SOLUTION · SQL SERVER CDC

SQL Server CDC: change data capture and real-time data replication

Real-time SQL Server CDC from the transaction log or Change Tracking, with the agent that matches your primary keys and SQL Agent policy.

Gluesync by MOLO17 provides SQL Server CDC through two dedicated source agents: a CDC agent that reads the transaction log, and a Change Tracking agent for tables with primary keys. Pick the agent that matches your tables and DBA policy, then deliver changes continuously to the databases, warehouses, lakes, and event streams your teams already use. Manage every pipeline from the Core Hub web UI, the Gluesync control plane.

Start a Gluesync trial Talk to us

VENDOR COMPATIBILITY

Battle-tested on every Microsoft SQL Server you run

The same Gluesync Microsoft SQL Server agent is tested against each vendor offering below, self-managed or fully managed, so capture behaves the same wherever Microsoft SQL Server runs.

MS SQL Server Source Target Tested
Amazon RDS for SQL Server Source Target Tested
Azure SQL Source Target Tested
Microsoft Azure SQL Database Source Target Tested
Microsoft Azure SQL Server Source Target Tested
Microsoft Azure Synapse Analytics Source Target Tested

WHO THIS IS FOR

SQL Server teams moving a system of record into modern platforms

SQL Server DBAs who must approve a source role with a written list of what to enable: SQL Agent for the CDC agent, CDC or Change Tracking on each database and table, a runtime user with db_owner on the database, and TLS 1.2 or higher
Data platform leads feeding Snowflake, BigQuery, Amazon Redshift, data lakes, Kafka, MongoDB, or Couchbase from SQL Server systems of record, who want one control plane for sources and targets instead of a Kafka Connect cluster
Engineers who prototyped the Debezium SQL Server connector or a hand-built Change Tracking reader, and now own offsets, schema history, and on-call for a pipeline the business treats as infrastructure
Teams replacing native replication scripts, Qlik Replicate, Fivetran HVR, or Oracle GoldenGate renewals, looking for a commercial SQL Server CDC path with best-in-class enterprise support, rated 4.9/5 by customers that reaches heterogeneous targets under one Core Hub

THE PROBLEM

SQL Server data that only moves in batches becomes stale

SQL Server is usually the system of record, so scheduled extracts, timestamp polling, and nightly loads leave reporting and downstream applications working from a copy that is already behind, while competing with production for I/O. Buyers searching for SQL Server CDC usually weigh three options: Change Tracking or CDC wired up by hand, a Debezium SQL Server connector on Kafka, or a commercial replication product. Each one leaves a different set of jobs with your team: SQL Agent capture jobs and change-table cleanup, Change Tracking retention and version handling, or Kafka Connect offsets and schema history.

Gluesync addresses that pattern with per-agent CDC. Two SQL Server source agents read changes through SQL Server's own mechanisms, and the same pipeline model, snapshots, and target agents sit on top of either one. You choose the agent by table shape and DBA policy, and each target agent writes to the destination you choose.

HOW IT WORKS

How Gluesync does SQL Server CDC

Two native capture techniques, one pipeline model

Both agents connect over the built-in JDBC driver with TLS. Each reads a different SQL Server feature, so the choice depends on your tables and your DBA's policy rather than on the target.

CDC agent: reads the transaction log. It replays the log blocks that SQL Server's Change Data Capture APIs surface, and SQL Server Agent runs the capture jobs that feed those APIs. CDC agent docs ↗
Change Tracking agent: reads row-level deltas from SQL Server Change Tracking metadata tables. It continuously queries the Change Tracking functions for incremental changes, and every tracked table needs a primary key. Change Tracking agent docs ↗

What your DBA will be asked to enable

Enablement differs by agent. The list below summarizes the steps; full statements are in each agent's setup guide.

CDC agent: SQL Server Agent running on the source. On Linux, /opt/mssql/bin/mssql-conf set sqlagent.enabled true, then a restart. A DBA with sysadmin turns it on.
CDC agent: CDC enabled on the database with sys.sp_cdc_enable_db, then on each captured table with sys.sp_cdc_enable_table.
Change Tracking agent: Change Tracking enabled on the database with ALTER DATABASE ... SET CHANGE_TRACKING = ON, then on each table with ALTER TABLE ... ENABLE CHANGE_TRACKING.
Runtime user: db_owner on the database. The CDC functions require it.
Connectivity: TLS 1.2 or higher on the instance, and TCP/IP enabled in SQL Server Configuration Manager. Named Pipes is not used.

Connections and topology

Connects over TCP/IP on port 1433 by default, with optional TLS and a certificate path. Named Pipes is not used and is not required.
Each agent connection takes one database name. Pipelines group the source agent, target agents, and entities as usual.

Snapshots and retention

Initial snapshot ahead of CDC: an entity option loads the current data, then switches to CDC automatically.
Snapshot reads are tunable, with configurable parallelism and logical partitioning for very large tables.
Before and after images are enabled automatically on the CDC agent, so updates carry the previous and the new row values.
Change Tracking retention is configurable. Set it to cover your longest planned outage.

Architecture around Core Hub

Lightweight agents sit close to each system, and Core Hub orchestrates them: web UI, REST APIs, and routing between source and target agents. A pipeline groups one source agent, its target agents, and the entities they replicate. Both SQL Server agents support snapshots, SqlBulkCopy bulk load, and the source and target role, so SQL Server can also be the destination. Core Hub and agents deploy with Docker, Docker Compose, or Kubernetes.

Explore the general CDC streaming architecture →

CAPTURE OPTIONS

Two SQL Server source agents: which one to pick

Either agent plugs into the same pipelines, snapshots, and targets under Core Hub. The choice comes down to your primary keys and your DBA's policy. Pick the CDC agent if some tables have no primary key, if you want before and after images from the log, or if SQL Server Agent can run on the source. Pick the Change Tracking agent if every table has a primary key and SQL Agent is not part of your plan, since its prerequisites do not include SQL Agent. For the reasoning behind the two mechanisms, read MS SQL Server CDC: transaction log versus change tracking.

Agent	Capture technique	Versions	Best for
MS SQL Server CDC agent ↗	Transaction log, read through SQL Server Change Data Capture and SQL Agent capture jobs	SQL Server 2008 and later; Enterprise edition	Tables with no primary key, before and after images from the log (enabled automatically on this agent), and DBA teams that can run SQL Server Agent on the source.
MS SQL Server Change Tracking agent ↗	Row-level deltas from SQL Server Change Tracking metadata, keyed on primary keys	SQL Server 2016 and later; all editions	Databases where every table has a primary key and SQL Agent is not part of the plan, since Change Tracking needs no capture jobs on the instance.

TARGETS AND TOPOLOGIES

Keep SQL Server as the system of record while modernizing destinations

Use the integrations directory to pair Microsoft SQL Server as source or target with relational engines, NoSQL stores, cloud warehouses, object and lake storage, or event streams, subject to each agent's documented source and target role. Both SQL Server agents also support the target role, so SQL Server can be a destination as well as an origin. Bulk loads use SqlBulkCopy into staging tables.

Gluesync keeps pace with your change volume at any scale. MOLO17 Professional Services can size the deployment with your team.

Target agents write in optimized batches, never row by row, and switch to native bulk load for both snapshots and CDC on targets such as Snowflake, Google BigQuery, Amazon Redshift, Microsoft SQL Server, and PostgreSQL.

Offload reporting and API reads from SQL Server to PostgreSQL, a NoSQL store, or a cache: see database offload
Stream SQL Server changes into Snowflake, BigQuery, Amazon Redshift, or a data lake with destination-native bulk loading: see warehouse sync
Publish SQL Server changes to Apache Kafka for microservices and integration
Run a snapshot plus CDC to keep SQL Server and a cloud target aligned until cutover: see cloud migration

FAIR, HIGH-LEVEL COMPARISON

Where Gluesync fits among SQL Server CDC approaches

Approach	What buyers usually get	Where Gluesync fits
CDC or Change Tracking scripted in-house	Full control. Your team writes the readers, retention checks, restart logic, and monitoring, and maintains them after each SQL Server upgrade	Productized agents with snapshots, pipeline controls, and Core Hub monitoring. Transaction log versus Change Tracking covers the operational trade-offs of each mechanism
Debezium SQL Server connector on Kafka	Open-source connector; you run Kafka Connect, offsets, schema history, sinks, and upgrades	Native SQL Server agents with Core Hub operations and MOLO17 enterprise support, without a Connect cluster. Read the Debezium alternative comparison
Qlik Replicate, Fivetran HVR, or GoldenGate-class commercial CDC	Mature replication portfolios with broad source coverage; packaging and licensing vary by product generation	Two SQL Server capture paths under one control plane, with heterogeneous targets. See migrating to Gluesync for supported paths
Cloud-managed CDC services (AWS DMS and similar)	Convenient inside one cloud; SQL Server capture options and targets follow that provider's constraints	Agents deploy where SQL Server runs, on-premises or in a managed service, with Docker, Docker Compose, or Kubernetes. Targets are not limited to one cloud provider
Timestamp polling or trigger-based sync	Simple to start; repeated reads compete with production for I/O and CPU	Agents read changes from SQL Server's own change feeds rather than scanning tables to find them; each agent reads a different SQL Server feature

SQL Server CDC research and implementation detail

What is SQL Server CDC with Gluesync?

Change data capture from Microsoft SQL Server through the CDC agent, which replays transaction log blocks surfaced by SQL Server's Change Data Capture APIs, or through the Change Tracking agent, which reads row-level deltas from SQL Server Change Tracking. Gluesync delivers the captured changes to configured targets via Gluesync agents and Core Hub.

Which SQL Server agent should we choose?

Choose the CDC agent if any table has no primary key, if you want before and after images from the log, or if SQL Server Agent can run on the source. Choose the Change Tracking agent if every table has a primary key and SQL Agent is not part of your plan. Decide per database, because the two mechanisms are not interchangeable.

Which SQL Server versions and editions are supported?

The CDC agent supports SQL Server 2008 and later on Enterprise edition. The Change Tracking agent supports SQL Server 2016 and later on all editions. The CDC agent also supports Standard edition from SQL Server 2022.

What does the DBA have to enable on the database?

For the CDC agent, a DBA with sysadmin rights turns on SQL Server Agent, and CDC is then enabled on the database and each table. For the Change Tracking agent, Change Tracking is enabled on the database and each table. The Gluesync runtime user needs db_owner on the database, and TLS 1.2 or higher must be enabled on the instance.

Do SQL Server tables need primary keys?

Not for the CDC agent, which works even when tables have no primary key. The Change Tracking agent requires a primary key on every table it tracks.

Does Gluesync support Azure SQL and Amazon RDS for SQL Server?

Yes. Both SQL Server agents are tested against Amazon RDS for SQL Server and the Microsoft Azure SQL family, including Azure SQL Database. Because the CDC agent depends on SQL Server Agent, the Change Tracking agent is the fit for Azure SQL Database.

Can Gluesync write to SQL Server, not only read from it?

Yes. Both SQL Server agents support the target role through the built-in JDBC driver, with TLS, and bulk loads use SqlBulkCopy. Identity columns are forced to match the source value.

CDC BY SOURCE DATABASE

Other sources Gluesync captures from

Evaluate Gluesync with your SQL Server transaction log or Change Tracking setup

Start a trial on your infrastructure, or talk to MOLO17 about the right SQL Server agent for your edition, primary keys, SQL Agent policy, and targets.

Start a Gluesync trial Talk to MOLO17