AIX > Administrator > High Availability

Restrictions and Considerations of PowerHA SystemMirror Enterprise Edition HyperSwap

Editor’s note: This is the final article in a three-part series about PowerHA SystemMirror for AIX Enterprise Edition. The first part, introduced the concept and benefits of HyperSwap with PowerHA SystemMirror. The second article, outlines how to integrate HyperSwap with a PowerHA SystemMirror Enterprise Edition cluster.

Having walked through the overview, benefits and configuration steps of HyperSwap, we’ll conclude with some considerations and its restrictions.

While planning for HyperSwap support with PowerHA SystemMirror, keep in mind some generic, performance-related and implementation considerations. These will play a major role in helping plan for a robust and redundant disaster recovery solution with PowerHA providing support for storage maintenance without interrupting application service. Consider these factors carefully:

  • Concurrent workloads across sites, such as Oracle RAC, are not supported.
  • SCSI reservations aren’t supported in mirror groups that use HyperSwap disks.
  • In order to use Live Partition Mobility (LPM), HyperSwap must be disabled for all online mirror groups. Once migration is complete, enable HyperSwap.
  • Swap time must be calculated. This is the amount of I/O delay time, in seconds, that PowerHA causes while performing a HyperSwap operation on a mirror group.

The swap time helps plan for and understand how much time it will take for HyperSwap operation to finish. The swap timeout value is specific for each mirror group in a cluster. The value varies for planned and unplanned HyperSwap. The swap timeout for planned HyperSwap is 120 seconds and cannot be changed. The timeout for unplanned HyperSwap is between zero and 180 seconds. The swap time for unplanned HyperSwap depends on various factors, including:

  • The number of nodes where the application is hosted; more nodes means more information being shared
  • Network latency and application network usage
  • The number of disks used by the application
  • The I/O response-time requirement for the application.

Restrictions and Limitations

The initial release of HyperSwap support with PowerHA SystemMirror has certain restrictions and limitations. For example, vSCSI disks aren’t supported, and it’s only available for IBM DS8800 or higher storage subsystems. Also, support for LPM is limited to base AIX Device Driver support.

In addition, automatic resynchronization isn’t supported for HyperSwap-enabled configurations. Instead, users must manually resume replication after a replication link heals. For IBM DS8800 in-band metro mirror peer-to-peer remote copy (PPRC) resources, automatic resynchronization is done through a SystemMirror join cleanup event. Other limitations include:

  • Disk replication relationships must adhere to a 1-to-1 relationship between the underlying logical subsystems.
  • The HyperSwap enable/disable operation on a rootvg mirror group applies to all nodes of the cluster.
  • The PowerHA rediscovery utility must be run after any storage-level PPRC configuration changes—such as adding, removing or changing new PPRC paths.
  • Enabling HyperSwap for a cluster repository disk requires an alternate disk to be specified.
  • Workloads that concurrently access the same, primary PPRC disk, but span across multiple sites, is not supported.
  • Active/Active workloads that perform I/O to both PPRC volumes (primary and secondary) are not supported.

Superior Solution

Despite these restrictions, HyperSwap support with PowerHA SystemMirror offers a compelling solution providing continuous availability against storage errors. It allows seamless and transparent switch over of application I/O to the secondary storage without affecting the application user or connectivity. Ultimately, the solution enables storage maintenance without application downtime. The advantages and benefits offered by HyperSwap support for PowerHA clearly outweigh the current limitations and restrictions.

Kunal Langer is a technical solutions architect, Power Systems, AIX and PowerHA, with the Systems and Technology Group Lab-Based Services and Training.

comments powered by Disqus



2019 Solutions Edition

A Comprehensive Online Buyer's Guide to Solutions, Services and Education.

IBM Systems Magazine Subscribe Box Read Now Link Subscribe Now Link iPad App Google Play Store
IBMi News Sign Up Today! Past News Letters