A financial services company is using AWS Database Migration Service (AWS OMS) to migrate Its databases from on-premises to AWS. A database…

Target latency is the amount of time that AWS DMS takes to apply changes from the source database

to the target database1. High target latency can indicate performance issues or replication errors in

the AWS DMS task.

One possible cause of high target latency is the lack of a primary key or index in the target database.

A primary key or index helps AWS DMS identify and apply changes to the corresponding rows in the

target database. Without a primary key or index, AWS DMS has to scan the entire table to find the

matching rows, which can increase the target latency and consume more CPU and memory

resources2.

Another possible cause of high target latency is the resource bottlenecks in the replication instance.

The replication instance is the compute resource that runs the AWS DMS task and connects to the

source and target endpoints. If the replication instance is under-provisioned or overloaded, it can

affect the replication performance and cause high target latency. Some factors that can contribute to

resource bottlenecks are insufficient network bandwidth, low disk space, high CPU utilization, or

large transaction sizes3.