Ingres Zero Downtime Migration

Perform Migrations with no Downtime

Ingres HVR offers zero downtime when maintenance is performed on a production system. This may be a migration, or the entire replacement of an application with new technology and typically involves several analysts and programmers.

Avoid Risky "Big-Bang" Migrations

Without Ingres HVR a system migration would involve a "big bang" during which the system is offline for several days, and is then error-prone in the following weeks.

Phased and Planned Approach

Using Ingres HVR a zero-downtime migration can be planned in phases. First analysts must define the logic needed to convert between the two systems. This is injected inside the Ingres HVR migration channel which is able to replicate between the old and the new systems. Once this channel is running the production can be moved module-by-module from the legacy system to the new system. For example when the reporting module is ready it can be tried against live data piped through Ingres HVR, and when it is fully tested reporting can be switched to production on the new system.

Ingres HVR Uniquely Suited to Migrations

Ingres HVR is uniquely suited to large and complicated migration, because;

• Ingres HVR can replicate between different DBMS types, such as Oracle, Microsoft SQL Server and Ingres.
• Ingres HVR can pipe changes into a XML based web service for more mapping flexibility.
• A simple one-to-one replication can be setup in seconds, mappings with medium complexity can be configured using drag-and-drop in the Ingres HVR GUI, and very complex mappings can be encoded in SQL or a special 'agent' and then injected into the replication pipeline.
• Ingres HVR can be replicated bi-directionally (both directions) and can detect and resolve collisions if the same object is changed simultaneously.