Choosing a lifecycle option for a pattern instance

From IBM® MQ Virtual System Pattern Type 1.0.0.6, you can choose to manage your IBM MQ log and queue manager data independently from the deployed pattern instance.

About this task

When you create patterns and deploy pattern instances, you can plan the lifecycle of your deployed instance, data management, migration, and reuse of patterns. The following two configuration options are provided to help you to customize your IBM MQ pattern creation and instance deployment:

Manage IBM MQ data manually configuration option: You can specify a location for your IBM MQ data directory, different from the default one, to prevent your IBM MQ data from being deleted when the deployed instance is deleted. If you do not select the option to manage the IBM MQ data directory manually, the deployment id is used in the IBM MQ data path at the time of instance creation. When the instance is deleted, the IBM MQ data is deleted too.
Use existing IBM MQ queue manager data directory configuration option: You can deploy an instance from a pattern and use data that exists from a previously deleted or stopped instance to re-create and resume the IBM MQ queue manager.

Both the Manage IBM MQ data manually and the Use existing IBM MQ queue manager data directory options are available with single instance, High availability active instance, and High availability standby instance pattern configurations.

The use of the two new configuration options is described in the following instance lifecycle settings:

Procedure

Default
If you do not select the Manage IBM MQ data manually check box when the pattern is created, the deployment id is included in the IBM MQ log and data path. When a pattern instance that is deployed from this pattern is deleted, the IBM MQ queue manager logs and data that are associated with the pattern instance, are removed. You can use this pattern configuration to deploy long running pattern instances when you want to manage IBM MQ versions through the IBM MQ and IBM PureApplication® System maintenance framework. IBM MQ networks with common configurations that are pre-defined in this way, can be deployed quickly. You cannot migrate data or modify a pattern instance that is created with default lifecycle settings, to manual or resume lifecycle settings.
Manual
Select the Manage IBM MQ data manually check box when you create the pattern. The deployment id is not included in the IBM MQ log and data path, so when a pattern instance that is deployed from this pattern is deleted, the IBM MQ queue manager logs and data that is associated with the pattern instance, remain in the location that you specify. The different location that you choose manually must be on the shared General Parallel File System server. You can use this pattern configuration when the IBM MQ log and data directory path locations must be defined before the pattern instance is deployed, for example, when the directories are registered to an external monitoring service, and when the IBM MQ log and data directories must not be deleted with the pattern instance. If you know that you might want to use the IBM MQ log and queue manager data from an instance you are deploying now, in subsequent deployments, you must use the Manage IBM MQ data manually option. This is a prerequisite for selecting the Use existing IBM MQ queue manager data directory option in subsequent deployments.
You are responsible for avoiding IBM MQ queue manager naming collisions in GPFS shared locations. However, if you try to reuse a name of an existing IBM MQ queue manager data directory location for a new instance, the crtmqm command transforms the final directory location to establish a unique path. You cannot migrate data or modify a pattern instance that is created with manual lifecycle settings, to default lifecycle settings.

Note: Consider the impact that any scripts that you create might have on initial and subsequent pattern instance deployments. For example, adding a simple script to create a queue during instance deployment might result in an error in the case where a subsequent deployment uses existing data to resume the queue manager and the queue is already defined. Either modify or remove the script to avoid the error.
Resume
Select the Manage IBM MQ data manually check box and use the Use existing IBM MQ queue manager data directory field. You can use this configuration when you want to re-create and resume an IBM MQ queue manager by using data that exists from a previous instance.
Note: The previous instance must be created with the Use existing IBM MQ queue manager data directory field left empty. If the field is not empty, it is evaluated, along with the data path or GPFS shared location, during the deployment to form the final directory for existing IBM MQ queue manager data, for example /<data_path>/qmgrs/<Use existing MQ qm data directory>. If the directory location is not found, the deployment fails.
This field is only valid for deployments that use GPFS shared storage for the IBM MQ queue manager log and data location because directory paths on storage that are tightly bound to the host virtual machine, are removed when the previous pattern instance is deleted. This field is set to the previous IBM MQ queue manager name, for example QM1. However, where multiple IBM MQ queue managers with the same name share a path, the final directory is transformed to establish a unique path, for example QM1.000.