cmapplyconf man page on HP-UX

Man page or keyword search:  
man Server   10987 pages
apropos Keyword Search (all sections)
Output format
HP-UX logo
[printable version]

cmapplyconf(1m)						       cmapplyconf(1m)

NAME
       cmapplyconf  -  verify and apply Serviceguard cluster configuration and
       package configuration files

SYNOPSIS
       cmapplyconf [-f] [-v] [[-k|-K] -C cluster_ascii_file]
       [[-p pkg_reference_file] | [-P pkg_ascii_file]...]

DESCRIPTION
       cmapplyconf verifies the cluster configuration and  package  configura‐
       tion   specified	  in   the   cluster_ascii_file	  and  the  associated
       pkg_ascii_file(s), creates or updates the  binary  configuration	 file,
       called  cmclconfig,  and distributes it to all nodes.  This binary con‐
       figuration file contains the cluster configuration information as  well
       as  package  configuration information for all packages specified. This
       file, which is used by the cluster daemons to manage the entire cluster
       and package environment, is kept in the $SGCONF directory.

       Only  a	superuser  with	 effective  user ID of zero (see id (1) and su
       (1)), can verify, create, or update the configuration.

       cmapplyconf verifies any configured external  script  program  in  each
       pkg_ascii_file for the package run and halt function by calling it with
       a "validate" parameter. The external script program is run on each mem‐
       ber  that  the package is configured to run on. A non-zero return value
       from any external script program will cause the command to fail.

       If the cluster_ascii_file specifies a quorum server as the cluster tie-
       breaker service, the quorum server must be running and all nodes in the
       cluster configuration must be authorized to access it. If more than one
       IP  address  is specified for the quorum server, the quorum server must
       be reachable from all configured nodes through all  the	IP  addresses.
       Otherwise the cmapplyconf command will fail.

       If  the	cluster_ascii_file  specifies  cluster lock lun devices as the
       cluster tie-breaker service, all nodes must be accessing the same phys‐
       ical device. The lock lun device file must be a block device file.  The
       cluster must be down to modify a cluster tie-breaking service.

       If changes to either the cluster configuration or to any of the package
       configuration  files  are  needed,  first  update the appropriate ASCII
       file(s) (cluster or package),  then  validate  the  changes  using  the
       cmcheckconf command and then use cmapplyconf again to verify and redis‐
       tribute the binary file to all nodes.  The cluster and package configu‐
       ration can be modified whether the cluster is up or down, although some
       configuration requires either the cluster or  the  package  be  halted.
       Please  refer  to  the  manual for more detail.	The cluster ASCII file
       only needs to be specified if configuring the  cluster  for  the	 first
       time, or if adding or deleting nodes to the cluster.  The package ASCII
       file only needs to be specified if the package is being	added,	or  if
       the package configuration is being modified.

       It is recommended that the user run the cmgetconf command to get either
       the cluster ASCII configuration file  or	 package  ASCII	 configuration
       file whenever changes to the existing configuration are required.

       Note  that cmapplyconf will verify and distribute cluster configuration
       or package files.  It will not cause the cluster	 daemon	 to  start  or
       application  packages  to run.  If the cluster is down, once the binary
       configuration file is distributed, use cmruncl  to  start  the  cluster
       daemons on all nodes.

       If the cluster is already up and running, cmapplyconf applies modifica‐
       tions to the existing  binary  configuration  file  while  the  cluster
       remains	active.	  The user needs to use the cmrunnode command to start
       the cluster activities on the newly added node(s), and use the cmmodpkg
       or cmrunpkg command to start the newly added package(s).

       If  cmapplyconf	is specified when the cluster or packages have already
       been configured, the cluster_ascii_file,	 and  package_ascii_file,  (or
       all ASCII files in the package_reference_file) will be scanned for con‐
       figuration changes.  If a node is specified in the  cluster_ascii_file,
       but  does not exist in the previous configuration, that node is consid‐
       ered as a new node and will be added to the new configuration.  A  node
       which exists in the previous configuration, but is not specified in the
       cluster_ascii_file, will be removed  from  the  cluster	configuration.
       The  node  needs to be halted before it can be removed from the cluster
       configuration.  The same kind of processing will apply to  the  package
       configuration  to  determine  whether  to  add or delete package nodes,
       package subnet, etc.  Not all package configuration changes require the
       package	to  be	halted.	  Similar processing will apply to the network
       configuration to determine whether to add or delete heartbeat  or  sta‐
       tionary	networks,  add or delete standby network interfaces change the
       network attributes, change the network polling  interval	 etc.  in  the
       cluster	configuration.	In general these changes can be made while the
       node and cluster are running, but note that at least one heartbeat net‐
       work must remain up and unchanged while the cluster is running. See the
       Managing Serviceguard manual for more information. It is recommended to
       use  the	 cmquerycl  -c	command	 especially to add new networks when a
       cluster is configured.  Please refer to cmquerycl  man  page  for  more
       detail.

       After  cmapplyconf  completes,  make sure to use cmmodpkg to enable all
       the package switching flags for packages that had been halted  earlier.
       Once package switching is enabled, the package will be started automat‐
       ically.

       Under Serviceguard Extension for RAC (HP-UX only), cmapplyconf  returns
       an  error,  if  a multi node package is configured with both SUBNET and
       CLUSTER_INTERCONNECT_SUBNET parameters that monitor the same subnet.

       Before starting configuration, make sure that all nodes have  the  same
       release of Serviceguard. If Serviceguard Extension for RAC, SGeRAC,
	(HP-UX	only) is installed, make sure that even SGeRAC release version
       matches.
	Configuration is only allowed if all the nodes in the cluster  are  on
       the same Serviceguard and SGeRAC, if applicable, release.

       Configurations  using  the  SITE_NAME  and SITE attributes (HP-UX only)
       must have the appropriate versions of Metrocluster and SGeRAC  software
       installed on all nodes.

   Options
       cmapplyconf supports the following options:

	      -v	Verbose output will be displayed.

	      -k	Using the -k option means that cmapplyconf only checks
			disk connectivity to the LVM volume  groups  that  are
			identified  in	the  ASCII file.  This option does not
			exit on Linux. Omitting the  -k	 option	 (the  default
			behavior)  means that cmapplyconf tests the connectiv‐
			ity of all LVM volume group on all the cluster	nodes.
			Using  -k can result in significantly faster operation
			of the command.	 Do not use -k the option when	remov‐
			ing  LVM  volume  groups from the cluster.  -k must be
			used with -C and can not be used with -K

	      -K	Using the -K option means that cmapplyconf only checks
			disk connectivity for cluster lock volume groups. This
			option does not exist on Linux.	  For  all  other  LVM
			volume	groups	no connectivity will be checked and no
			modification will be done to their  state  of  cluster
			awareness.  Omitting the -K and -k option (the default
			behavior) means that cmapplyconf tests the  connectiv‐
			ity of all LVM volume groups on all the cluster nodes.
			Using -K can result in significantly faster  operation
			of  the	 command.  -K can be used only when cluster is
			already configured and is used with -C -K can  not  be
			used with -k -K does not affect lock LUN checking.

	      -f	Force  the distribution even if a binary configuration
			file exists on any nodes.  The old  binary  configura‐
			tion  file  will be replaced.  If the -f option is not
			specified and a binary	file  exists  on  one  of  the
			nodes,	the  user  will	 be  asked  whether or not the
			existing  file	should	be  replaced.  If  a  negative
			response  is  given, neither the configuration nor the
			binary configuration file will be modified.  Note that
			in the cases when either the cluster or the package is
			not halted as required, the -f option will  not	 force
			the operation to go through.

	      -C cluster_ascii_file
			Name  of  the  cluster	ASCII  file to use.  This is a
			required parameter if the cluster has never been  con‐
			figured	 before.  If  not specified, the local cluster
			configuration is used.	An ASCII file for a new	 clus‐
			ter  is	 created  with the cmquerycl command. An ASCII
			file for an existing cluster is created with cmgetconf
			command.  See cmquerycl(1m) and cmgetconf(1m).

	      -P pkg_ascii_file...
			Name of the package configuration file(s) to use.  For
			a new package, a package configuration	template  file
			can be created by using the cmmakepkg command and must
			be customized to include specific information for  the
			package.  See cmmakepkg(1m).  A configuration file for
			an existing package can	 be  generated	by  using  the
			cmgetconf command. See cmgetconf(1m).

	      -p pkg_reference_file
			Name of the file containing a list of package configu‐
			ration file(s) to be used.  This file may be necessary
			if  the number of pkg_ascii_file names given with mul‐
			tiple -P options do not fit on the command line.  This
			option cannot be used with the -P option.

RETURN VALUE
       Upon completion, cmapplyconf returns one of the following values:

	       0   Successful completion.
	       1   Command failed.

EXAMPLES
       The  high availability environment contains an ASCII cluster configura‐
       tion file, clusterA.config and two packages, pkg1 and  pkg2,  specified
       in ASCII files pkg1.config and pkg2.config.

       To create and distribute the binary configuration file, use the follow‐
       ing command:

	      cmapplyconf -C clusterA.config -P pkg1.config -P pkg2.config

       To specify a long list of package configuration files, use the  follow‐
       ing command:

	      cmapplyconf -C clusterA.config -p file

       where file contains:

	      pkg1.config
	      pkg2.config
	      pkg3.config

       To  add	a  node,  node1 to the existing cluster configuration, use the
       following command:

	      cmgetconf -c clusterA clusterA.config

       Modify the ASCII file to add the node information accordingly

	      cmapplyconf -C clusterA.config

	      cmrunnode node1

       To delete a node, node2 from the existing  cluster  configuration,  use
       the following command:

	      cmgetconf -c clusterA clusterA.config

	      cmhaltnode node1

       Modify the ASCII file to delete the node information accordingly

	      cmapplyconf -C clusterA.config

       Apply  the  configuration while restricting the connective check to the
       volume groups specified in the clusterA.config file  and	 marking  them
       cluster aware (if not already),

	      cmapplyconf -k -C clusterA.config

       Apply  the  configuration while restricting the connective check to the
       cluster lock volume groups and  marking	them  cluster  aware  (if  not
       already),

	      cmapplyconf -K -C clusterA.config

       To  add	a new heartbeat network to the existing cluster configuration,
       use the following command:

	      cmquerycl -c clusterA -C clusterA.config

       Modify the ASCII file to uncomment the new HEARTBEAT_IP

	      cmapplyconf -C clusterA.config

       To remove a network from the existing cluster configuration, use either
       of the following commands to get the current cluster configuration:

	      cmquerycl -c clusterA -C clusterA.config

	      or cmgetconf -c clusterA -C clusterA.config

       Modify the ASCII file to delete the network entries

	      cmapplyconf -C clusterA.config

       To  verify  and	distribute modification to package pkg1 from a running
       cluster, with pkg1.config containing the changes to the package,	 pkg1,
       use the following command:

	      cmapplyconf -f -P pkg1.config

       To start that package, run

	      cmmodpkg -e pkg1

LIMITATIONS
       Configurations for packages of SYSTEM_MULTI_NODE type cannot be applied
       at the same time as the cluster configuration and/or any other  package
       configurations.	 Therefore,  when applying configurations for packages
       of SYSTEM_MULTI_NODE type, the following limitations apply:
	      The option may not be specified.
	      If using the option, only configuration files  for  packages  of
		     SYSTEM_MULTI_NODE type may be specified.
	      If  using	 the  option, only configuration files for packages of
		     SYSTEM_MULTI_NODE type may be specified  in  the  package
		     reference file.

       Under  Serviceguard Extension for RAC (HP-UX only), cmapplyconf returns
       an error if you try to add nodes to or delete nodes  from  the  cluster
       while  there are any SLVM volume groups activated in shared mode. Deac‐
       tivate all shared SLVM volume groups before running cmapplyconf to  add
       new cluster nodes or delete existing nodes from the cluster.

AUTHOR
       cmapplyconf was developed by HP.

SEE ALSO
       cmcheckconf(1m),	    cmgetconf(1m),    cmmakepkg(1m),	cmquerycl(1m),
       cmruncl(1m).  cmhaltcl(1m), cmrunnode(1m).  cmhaltnode(1m).

		    Requires Optional Serviceguard Software    cmapplyconf(1m)
[top]

List of man pages available for HP-UX

Copyright (c) for man pages and the logo by the respective OS vendor.

For those who want to learn more, the polarhome community provides shell access and support.

[legal] [privacy] [GNU] [policy] [cookies] [netiquette] [sponsors] [FAQ]
Tweet
Polarhome, production since 1999.
Member of Polarhome portal.
Based on Fawad Halim's script.
....................................................................
Vote for polarhome
Free Shell Accounts :: the biggest list on the net