ActioNetAgile: EXADATA Migration

One of our Federal Agency customers has a large set of Oracle database(s), with over 300 Terabytes of data. The central database(s) replicated data from a decentralized network of smaller database(s), spread all over the world. The near real-time replication of data between central and remote database(s) was built on top of Oracle product suite. The system is very complex and mission-critical. It exchanges data with other Federal Agencies and is directly tied to our country’s National Security apparatus. Any disruption to its functioning will have ripple effects reverberating throughout the world. ActioNet is responsible for operation and maintenance of this central and distributed oracle databases(s), around the clock.

The Project

The Oracle databases system described above needed to operate efficiently and with the least number of disruptions. However, due to archaic hardware and earlier versions of Oracle, the system suffered frequent outages and capacity issues. The system had to be upgraded to the next generation of Oracle’s engineered systems, EXADATA. The project involved migration of data from the existing Oracle 10g database(s) into EXADATA, running Oracle 11g. In addition, the migrated EXADATA database(s) had to be placed in another Data Center (Destination), situated geographically several hundred miles away from the existing Data Center (Source).

The Solution

ActioNet teamed with Oracle and other technology partner(s), as well as other vendors to establish a set of use-case scenarios and identified the most appropriate scenario to implement. Because there are several database(s), the solution to each scenario had to be adjusted slightly based on several factors, such as current patch level, data size, etc. The most relevant scenario and solution are depicted above.

ActioNet worked with the customer and partner(s) to establish a detailed migration plan, including specific steps and expected outcome(s). One of the partner(s) conducted independent verification of deployment steps (IV&V). The IV&V team’s feedback was incorporated into the migration plan. In order to minimize risk during migration, we selected the smallest database with the least number of external dependencies to conduct the first migration. We incorporated lessons-learned during this migration into the plan. We progressively handled database(s) of larger size and complexity, completing all migrations per customer timelines.

Another huge logistical challenge was the actual duration of migration. Due to the large data-set size, copying data from one location to another took a long time. In addition, the outage window had to be timed around Federal Holiday week-end(s), to cause the least amount of disruption to operations. The project culminated in a huge success when we migrated the most complex database with over 42 TB of data as per the plan and with no issues.