Traditional approaches for production database migrations don't scale. They require a lot of time, money and effort.
X
Initial data copy for large data sets is slow and fragile - hours to days.
​
Increases chances of last-minute failures.
​
Complicates planning, testing, and prolongs source impact.
X
Incremental sync or CDC has limited throughput and replication delay in 10's of minutes.
​​
Off-the-shelf CDC is difficult to orchestrate and manage.
​
Writing performant and reliable custom ETL takes months.
X
Cutover risks becoming long downtime.
​
Common data integrity checks are weak and superficial.
​
Separate process for backout adds an extra month to planning.
Dsync accelerates timelines and reduces risks for migrations
to MongoDB Atlas from Cosmos DB and DynamoDB
with one single command
dsync -s $SOURCE -d $ATLAS
01
Parallelized and resumable Initial Data Copy
➤ 10x faster than other tools on the market or DIY scripts
➤ Dsync is secure and doesn't persist any transient data​
➤ Easily track progress using Command-Line Interface or Web UI
02
Real-time incremental sync and CDC
➤ Optimized performance of 1000+ ops/sec for active workloads
➤ Automatically resumes from the last checkpoint on restart
➤ Seamless transition from Initial Data Copy
➤ No code changes needed for deletes in Cosmos DB
03
Cutover
➤ No downtime out-of-the-box
➤ Embedded data integrity checks
➤ Simple backout by reversing direction
