lodemark
lodemark
All case studiesCase study · Cloud migration

Out of the data center, onto AWS -
in two months, with zero downtime

How we migrated a production platform from its on-premises data center to AWS - on a deadline we didn't choose, with almost no internal knowledge to lean on, and without users noticing a thing.

AWSCloud MigrationZero Downtime
Technologies
AWSEC2RDSS3Route 53MySQLPostgreSQLCassandraMongoDBDocumentDB
2
months from audit to full migration
<5
minutes of downtime - DNS propagation only
40
microservices moved in priority order
4
databases migrated and verified
The challenge

A forced exit, on someone else's clock

Our client's entire platform - around 40 microservices and four databases - ran in a single on-premises data center. Then the operator terminated the contract. The deadline wasn't negotiable: take your infrastructure and leave.

The harder problem was what wasn't there. More than 90% of the engineers who had built and operated these systems had already left the company, taking most of the internal knowledge with them. Documentation was thin, and the business data living in those systems - spread across MySQL, PostgreSQL, Cassandra, and MongoDB - was an asset the company could not afford to lose.

What we did

Map everything. Then move it, piece by piece.

We started where every migration should: understanding what actually existed, and - most importantly - where the data lived.

Audited and mapped every on-premises system - what it did, what it depended on, and how critical it was
Located every data store across MySQL, PostgreSQL, Cassandra, and MongoDB
Set up AWS accounts, prioritized the 40 microservices, and moved them one by one
Scheduled all high-risk cutovers in the lowest-traffic windows for the US-based user base
Migrated every database with full data verification before each switchover - the only downtime was DNS cutover, under five minutes

From there, we moved the platform to AWS service by service, planning each cutover so even a worst-case failure would touch as few users as possible.

The result

Out in two months. Nobody noticed.

Two months in, all 40 microservices were out of the on-premises data center, and the databases and backups were running on AWS - including managed RDS and DocumentDB. There was no downtime beyond DNS propagation, no revenue impact, and no support tickets asking what happened - because from the users' perspective, nothing did. The business data that couldn't be lost, wasn't.

"We lost our data center and most of the team that built it. Lodemark got us onto AWS in two months, and our users never knew it happened."

CTO, Health & Fitness client
Ship without fear.
Scale without chaos.

Tell us about your infrastructure - we'll reply within one business day.