Performance Analysis

Performance Analysis

Analysis Shows That Axxana’s Phoenix for Far Sync Significantly Improves Performance

Oracle’s Far Sync feature in Active Data Guard offers the promise of zero data loss and therefore high application availability at any distance.

However, depending on its distance from the primary site, it exacts a performance penalty that some organizations cannot tolerate.

Axxana’s Phoenix for Oracle Far Sync overcomes this issue by allowing organizations to implement the Far Sync instance within the primary data center, while still guaranteeing zero loss of data, even in total data center or regional disasters.

To demonstrate the performance benefits of a Phoenix for Oracle Far Sync configuration, Axxana’s research and development team recently compared differences in latency between a non-Phoenix configuration—where the Far Sync instance is in a third site—and one that uses Axxana’s Phoenix Black Box to house the Far Sync instance at the primary site. The analysis findings were impressive.

As expected, with synchronous replication (i.e., in Maximum Availability mode), network latency significantly affects the application. Even with one millisecond of latency (which may happen within a data center), performance drops noticeably. With the higher latencies that can be regularly experienced between a primary and nearby data center, performance can easily drop by 50 percent or more, making the use of synchronous replication impractical.

However, by using Axxana’s Phoenix for Oracle Far Sync to colocate the Far Sync instance within the same network segment as the primary database, organizations can achieve latencies in the order of tens of microseconds, leading to minimal impact on performance, almost on par with asynchronous Maximum Performance configuration.

Phoenix for Oracle Far Sync also addresses other important drawbacks of a classic Far Sync configuration. By allowing organizations to deploy Far Sync right within their primary data center, Phoenix for Oracle Far Sync eliminates capital and operational costs associated with maintaining a third Far Sync site, reduces communications costs, and ensures zero data loss and the highest availability in disaster scenarios where Oracle Far Sync alone cannot (e.g., rolling disasters and regional disasters).

To find out more about the test environment and Axxana’s results, please see the technical note, Axxana’s Phoenix for Oracle Far Sync: Performance Analysis.