ) and a literal representation of this date. For example, 1005833787 (15/11/2001 - 15:16:27).

swload_id

This string indicates the Foundation Services software version. The string 1 is displayed for the Foundation Services.

CGTP @

This is the address of the node of the cgtp0 interface

USAGE

You must log in as superuser to use the so, reload, and squalif commands of the nhcmmstat tool.

EXAMPLES

This section contains examples of how to use the nhcmmstat tool:

Example 1. To Get Information About the Master Node
  • Log in to a peer node.

  • Type:

    # nhcmmstat -c master
    

    An output similar to the following is displayed:

    ------------------------------
    node_id     = 20 
    domain_id   = 250
    name        = netraMEN2-cgtp0
    role        = MASTER
    qualified   = YES
    synchro.    = READY
    frozen      = NO
    excluded    = NO
    eligible    = YES
    incarn.     = 1015949483 (12/03/2002 - 17:11:23)
    swload_id   = 1
    CGTP @      = 10.250.3.20
    ------------------------------
    
Example 2. To Get the nodeid of the Current Node
  • Log in to a peer node.

    This node becomes the current node.

  • Type:

    # nhcmmstat -c local
    

    An output similar to the following is displayed:

    Local Node id is 10
    
    

    In this example the nodeid of the current node is 10.

You can also find the nodeid of a node by using the ifconfig command. The nodeid corresponds to the host part of the nodes IP address. For more information, see the ifconfig(1m) man page.

Example 3. To Get Information About a Specific Node
  • Log in to a peer node.

  • Type:

    # nhcmmstat -c info -n nodeid
    

    An output similar to the following is displayed:

    ------------------------------
    node_id     = nodeid
    domain_id   = 250
    name        = netraMEN1-cgtp0
    role        = VICE-MASTER
    qualified   = YES
    synchro.    = READY
    frozen      = NO
    excluded    = NO
    eligible    = YES
    incarn.     = 1008266390 (13/12/2001 - 18:59:50)
    swload_id   = 1
    CGTP @      = 10.250.3.10
    ------------------------------
Example 4. To Force the Qualification of a Node Asynchronously
  • Log in to a master-eligible node.

  • Type:

    # nhcmmstat -c squalif
    
  • The nhcmmstat and squalif tool forces the requalification of the current node to make it the master node. This function can only be successful when there is no active master node in the cluster and the current node is a master-eligible node.

    The squalif tool is asynchronous. The tool is not blocked while qualification is taking place.

Example 5. To Get Information About All Peer Nodes
  • Log in to any peer node.

  • Type:

    # nhcmmstat -c all
    

    Information about each peer node is printed to the console.

Example 6. To Trigger a Switchover
  • Log in to the master node.

  • Type:

    # nhcmmstat -c so
    

    An output similar to the following is displayed:

    [USER CB] master elected = 10
    ------------------------------
    node_id     = 10 
    domain_id   = 250
    name        = netraMEN1-cgtp0
    role        = MASTER
    qualified   = YES
    synchro.    = NEEDED !!!
    frozen      = NO
    excluded    = NO
    eligible    = YES
    incarn.     = 1008266390 (13/12/2001 - 18:59:50)
    swload_id   = 1
    CGTP @      = 10.250.3.10
    ------------------------------
    [USER CB] vicemaster elected = 20
    ------------------------------
    node_id     = 20   [This is the current node]
    domain_id   = 250
    name        = netraMEN2-cgtp0
    role        = VICE-MASTER
    qualified   = YES
    synchro.    = NEEDED !!!
    frozen      = NO
    excluded    = NO
    eligible    = YES
    incarn.     = 1008266566 (13/12/2001 - 19:02:46)
    swload_id   = 1
    CGTP @      = 10.250.3.20
    ------------------------------
    

    If there is a vice-master qualified to become master, it is elected master, the master becomes the vice-master, and the disks are synchronized. If there is no potential master, nhcmmstat does not perform a switchover.

ATTRIBUTES

See attributes(5) for descriptions of the following attributes:

ATTRIBUTE TYPEATTRIBUTE VALUE
ArchitectureSPARC
AvailabilitySUNWnhcmb
Interface StabilityEvolving

SEE ALSO

Intro(3CMM), nhcmmd(1M), cluster_nodes_table(4), and nhinstall(1M).


Netra HAS FS 2.1Go To TopLast Changed September 2004