Comparing database migration tools in AWS, GCP, and Yandex Cloud

Moving to the cloud or changing providers is always open-heart surgery. An engineer has to solve a lot of issues: how to move terabytes of data with minimal downtime? How to convert the schema? How to set up CDC (Change Data Capture) so as not to lose a single transaction during the switch?

Each major cloud provider has its own answer. Approaches are radically different: some provide just a “pipe” for data, some — a full ETL combine, some make migration serverless and as trouble-free as possible.

We examine the “native” tools of the three giants:
AWS Database Migration Service (DMS), Google Cloud Database Migration Service, Yandex Data Transfer.

AWS Database Migration Service (DMS)

DMS is a veteran and a Swiss army knife of the industry. It is the most mature and flexible tool on the market, supporting a huge number of scenarios: from classic lift-and-shift to streaming replication into a Data Lake.

Architecture

Classic: a Replication Instance (a VM in AWS) is created, after which source and target endpoints are configured. All the load — reading logs, transforming data, and writing — falls on that instance.

What can be migrated?

This has the broadest matrix in the industry.

Sources: Oracle, MS SQL Server, PostgreSQL, MySQL, MariaDB, Db2, SAP ASE, MongoDB, Cassandra, S3, and others.
Targets: everything listed above + Redshift, DynamoDB, Neptune, OpenSearch, Kinesis, Kafka.

Key capabilities

• AWS Schema Conversion Tool (SCT): a powerful tool for automatic conversion of schemas, procedures, and functions. The best solution for Oracle/SQL Server → PostgreSQL/Aurora migrations.
• Deep CDC: can read Oracle archived redo logs, imposing almost no load on production.

Pros

• Maximum flexibility, huge scenario coverage
• Ideal for heterogeneous migrations

Cons

− Requires managing infrastructure (instances, monitoring, sizing)
− Billed by the hour for the replication instance
− There is a serverless variant, but it is not yet the main option

Google Cloud Database Migration Service

Google chose the philosophy: migration should be so simple that anyone can do it. Minimal configuration, maximum automation. Practically fully serverless.

Architecture

Fully serverless. No virtual machines are needed: you launch the setup from the master interface in the Google Cloud Console, and GCP manages all resources itself.

What to migrate?

Designed for migrations into Google Cloud.

Sources: MySQL, PostgreSQL, SQL Server, Oracle.
Targets: Cloud SQL (MySQL, PostgreSQL, SQL Server), AlloyDB.

Key features

• Improved support for Oracle → PostgreSQL/AlloyDB, including schema and code conversion.
• Homogeneous migrations (for example, MySQL → MySQL) are free.
• Heterogeneous — first 500 GiB of backfill are free.

Pros

• Fully serverless — no infrastructure to manage
• Friendly interface, quick start
• Often free or very cheap

Cons

− Works only with Google services
− Fewer fine-grained data transformation options

Yandex Data Transfer

A service born from Yandex’s internal tools intended for moving huge volumes of data between shards and services. It has turned into a universal data bus — simple, powerful, and convenient.

Architecture

Managed approach: you create endpoints and transfer tasks, the infrastructure is hidden. At the same time, settings are quite flexible.

Sources and targets

A strong point is open-source and analytics.

Sources: PostgreSQL, MySQL, ClickHouse, MongoDB, Kafka, Oracle CDC, YDB, Elasticsearch.
Targets: PostgreSQL, MySQL, ClickHouse, Greenplum, Kafka, YDB, Object Storage, Data Streams.

Key features

• Best integration with ClickHouse among all clouds: correct upserts, ReplacingMergeTree, partitioning.
• Native support for Kafka as both source and target.
• Support for complex Many-to-One, One-to-Many, sharding scenarios.

Pros

• Ideal for analytics and OLTP → ClickHouse
• Very convenient for Kafka architectures
• Operates within the Russian legal framework (152-ФЗ)

Cons

− Support for enterprise legacy (old Oracle, MS SQL) is weaker
− No automatic schema conversion like AWS SCT

Comparison table

Conclusion: what to choose?

AWS DMS

If you have an enterprise migration: old Oracle/SQL Server, complex schemas, heterogeneous configurations, and strict requirements.

Google Cloud DMS

If simple deployment, speed, minimal cost, and migration into Cloud SQL/AlloyDB are important. A great option for startups, developers, and small companies.

Yandex Data Transfer

If you operate under Russian jurisdiction or build analytics on ClickHouse and streaming on Kafka. A universal and inexpensive data bus.

Each tool is ideal in its own context. Choose the one that best fits your project’s architecture and business requirements.