StarQuest: keeping data in motion

 

StarQuest Technical Documents

SQDR Plus IBM i High Availability Role Swap

Last Update: 15 March 2017
Product: SQDR Plus
Version: 4.0 and later
Article ID: SQV00PL030

Abstract

Many IBM i environments use a High Availability (HA) solution such as Vision Solutions MIMIX to maintain a mirror image of the IBM i host system on a backup system.

This technical document describes the recommended procedure to achieve minimal disruption when performing a planned host role swap. It assumes that the SQDR schema, userID's, and exit program are present on both IBM i servers, that the credentials for the SQDR userID's are identical, and ideally the host components are at the same version level.

In the scenario below, we describe the process of reverting back to the production IBM i server after using the backup server for some time, but a similar procedure can be used when moving from production to the backup server, assuming that swap was planned rather than a reaction to an emergency situation.

Solution


PRIOR TO SWITCH:

  1. Make sure that no changes are being made to the source tables.
  2. Perform an snapshot replication of SQDR.SQ_IXLIST (the "primary key" advice table) from the backup IBM i system to the production IBM i system. This insures that index configurations are current and avoids flagging any subscriptions as altered.
  3. Review the IP naming used for the Control Center and Data Replicator Manager connections - if DNS names are being used, then the switch should be transparent. If IP addresses are being used, it be necessary to update the configuration.
  4. In Data Replicator Manager, pause SQDR groups associated with the IBM i source by selecting the Source, right-click and select Pause. Verify that there no active connections to the source.

Perform the following steps In SQDR Control Center:

  1. Quiesce the server to insure all transactions have been "drained" from the original source system by selecting the Agent and choosing Database/Recovery/Quiesce Agent.
  2. Disable the Agent by selecting the Agent and choosing Disable Agent from the Database menu.
  3. Set the Recovery Mode to Warm by selecting the Agent and choosing Database/Recovery/Set Startup Mode...

At this point the host switch occurs. After the host switch:

  1. Confirm that the DNS name now resolves to the desired host system (tracert, jping)
  2. If necessary, update Host Components for the Agent.
  3. Start the Agent in the Control Center and wait for its icon to turn green.
  4. Notify the IBM i administrator that users may resume transactions
  5. In Data Replicator Manager, resume the paused groups by selecting the Source, right-click and select Resume.

After the switch is judged final:

On the backup system, "park" the exit program by issuing the following SQL statement on the host system:

DELETE FROM SQDR.SQ_READERS

The SQDR exit program prevents deletes of unscraped journal receivers. Issuing this SQL statement prevents the accumulation of journal receivers on the backup system, since SQDR Plus will now be using the journal receivers on the production server. Be sure to perform this operation only on the backup server, after the conclusion of the testing,

 

DISCLAIMER

The information in technical documents comes without any warranty or applicability for a specific purpose. The author(s) or distributor(s) will not accept responsibility for any damage incurred directly or indirectly through use of the information contained in these documents. The instructions may need to be modified to be appropriate for the hardware and software that has been installed and configured within a particular organization.  The information in technical documents should be considered only as an example and may include information from various sources, including IBM, Microsoft, and other organizations.