Configuring geographically dispersed DB2 pureScale clusters

Because disasters like power failures and fires might easily disable a single data center, many large IT organizations configure two sites, far enough apart to be on separate power grids. This configuration minimizes the risk of total outage, and allows business to carry on at one site, even if the other is impacted by a disaster.

Jeremy Brumer (, DB2 pureScale System Verification Test, IBM

Jeremy Brumer is the team lead for the DB2 pureScale System Verification Test organization, and is based at the IBM Canada Lab. He has over 10 years experience with DB2 products, and has been focused for the last several years on high availability solutions, including HADR, Microsoft Failover Clustering, and DB2 pureScale.

Roy Cecil (, DB2 Performance Team, IBM

Roy Cecil photoRoy Cecil has 10 years of experience in Software Development and Software and Systems Performance Engineering. He has been working since 2008 on DB2 V9.8 performance improvement, and most recently on improving the DB2 Recovery Performance. Roy is also an expert in UNIX and database systems and is based out of the Dublin Software Labs in Ireland. .

Fabrizio Fabbri (, DB2 Software Development Team, IBM

Fabrizio Fabbri photoFabrizio Fabbri is a software developer and works with the DB2 kernel development team. He is currently working with the DB2 pureScale development team on cluster management and shared file system integration.

Steve Rees (, Senior Performance Manager, IBM

Steve Rees photoSteve Rees is Senior Performance Manager for DB2 pureScale software. He has worked on DB2 performance for over 15 years, with the last three on DB2 pureScale software.

21 April 2011

This article describes the geographically dispersed DB2® pureScale™ cluster (GDPC). Like the Geographically Dispersed Parallel Sysplex™ configuration of DB2 for z/OS®, GDPC provides the scalability and application transparency of a regular single-site DB2 pureScale cluster, but in a cross-site configuration that enables active/active system availability, even in the face of many types of disasters.

The active/active part is important because it means that during normal operation, the DB2 pureScale members at both sites are sharing the workload between them as usual, with workload balancing (WLB) maintaining an optimal level of activity on all members, both within and between sites. This means that the second site is not a standby site, waiting for something to go wrong. Instead, the second site is pulling its weight, returning value for investment even during day-to-day operation.

This article describes the prerequisites for a geographically dispersed DB2 pureScale cluster, followed by the steps to one deploy one, as well as some of the performance implications of different site-to-site distances and different workload types.

The article covers the following topics:

  • GDPC concepts
  • GDPC infrastructure and prerequisites
  • GDPC setup and configuration
  • Performance factors
  • Detailed configuration steps


Article in PDF formatDBP_pureScale_GDPC_0419.pdf820KB



Get products and technologies


Zone=Information Management
SummaryTitle=Configuring geographically dispersed DB2 pureScale clusters