Fast-growing enterprise
A fast-growing enterprise had to move a mission-critical application to a new AWS account with no downtime and no data loss. We copied its servers and databases across, tested them, then switched traffic.
The facts
The result
Zero downtimeNo downtime and no data loss during the move
Six steps from one account to the other
The order matters: nothing moved until the copy had been tested.
- Step 1: CopyImages of the servers and snapshots of the databases.
- Step 2: ShareBoth shared with the new account through IAM roles and policies.
- Step 3: RebuildMatching servers, with the same security groups and network settings.
- Step 4: RestoreDatabases restored from the snapshots, with settings aligned.
- Step 5: TestLoad, data and function tested before any traffic moved.
- Step 6: SwitchDNS and routing pointed at the new account, and the old one cleaned up.
The challenge
- A sensitive application that could not stop.
- Identical settings, security and access in the new account.
- The risk of lost data or a broken service.
What we did
- Planned the move so nothing stopped and nothing was lost.
- Rebuilt every server and database in the new account exactly as it was.
- Tested before switching, then removed what the old account no longer needed.
How it was built
- Compute
- Amazon EC2, rebuilt from machine images
- Database
- Amazon RDS, restored from snapshots
- Access
- AWS IAM roles and policies
- Network
- Security groups
- Key pairs
- DNS records
What came of it
- EnvironmentThe application now runs in a more optimized AWS environment that fits its needs.
- Cost and securityMeasures taken during the move improved performance and cost efficiency.