How to Fix Slow AWS DMS Full Load and Growing CDC Lag
A company is migrating their SAP ERP system running on an on-premises IBM Db2 database to Amazon RDS for Db2. The migration must have minimal downtime. The current database size is 2 TB. The network link between on-premises and AWS is 1 Gbps. The migration team plans to use AWS DMS with a full load followed by ongoing replication. During a test migration, they notice that the full load takes longer than expected, causing the change data capture (CDC) lag to grow. They need to adjust the migration approach to meet the downtime window. What should they do?
Quick Answer
The answer is to increase the compute capacity of the DMS replication instance. This directly addresses the root cause of slow AWS DMS full load performance and growing CDC lag because a larger instance provides more CPU and memory, which accelerates the initial data transfer from the 2 TB on-premises Db2 database and prevents the replication engine from falling behind during ongoing changes. On the AWS Certified SAP on AWS Specialty PAS-C01 exam, this scenario tests your understanding that DMS throughput is often bottlenecked by the replication instance size, not the source database, especially when the network link is adequate. A common trap is to assume you must optimize the source database or switch to offline methods like Snowball, but those break the requirement for minimal downtime and ongoing replication. Remember the memory tip: when CDC lag grows, think “compute first”—scale up the DMS instance before touching the source or network.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
Increase the compute capacity of the DMS replication instance
Increasing the compute capacity of the DMS replication instance provides more processing power to handle the full load faster, reducing the time taken and preventing CDC lag from growing. Option A is incorrect because using multiple DMS tasks in parallel can cause conflicts and may not improve performance due to constraints on the source database. Option C is incorrect because an RDS read replica cannot offload the initial load from an on-premises source. Option D is incorrect because using Snowball for offline transfer would interrupt ongoing replication and is not suitable for minimal downtime.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Use multiple DMS tasks in parallel for the full load
Why it's wrong here
Using multiple DMS tasks in parallel can cause conflicts and may not improve performance due to constraints on the source database.
- ✓
Increase the compute capacity of the DMS replication instance
Why this is correct
Increasing the compute capacity of the DMS replication instance provides more processing power to handle the full load faster, reducing the time taken and preventing CDC lag from growing.
- ✗
Create an RDS read replica to offload the initial load
Why it's wrong here
An RDS read replica cannot offload the initial load from an on-premises source.
- ✗
Use AWS Snowball to transfer the initial load and then resume CDC
Why it's wrong here
Using Snowball for offline transfer would interrupt ongoing replication and is not suitable for minimal downtime.
Go deeper
Related to this question
About these practice questions
Courseiva writes every PAS-C01 question from scratch — 1,616 in total, each with an explanation and a wrong-answer breakdown. None are copied from real exams or dumps. Learn why practice questions differ from exam dumps →
Same concept, more angles
1 more way this is tested on PAS-C01
These questions test the same concept from different angles. Work through them to make sure you can recognise it however the exam phrases it.
Variation 1. An organization is migrating their SAP ERP system running on SAP ASE (Sybase) to Amazon RDS for SAP ASE. They have a 5 TB database and a 1 Gbps network. They need to keep the migration window under 6 hours. They plan to use AWS DMS with full load and CDC. During the full load, they notice that the DMS replication instance is using 100% CPU and the full load is progressing slowly. What should they do to meet the migration window?
medium- ✓ A.Increase the instance size of the DMS replication instance
- B.Switch to using AWS Snowball for the initial load
- C.Increase the compute capacity of the source database
- D.Disable CDC and perform only a full load
Why A: Increasing the DMS replication instance size provides more CPU and memory, which directly speeds up the full load process. Option B is incorrect because AWS Snowball is used for offline data transfer and would not help with ongoing replication; it also introduces additional setup and transfer time, making it unsuitable for a <6 hour window. Option C is incorrect because the bottleneck is the DMS replication instance's CPU, not the source database's compute capacity—increasing source resources does not alleviate DMS resource constraints. Option D is incorrect because disabling CDC would cancel ongoing replication, requiring a full reload later, thus increasing downtime beyond the target window.
JA
Written by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
This PAS-C01 practice question is part of Courseiva's free Amazon Web Services certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the PAS-C01 exam.