High availability, backup, resiliency, and recovery enhancements
Db2 12.1.2 includes enhancements to high availability and cluster management with Pacemaker. These changes improve the performance of both Db2 pureScale and Data Partitioning Feature (DPF)-configurations.
- Db2 Distributed (on-premises)
- The container micro-service deployments of Db2 and Db2 Warehouse for Kubernetes-based clusters, which includes Red Hat OpenShift, the Amazon Elastic Kubernetes Service (EKS), the Red Hat OpenShift Service on AWS (ROSA), and the Azure Kubernetes Service (AKS).
Enhancement | Description |
---|---|
CPU scheduling for Pacemaker | Db2
clustered environments that use Pacemaker automation now enable CPU scheduling on the system to
protect critical HA resources during high CPU load times. This new feature ensures that Pacemaker
gets CPU time as needed and prevents false fail overs and time outs. Affected Pacemaker-managed
environments include the following highly available configurations:
|
Update the db2adutl utility to display the size of the backup image or archived log file in IBM Spectrum Protect. | Users of IBM Spectrum Protect [formerly Tivoli Storage Manager (TSM)] can now use the db2adutl utility to query their shared remote file storage for the size of backup images, load copy images, and log files. |
UPGRADE DATABASE now works with topology change | UPGRADE DATABASE operations no longer require that a source Db2 11.5.x and earlier database and a target Db2 12.1.2 databases have the same topology. For more information, see Upgrading a Db2 server and database with topology change. |
Pacemaker software refresh | Pacemaker is upgraded from version 2.1.7 to 2.1.9. For more information, see Supported software and hardware for Pacemaker. This new version is used by all integrated solutions that use Pacemaker. |
Db2 Database Partitioning Feature (DPF) enhancements
Enhancement | Description |
---|---|
Pacemaker DPF resiliency improvement | With the release of Db2 12.1.2, resilience to failures is enhanced for HA-enabled DPF clusters, and the recovery process from complete cluster outage scenarios is automated. |
New db2cm -verify
option for checking operating system level and prerequisite packages of DPF
clusters |
After either updating their DPF-configured Db2 cluster or making
changes to installed packages, users can now run the following command to verify that prerequisites
are met by all hosts in the cluster or a subset of
hosts: For more
information, see db2cm command reference for DPF-configured clusters. |
Pacemaker software refresh | Pacemaker is upgraded from version 2.1.7 to 2.1.9. For more information, see Supported software and hardware for Pacemaker. This new version is used by all integrated solutions that use Pacemaker. |
Db2 pureScale enhancements
Enhancement | Description |
---|---|
Db2 pureScale on AWS now supports RDMA | With the release of 12.1.2, you can now deploy a self-managed, RDMA configuration of Db2 pureScale directly from the Amazon Web Services (AWS) Marketplace. For more information, see Db2 pureScale on Amazon Web Services. |
Enhanced FORCE_ALL option in the ADMIN_MOVE_TABLE stored procedure. |
In a Db2 pureScale® environment, running the ADMIN_MOVE_TABLE stored procedure under a heavy workload can cause the SWAP phase to fail. Failure is caused by deadlocks or timeouts that can happen, even when the FORCE_ALL option is enabled. These failures are most common when numerous blocking applications are active on the source table . With the enhanced functionality of FORCE_ALL, the ADMIN_MOVE_TABLE stored procedure can now terminate a broader range of blocking applications, increasing the likelihood of successful completion under high-contention scenarios. Examples of newly handled blocking scenarios include:
|
Pacemaker software refresh | Pacemaker is upgraded from version 2.1.7 to 2.1.9. For more information, see Supported software and hardware for Pacemaker. This new version is used by all integrated solutions that use Pacemaker. |