What’s New with Db2 Mirror, PowerHA, and BRMS
November 8, 2023 Alex Woodie
There’s something for everyone in the new releases of IBM i that IBM will be shipping later this month. For administrators who revel in having well-run high availability, continuous availability, and tape backups, the new releases will not disappoint.
The most substantial enhancements appear to have been saved for Db2 Mirror for i, which is IBM’s continuous availability solution for IBM i. The new management and monitoring features should bring some major improvements to how customers interact with the product.
Some of the biggest changes are due to the addition of the new “Health Center” component to the Db2 Mirror GUI. According to IBM, the Health Center will monitor critical and product-related resources on the primary server, including SYSBAS, all database IASPs that are active, and network lines.
Customers can set the critical resource check interval to anything between five and 60 seconds. The checks include determining the status of the RDMA communication links between the nodes; checking the status of disk unit connectivity; and checking storage availability.
In the event of a communication outage, the Health Center will suspend replication as an unplanned outage, after a predetermined pause is completed, which can be anything between 15 seconds and an hour, according to IBM. Unplanned outages are eligible for automatic takeover, IBM says. When the storage threshold check is triggered, replication is suspended, but it’s not treated as an unplanned outage.
Health Center also includes conducts product resource checks, such as Db2 Mirror jobs, the expiration of product license keys, or other less-critical. If an issue is detected with Db2 Mirror product resources, a message is sent to the QSYSOPR queue. Health Center may also attempt to automatically recover a Db2 Mirror job if an issue is detected. You can read more about Health Center here.
IBM also improved how Db2 Mirror responds to jobs that prevent the successful vary off of a Db2 Mirror IASP, which could happen if a job is held, executing a long running commit or rollback, or running a very long SQL job.
With the new release, when a job is preventing the IASP from varying off, the administrator can now view the list of offending jobs by querying the MIRROR_SUSPENDING_JOBS table function. Once she knows which jobs are the cause of the holdup, the administrator can take the appropriate action.
Speaking of SQL jobs for monitoring Db2 Mirror, IBM has also delivered a new scalar function in QSYS2 for comparing Db2 Mirror versions a customer has running. The COMPARE_MIRROR_VERSION function checks to see whether the xxx.yyy.zzz version identifiers are the same.
Db2 Mirror administrators will also appreciate the addition to the Db2 Mirror Services SQL function. According to IBM, the table now shows, for each Db2 Mirror service, the Db2 for i PTF Group levels and the earliest possible release where a version of the service is available. The detailed information shown in the QSYS2.SERVICES_INFO table covers the dozens of services contained in the communication, product, reclone, replication, resynchronization, and serviceability categories.
Finally, this release also brings some performance enhancements, specifically in the areas of the replication of some object types and during resynchronization. For more information check out the Db2 Mirror wiki page.
Next up, IBM has also issued some enhancements to IBM PowerHA SystemMirror for i – which, of course, is now developed by Fortra (formerly HelpSystems) on an outsourcing contract with IBM.
The official IBM announcement letter for IBM i 7.5 TR3 is a little bit light on details, including the version and release number. Fortra also hasn’t updated its website with information on the new releases of PowerHA. What we do know is that PowerHA’s has been enhanced with several new job tracking features that should make it easier to get started and recover from an incident.
For starters, IBM has rolled out new user-defined job tracking that makes it easier for users to define which job queues to track. The new release also makes it easier for users to determine what the status of a job is at the time of a failure. The new feature work for both jobs that were active when the failure hit, as well as recently active jobs.
Finally, users are getting a streamlined way of resubmitting any jobs that may have failed when the server unexpectedly went offline. That should pave the way for a quicker system recovery, IBM says.
It’s not clear what the version and release numbers will be associated with these new features, but one thing is clear: it will become available on December 15, whereas the remainder of the enhancements in IBM i 7.5 TR3 and 7.4 TR9 will ship on November 17.
Last but never least is Backup, Recovery and Media Services for i, or BRMS, the popular IBM i backup product used by a substantial portion of the IBM i installed base. IBM has also outsourced to Fortra the development of BRMS.
One of the big new features is the addition of IASP History Synchronization, which IBM says will give users a set of SQL views and procedures to reflect the use of high availability software in the environment. The views and procedures will also allow the BRMS recovery report capability to combine two systems backups into a single recovery report, which will help to preview the prevent duplication of replicated data, while still having it be available for recovery if needed.
BRMS administrators will appreciate a new SQL procedure that shows the BRMS functional authority usage for specified users and BRMS functional areas. Finally, IBM (or Fortra) has delivered an improved user experience when using VTL devices over iSCSI. You can read more about BRMS enhancements here.