Conclusions

HP has successfully shown the viability of the Extended Cluster for RAC in a virtualized environment over distance. In particular, HP demonstrated the feasibility of a SGeRAC configuration deployed across two discrete data centers separated by 100 km. The 100-km distance significantly diminishes the likelihood of a single disaster impacting more than one data center at any one time.

Traditionally, disaster tolerance in an IT environment had been accomplished through a configuration of “active/standby”, that is, resources are not utilized until needed. The 100-km RAC solution destroys this paradigm—there is now full utilization of all resources in all data centers at all times, that is, an “active/active” environment. Data is replicated between the two data centers and is fully synchronized, allowing location-independent access to the applications. The HP VSE permits both data centers to see the data simultaneously and to leverage the aggregated resources from both locations continuously. The HP VSE is the only solution available on the market today that tightly integrates high availability, partitioning, and utility pricing.

HP is thus able to deliver a highly reliable solution incorporating SGeRAC, the VSE, BEA WebLogic Server, and Oracle9i RAC. The refinement of VSE technologies makes possible the full utilization of IT resources beyond 100 km as well.3 The leadership of HP in virtualization technologies ensures that full resource utilization will soon become mainstream, regardless of the distance separating data centers.

HP has thus made extending a RAC solution a reality without sacrificing resource efficiency.

References

DWDM: A White Paper by Joseph Algieri and Xavier Dahan, available on request from HP

3HP Continentalclusters for RAC is available for customers requiring a solution that extends beyond 100 km and incorporates the features of extended RAC across two data centers located anywhere in the world.

13