Create DR/HA RDQMs

You use the crtmqm command to create a replicated data queue manager (RDQM) in a DR/HA configuration.

About this task

We can create a DR/HA RDQM as a user in the mqm group if the user can use sudo. Otherwise create the RDQM as root.

We must create a number of DR/HA RDQMs:

• On the HA group on the 'main' site:
  • On the node where we want the queue manager to run under normal conditions, create the primary/primary DR/HA RDQM.
  • On each of the other two nodes in the HA group, create a primary/secondary DR/HA RDQM.

• On the HA group on the 'recovery' site:
  • On the node where the queue manager will run if it fails over to the recovery site, create the secondary/primary DR/HA RDQM. We can use the command output when you created the primary/primary queue manager on the 'main' site.
  • On each of the other two nodes in the HA group, create a secondary/secondary DR/HA RDQM.

All of the queue manager instances must have the same name and be allocated the same amount of storage.

Procedure

• To create the primary/primary DR/HA RDQM:
  1. Enter the following command:

     crtmqm -sx -rr p
                   [-rl DRLocalIP1,DRLocalIP2,DRLocalIP3]
                   (-ri DRRemoteIP1,DRRemoteIP2,DRRemoteIP3 | -rn GroupName)
                   -rp DRPort
                   [-z] [-q] [-c Text] [-d DefXmitQ] [-h MaxHandles]
                   [-g ApplicationGroup] [-oa user|group]
                   [-t TrigInt] [-u DeadQ] [-x MaxUMsgs]
                   [-lp LogPri] [-ls LogSec]
                   [-lc | -ll | -lla | -lln] [-lf LogFileSize]
                   [-p Port] [-fs FilesystemSize] QMgrName
     

     Where:

     -sx

     Indicates that the initial HA role is primary.

     -rr p

     Indicates that the initial DR role is primary.

     -rl DRLocalIP1, DRLocalIP2, DRLocalIP3

     Optionally specify the IP addresses of the DR interfaces on the three nodes on the local site (that is, the 'main' site). If not specified, the IP addresses specified in the rdqm.ini file are used.

     -ri DRRemoteIP1, DRRemoteIP2, DRRemoteIP3

     Specify IP addresses of the DR interfaces on the three nodes on the remote site (that is, the 'recovery' site). We must specify either this or the -rn parameter.

     -rn GroupName

     Specify the remote HA group name as specified in the rdqm.ini file. You must specify either -ri or -rn.

     -rp Port

     Specifies the port to use for DR replication.

     other_crtmqm_options

     We can optionally specify one or more of these general crtmqm options:

     • -z
     • -q
     • -c Text
     • -d DefaultTransmissionQueue
     • -h MaxHandles
     • -g ApplicationGroup
     • -oa user | group
     • -t TrigInt
     • -u DeadQ
     • -x MaxUMsgs
     • -lp LogPri
     • -ls LogSec
     • -lc | -l
     • -lla | -lln
     • -lf LogFileSize
     • -p Port

     -fs size

     Optionally specifies the size of the filesystem to create for the queue manager, that is, the size of the logical volume that is created in the drbdpool volume group. Another logical volume of that size is also created, to support the reverting to snapshot operation, so the total storage for the DR RDQM is just over twice that specified here.

     QMname

     Specifies the name of the replicated data queue manager. The name is case sensitive.

     After the command completes, it outputs the command that we can input on the recovery site to create the secondary/primary instance of the queue manager.

• To create a primary/secondary DR/HA RDQM on the other two nodes in the HA group:
  1. Enter the following command on each node:

     crtmqm -sxs -rr p
                   [-rl DRLocalIP1,DRLocalIP2,DRLocalIP3]
                   (-ri DRRemoteIP1,DRRemoteIP2,DRRemoteIP3 | -rn GroupName)
                   -rp DRPort
                   [-fs FilesystemSize] QMgrName

     Where:

     -sxs

     Indicates that the initial HA role is secondary.

     -rr p

     Indicates that the initial DR role is primary.

     -rl DRLocalIP1, DRLocalIP2, DRLocalIP3

     Optionally specify the IP addresses of the DR interfaces on the three nodes on the local site (that is, the 'main' site). If not specified, the IP addresses specified in the rdqm.ini file are used.

     -ri DRRemoteIP1, DRRemoteIP2, DRRemoteIP3

     Specify IP addresses of the DR interfaces on the three nodes on the remote site (that is, the 'recovery' site). We must specify either this or the -rn parameter.

     -rn GroupName

     Specify the remote HA group name as specified in the rdqm.ini file. You must specify either -ri or -rn.

     -rp Port

     Specifies the port to use for DR replication.

     -fs size

     Specifies the size of the filesystem to create for the queue manager, that is, the size of the logical volume that is created in the drbdpool volume group. If we have specified a non-default size when creating the primary/primary RDQM, we must specify the same value here.

     QMname

     Specifies the name of the primary/secondary RDQM. This must be the same as the name you specified for the primary/primary instance of the RDQM. Note that the name is case sensitive.

• To create a secondary/primary DR/HA RDQM on the node where the queue manager will run if it fails over to the recovery site:
  1. Use the command output when you created the primary/primary DR/HA on the main site, or enter the following command:

     crtmqm -sx -rr s
                   [-rl DRLocalIP1,DRLocalIP2,DRLocalIP3]
                   (-ri DRRemoteIP1,DRRemoteIP2,DRRemoteIP3 | -rn GroupName)
                   -rp DRPort
                   [-fs FilesystemSize] QMgrName
     

     -sx

     Indicates that the initial HA role is primary.

     -rr s

     Indicates that the initial DR role is secondary.

     -rl DRLocalIP1, DRLocalIP2, DRLocalIP3

     Optionally specify the IP addresses of the DR interfaces on the three nodes on the local site (that is, the 'recovery' site). If not specified, the IP addresses specified in the rdqm.ini file are used.

     -ri DRRemoteIP1, DRRemoteIP2, DRRemoteIP3

     Specify IP addresses of the DR interfaces on the three nodes on the remote site (that is, the 'main' site). We must specify either this or the -rn parameter.

     -rn GroupName

     Specify the remote HA group name as specified in the rdqm.ini file. You must specify either -ri or -rn.

     -rp Port

     Specifies the port to use for DR replication.

     -fs size

     Optionally specifies the size of the filesystem to create for the queue manager, that is, the size of the logical volume that is created in the drbdpool volume group. Another logical volume of that size is also created, to support the reverting to snapshot operation, so the total storage for the DR RDQM is just over twice that specified here.

     QMname

     Specifies the name of the replicated data queue manager. The name is case sensitive.

• To create a secondary/secondary HA/DR RDQM on the other two nodes on the recovery site:
  1. Enter the following command on each node:

     crtmqm -sxs -rr s
                   [-rl DRLocalIP1,DRLocalIP2,DRLocalIP3]
                   (-ri DRRemoteIP1,DRRemoteIP2,DRRemoteIP3 | -rn GroupName)
                   -rp DRPort
                   [-fs FilesystemSize] QMgrName
     

     -sxs

     Indicates that the initial HA role is primary.

     -rr s

     Indicates that the initial DR role is secondary.

     -rl DRLocalIP1, DRLocalIP2, DRLocalIP3

     Optionally specify the IP addresses of the DR interfaces on the three nodes on the local site. If not specified, the IP addresses specified in the rdqm.ini file are used.

     -ri DRRemoteIP1, DRRemoteIP2, DRRemoteIP3

     Specify IP addresses of the DR interfaces on the three nodes on the remote site. We must specify either this or the -rn parameter.

     -rn GroupName

     Specify the remote HA group name as specified in the rdqm.ini file. You must specify either -ri or -rn.

     -rp Port

     Specifies the port to use for DR replication.

     -fs size

     Optionally specifies the size of the filesystem to create for the queue manager, that is, the size of the logical volume that is created in the drbdpool volume group. Another logical volume of that size is also created, to support the reverting to snapshot operation, so the total storage for the DR RDQM is just over twice that specified here.

     QMname

     Specifies the name of the replicated data queue manager. The name is case sensitive.

What to do next

After you have created all the DR/HA RDQMs, we must check the status on the primary/primary and secondary/primary instances to check all are correct. Use the rdqmstatus command on the nodes. The nodes should be displaying normal status as described in View DR/HA RDQM and HA group status. If they are not displaying this status, delete the secondary/primary instance and recreate it, taking care to use the correct arguments.

• Delete a DR/HA RDQM
  You use the dltmqm command to delete a DR/HA replicated data queue manager (RDQM).

Parent topic: RDQM disaster recovery and high availability

Related tasks

• Create DR/HA RDQMs

Related information

• crtmqm