Phased Delta Migration with Failback Replication from PostgreSQL

A involves in several phases. Data can be sliced per tenant, per service, per region, or per table to suit the needs of the migration. For each given migration phase, you use to perform an initial bulk load of the data, you use to update the target database via forward replication and to activate failback replication, and then you cut over application traffic to CockroachDB after schema finalization and data verification. This process is repeated for each phase of data.

• Data is migrated to the target .

• This approach utilizes .

• is achieved via failback replication.

This approach is comparable to the , but dividing the data into multiple phases allows each downtime window to be shorter, and it allows each phase of the migration to be less complex. Depending on how you divide the data, it also may allow your downtime windows to affect only a subset of users. For example, dividing the data per region could mean that, when migrating the data from Region A, application usage in Region B may remain unaffected. This approach may increase overall migration complexity: its duration is longer, you will need to do the work of partitioning the data, and you will have a longer period when you run both the source and the target database concurrently.

keeps the source database up to date with changes that occur in the target database once the target database begins receiving write traffic. Failback replication ensures that, if something goes wrong during the migration process, traffic can easily be returned to the source database without data loss. Like forward replication, in this approach, failback replication is run on a per-phase basis. It can persist indefinitely, until you’re comfortable maintaining the target database as your sole data store.

This approach is best for databases that are too large to migrate all at once, and in business cases where downtime must be minimal. It’s also suitable for risk-averse situations in which a safe rollback path must be ensured. It can only be performed if your team can handle the complexity of this approach, and if your source database can easily be divided into the phases you need.

This page describes an example scenario. While the commands provided can be copy-and-pasted, they may need to be altered or reconsidered to suit the needs of your specific environment.

Example scenario

You have a moderately-sized (500GB) database that provides the data store for a web application. You want to migrate the entirety of this database to a new CockroachDB cluster. You will divide this migration into four geographic regions (A, B, C, and D).

The application runs on a Kubernetes cluster with an NGINX Ingress Controller.

Your business could not accommodate major performance issues that could arise after the migration. Therefore, you want to enable failback replication so that you can easily return to using your original database with minimal interruption.

Estimated system downtime: 3-5 minutes per region.

Before the migration

• Install the tools.
• Review the and documentation.
• and .
• Recommended: Perform a dry run of this full set of instructions in a development environment that closely resembles your production environment. This can help you get a realistic sense of the time and complexity it requires.
• Understand the prerequisites and limitations of the MOLT tools:

Limitations

MOLT Fetch limitations

• Only tables with types of , , or can be sharded with .

• GEOMETRY and GEOGRAPHY types are not supported.

• OID LOB types in PostgreSQL are not supported, although similar types like BYTEA are supported.

MOLT Replicator limitations

• Replication modes require connection to the primary instance (PostgreSQL primary, MySQL primary/master, or Oracle primary). MOLT cannot obtain replication checkpoints or transaction metadata from replicas.
• Running DDL on the source or target while replication is in progress can cause replication failures.
• TRUNCATE operations on the source are not captured. Only INSERT, UPDATE, UPSERT, and DELETE events are replicated.
• Changes to virtual columns are not replicated automatically. To migrate these columns, you must define them explicitly with .

MOLT Verify limitations

• MOLT Verify compares 20,000 rows at a time by default, and row values can change between batches, potentially resulting in temporary inconsistencies in data. To configure the row batch size, use the --row_batch_size .
• MOLT Verify checks for collation mismatches on columns. This may cause validation to fail when a is used as a primary key and the source and target databases are using different .
• MOLT Verify might give an error in case of schema changes on either the source or target database.
• cannot yet be compared.

Step 1: Prepare the source database

In this step, you will:

• Create a dedicated migration user on your source database.
• Configure the source database for replication.

Create migration user on source database

Create a dedicated migration user (for example, MIGRATION_USER) on the source database. This user is responsible for reading data from source tables during the migration. You will pass this username in the source connection string.

CREATE USER migration_user WITH PASSWORD 'password';

Grant the user privileges to connect, view schema objects, and select the tables you migrate.

GRANT CONNECT ON DATABASE migration_db TO migration_user;
GRANT USAGE ON SCHEMA migration_schema TO migration_user;
GRANT SELECT ON ALL TABLES IN SCHEMA migration_schema TO migration_user;
ALTER DEFAULT PRIVILEGES IN SCHEMA migration_schema GRANT SELECT ON TABLES TO migration_user;

Grant the SUPERUSER role to the user (recommended for replication configuration):

ALTER USER migration_user WITH SUPERUSER;

Alternatively, grant the following permissions to create replication slots, access replication data, create publications, and add tables to publications:

ALTER USER migration_user WITH LOGIN REPLICATION;
GRANT CREATE ON DATABASE migration_db TO migration_user;
ALTER TABLE migration_schema.table_name OWNER TO migration_user;

Run the ALTER TABLE command for each table to replicate.

Configure source database for replication

Connect to the primary instance (PostgreSQL primary, MySQL primary/master, or Oracle primary), not a replica. Replicas cannot provide the necessary replication checkpoints and transaction metadata required for ongoing replication.

Verify that you are connected to the primary server:

SELECT pg_is_in_recovery();

You should get a false result:

 pg_is_in_recovery
-------------------
 f

Enable logical replication by setting wal_level to logical in postgresql.conf or in the SQL shell. For example:

ALTER SYSTEM SET wal_level = 'logical';

Step 2: Prepare the target database

Define the target tables

Convert the source table definitions into CockroachDB-compatible equivalents. CockroachDB supports the PostgreSQL wire protocol and is largely .

• The source and target table definitions must match. Review to understand which source types can be mapped to CockroachDB types.

  For example, a PostgreSQL source table defined as CREATE TABLE migration_schema.tbl (pk INT PRIMARY KEY); must have a corresponding schema and table in CockroachDB:

  CREATE SCHEMA migration_schema;
  CREATE TABLE migration_schema.tbl (pk INT PRIMARY KEY);
  

  • MOLT Fetch can automatically define matching CockroachDB tables using the option.

  • If you define the target tables manually, review how MOLT Fetch handles . You can use the MOLT Schema Conversion Tool to create matching table definitions.

• Every table must have an explicit primary key.

  Avoid using sequential keys. To learn more about the performance issues that can result from their use, refer to the . If a sequential key is necessary in your CockroachDB table, you must create it manually, after using to load and replicate the data.

• Review to understand how computed columns and partitioned tables can be mapped to the target, and how target tables can be renamed.

• By default on CockroachDB, INT is an alias for INT8, which creates 64-bit signed integers. PostgreSQL and MySQL default to 32-bit integers. Depending on your source database or application requirements, you may need to change the integer size to 4. For more information, refer to .

Schema Conversion Tool

The (SCT) converts source table definitions to CockroachDB-compatible syntax. It requires a free .

1. Upload a source .sql file to convert the syntax and identify in the table definitions.

2. Import the converted table definitions to a CockroachDB cluster:

   • When migrating to CockroachDB Cloud, the Schema Conversion Tool automatically .
   • When migrating to a self-hosted CockroachDB cluster, and pipe the directly into .

Drop constraints and indexes

To optimize data load performance, drop all non-PRIMARY KEY and on the target CockroachDB database before migrating:

• (you do not need to drop this constraint when using drop-on-target-and-recreate for table handling)

Do not drop constraints.

You can recreate the constraints and indexes after loading the data.

Create the SQL user

Create a SQL user in the CockroachDB cluster that has the necessary privileges.

To create a user crdb_user in the default database (you will pass this username in the target connection string):

CREATE USER crdb_user WITH PASSWORD 'password';

Grant database-level privileges for schema creation within the target database:

GRANT ALL ON DATABASE defaultdb TO crdb_user;

Grant permission to modify cluster settings:

GRANT SYSTEM MODIFYCLUSTERSETTING TO crdb_user;

Grant usage on schemas being migrated:

GRANT USAGE ON SCHEMA migration_schema TO crdb_user;
GRANT USAGE ON SCHEMA public TO crdb_user;

Grant user privileges to create tables in the migration_schema schema and internal MOLT tables like _molt_fetch_exceptions in the public CockroachDB schema:

Ensure that you are connected to the target database.

GRANT CREATE ON SCHEMA migration_schema TO crdb_user;
GRANT CREATE ON SCHEMA public TO crdb_user;

If you manually defined the target tables (which means that drop-on-target-and-recreate will not be used), grant the following privileges on the schema:

GRANT SELECT, INSERT, UPDATE, DELETE ON ALL TABLES IN SCHEMA migration_schema TO crdb_user;
ALTER DEFAULT PRIVILEGES IN SCHEMA migration_schema
GRANT SELECT, INSERT, UPDATE, DELETE ON TABLES TO crdb_user;

Grant the same privileges for internal MOLT tables in the public CockroachDB schema:

GRANT SELECT, INSERT, UPDATE, DELETE ON ALL TABLES IN SCHEMA public TO crdb_user;
ALTER DEFAULT PRIVILEGES IN SCHEMA public
GRANT SELECT, INSERT, UPDATE, DELETE ON TABLES TO crdb_user;

If you will be running Fetch with drop-on-target-and-recreate, and the target schema has pre-existing tables that were created by a different user, you may need to change table ownership to allow the migration user to drop those tables. Make the following alteration for each table:

ALTER TABLE migration_schema.table1 OWNER TO crdb_user;

Depending on the MOLT Fetch you will use, grant the necessary privileges to run either IMPORT INTO or COPY FROM on the target tables:

IMPORT INTO privileges

Grant SELECT, INSERT, and DROP (required because the table is taken offline during the IMPORT INTO) privileges on all tables being migrated:

GRANT SELECT, INSERT, DROP ON ALL TABLES IN SCHEMA migration_schema TO crdb_user;

If you plan to use cloud storage with implicit authentication for data load, grant the EXTERNALIOIMPLICITACCESS :

GRANT SYSTEM EXTERNALIOIMPLICITACCESS TO crdb_user;

COPY FROM privileges

Grant privileges to the user:

GRANT admin TO crdb_user;

Set session variable

Ensure that the statement_timeout is set to 0s for the user:

ALTER ROLE crdb_user SET statement_timeout = '0s';

Replication privileges

Grant permissions to create the staging schema for replication:

ALTER USER crdb_user CREATEDB;

Configure GC TTL

Before starting the initial data load, configure the on the source CockroachDB cluster to ensure that historical data remains available when replication begins. The GC TTL must be long enough to cover the full duration of the data load.

Increase the GC TTL before starting the data load. For example, to set the GC TTL to 24 hours:

ALTER DATABASE defaultdb CONFIGURE ZONE USING gc.ttlseconds = 86400;

The GC TTL duration must be higher than your expected time for the initial data load.

Once replication has started successfully (which automatically protects its own data range), you can restore the GC TTL to its original value. For example, to restore to 5 minutes:

ALTER DATABASE defaultdb CONFIGURE ZONE USING gc.ttlseconds = 300;

For details, refer to .

Migrate each phase

Steps 3-12 are run for each phase of the data migration. When migrating the first phase, you will run through these steps for Region A. You will repeat these steps for the other regions during each subsequent migration phase.

Step 3: Load data into CockroachDB

In this step, you will:

• Configure MOLT Fetch with the flags needed for your migration.
• Run MOLT Fetch.
• Understand how to continue a load after an interruption.

Configure MOLT Fetch

The includes detailed information about how to , and how to .

When you run molt fetch, you can configure the following options for data load:

• : Specify URL‑encoded source and target connections.
• : Specify schema and table names to migrate. Important for a phased migration.
• : Export data to cloud storage or a local file server.
• : Specifies whether data will only be loaded into/from intermediate storage.
• : Divide larger tables into multiple shards during data export.
• : Choose between IMPORT INTO and COPY FROM.
• : Determine how existing target tables are initialized before load.
• : Define any row-level transformations to apply to the data before it reaches the target.
• : Configure metrics collection during initial data load.

Read through the documentation to understand how to configure your molt fetch command and its flags. Follow , especially those related to security.

At minimum, the molt fetch command should include the source, target, data path, and flags. For a phased migration, you may also choose to include or flags:

molt fetch \
--source $SOURCE \
--target $TARGET \
--table-filter '.*user.*' \
--bucket-path 's3://bucket/path' \
--ignore-replication-check

However, depending on the needs of your migration, you may have many more flags set, and you may need to prepare some accompanying .json files.

Run MOLT Fetch

Perform the initial load of the source data.

1. Issue the command to move the source data to CockroachDB. This example command passes the source and target connection strings as environment variables, writes intermediate files to S3 storage, and uses the truncate-if-exists table handling mode to truncate the target tables before loading data. It also limits the migration to a single schema and filters three specific tables to migrate. The defaults to IMPORT INTO.

   You must include --pglogical-replication-slot-name and --pglogical-publication-and-slot-drop-and-recreate to automatically create the publication and replication slot during the data load.

   molt fetch \
   --source $SOURCE \
   --target $TARGET \
   --schema-filter 'migration_schema' \
   --table-filter 'employees|payments|orders' \
   --bucket-path 's3://migration/data/cockroach' \
   --table-handling truncate-if-exists \
   --pglogical-replication-slot-name molt_slot \
   --pglogical-publication-and-slot-drop-and-recreate
   

2. Check the output to observe fetch progress.

   If you included the --pglogical-replication-slot-name and --pglogical-publication-and-slot-drop-and-recreate flags, a publication named molt_fetch is automatically created:

   {"level":"info","time":"2025-02-10T14:28:11-05:00","message":"dropping and recreating publication molt_fetch"}
   

   A starting fetch message indicates that the task has started:

   {"level":"info","type":"summary","num_tables":3,"cdc_cursor":"0/43A1960","time":"2025-02-10T14:28:11-05:00","message":"starting fetch"}
   

   data extraction messages are written for each table that is exported to the location in --bucket-path:

   {"level":"info","table":"migration_schema.employees","time":"2025-02-10T14:28:11-05:00","message":"data extraction phase starting"}
   
   {"level":"info","table":"migration_schema.employees","type":"summary","num_rows":200000,"export_duration_ms":1000,"export_duration":"000h 00m 01s","time":"2025-02-10T14:28:12-05:00","message":"data extraction from source complete"}
   

   data import messages are written for each table that is loaded into CockroachDB:

   {"level":"info","table":"migration_schema.employees","time":"2025-02-10T14:28:12-05:00","message":"starting data import on target"}
   
   {"level":"info","table":"migration_schema.employees","type":"summary","net_duration_ms":1899.748333,"net_duration":"000h 00m 01s","import_duration_ms":1160.523875,"import_duration":"000h 00m 01s","export_duration_ms":1000,"export_duration":"000h 00m 01s","num_rows":200000,"cdc_cursor":"0/43A1960","time":"2025-02-10T14:28:13-05:00","message":"data import on target for table complete"}
   

   A fetch complete message is written when the fetch task succeeds:

   {"level":"info","type":"summary","fetch_id":"f5cb422f-4bb4-4bbd-b2ae-08c4d00d1e7c","num_tables":3,"tables":["migration_schema.employees","migration_schema.payments","migration_schema.payments"],"skipped_unmigratable_tables":[],"cdc_cursor":"0/3F41E40","net_duration_ms":6752.847625,"net_duration":"000h 00m 06s","time":"2024-03-18T12:30:37-04:00","message":"fetch complete"}
   

Continue MOLT Fetch after an interruption

If MOLT Fetch fails while loading data into CockroachDB from intermediate files, it exits with an error message, fetch ID, and continuation token for each table that failed to load on the target database.

{"level":"info","table":"public.employees","file_name":"shard_01_part_00000001.csv.gz","message":"creating or updating token for duplicate key value violates unique constraint \"employees_pkey\"; Key (id)=(1) already exists."}
{"level":"info","table":"public.employees","continuation_token":"a1b2c3d4-e5f6-7890-abcd-ef1234567890","message":"created continuation token"}
{"level":"info","fetch_id":"f5cb422f-4bb4-4bbd-b2ae-08c4d00d1e7c","message":"continue from this fetch ID"}
{"level":"error","message":"Error: error from fetching table for public.employees: error importing data: duplicate key value violates unique
  constraint \"employees_pkey\" (SQLSTATE 23505)"}

You can use this information to where it was interrupted.

Continuation is only possible under the following conditions:

• All data has been exported from the source database into intermediate files on or .
• The initial load of source data into the target CockroachDB database is incomplete.
• The load uses IMPORT INTO rather than COPY FROM.

Only one fetch ID and set of continuation tokens, each token corresponding to a table, are active at any time. See .

The following command reattempts the data load starting from a specific continuation file, but you can also use individual continuation tokens to .

molt fetch \
--source $SOURCE \
--target $TARGET \
--schema-filter 'migration_schema' \
--table-filter 'employees|payments|orders' \
--bucket-path 's3://migration/data/cockroach' \
--table-handling truncate-if-exists \
--ignore-replication-check \
--fetch-id f5cb422f-4bb4-4bbd-b2ae-08c4d00d1e7c \
--continuation-file-name shard_01_part_00000001.csv.gz

Step 4: Verify the initial data load

In this step, you will use to confirm that the source and target data is consistent. This ensures that the data load was successful. Use MOLT Verify’s or to select only the tables that are relevant for the given phase.

Run MOLT Verify

1. Run the command, specifying the source and target connection strings and the tables to validate.

   molt verify \
   --source $SOURCE \
   --target $TARGET \
   --table-filter 'employees|payments|orders'
   

2. Check the output to observe verify progress.

   A verification in progress indicates that the task has started:

   {"level":"info","time":"2025-02-10T15:35:04-05:00","message":"verification in progress"}
   

   starting verify messages are written for each specified table:

   {"level":"info","time":"2025-02-10T15:35:04-05:00","message":"starting verify on public.employees, shard 1/1"}
   

   A finished row verification message is written after each table is compared. If num_success equals num_truth_rows and the error counters (num_missing, num_mismatch, num_extraneous, and num_column_mismatch) are all 0, the table verified successfully. Any non-zero values in the error counters indicate data discrepancies that need investigation. For details on each field, refer to the page.

   {"level":"info","type":"summary","table_schema":"public","table_name":"employees","num_truth_rows":200004,"num_success":200004,"num_conditional_success":0,"num_missing":0,"num_mismatch":0,"num_extraneous":0,"num_live_retry":0,"num_column_mismatch":0,"time":"2025-02-10T15:35:05-05:00","message":"finished row verification on public.employees (shard 1/1)"}
   

   A verification complete message is written when the verify task succeeds:

   {"level":"info","net_duration_ms":699.804875,"net_duration":"000h 00m 00s","time":"2025-02-10T15:35:05-05:00","message":"verification complete"}
   

Step 5: Finalize the target schema

Add constraints and indexes

Add any constraints or indexes that you previously removed from the CockroachDB schema to facilitate data load.

If you used the --table-handling drop-on-target-and-recreate option for data load, only and constraints are preserved. You must manually recreate all other constraints and indexes.

For the appropriate SQL syntax, refer to and . Review the on CockroachDB.

Step 6: Begin forward replication

In this step, you will:

• Configure MOLT Replicator with the flags needed for your migration.
• Start MOLT Replicator.
• Understand how to continue replication after an interruption.

Configure MOLT Replicator (forward replication)

When you run replicator, you can configure the following options for replication:

• Replication connection strings: Specify URL-encoded source and target database connections.

• Replicator flags: Specify required and optional flags to configure replicator behavior.

• Tuning parameters: Optimize replication performance and resource usage.

• Replicator metrics: Monitor replication progress and performance.

Replication connection strings

MOLT Replicator uses --sourceConn and --targetConn to specify the source and target database connections.

--sourceConn specifies the connection string of the source database:

--sourceConn 'postgresql://{username}:{password}@{host}:{port}/{database}'

--targetConn specifies the target CockroachDB connection string:

--targetConn 'postgresql://{username}:{password}@{host}:{port}/{database}'

Follow best practices for securing connection strings. Refer to Secure connections.

Replicator flags

Configure the following flags for continuous replication. For details on all available flags, refer to .

Flag	Description
	Required. PostgreSQL replication slot name. Must match the slot name specified with --pglogical-replication-slot-name in the MOLT Fetch command.
	Required. PostgreSQL publication name. Must match the publication name created either manually or automatically with --pglogical-publication-and-slot-drop-and-recreate in the MOLT Fetch command.
	Required. Target schema name on CockroachDB where tables will be replicated. Schema name must be fully qualified in the format database.schema.
	Required. Staging schema name on CockroachDB for replication metadata and checkpoints. Schema name must be fully qualified in the format database.schema.
	Automatically create the staging schema if it does not exist. Include this flag when starting replication for the first time.
	Enable Prometheus metrics at a specified {host}:{port}. Metrics are served at http://{host}:{port}/_/varz.
	Path to a that enables data filtering, routing, or transformations. For examples, refer to .

Tuning parameters

Configure the following to optimize replication throughput and resource usage. Test different combinations in a pre-production environment to find the optimal balance of stability and performance for your workload.

The following parameters apply to PostgreSQL, Oracle, and CockroachDB (failback) sources.

Flag	Description
	Control the maximum number of concurrent target transactions. Higher values increase throughput but require more target connections. Start with a conservative value and increase based on target database capacity.
	Balance throughput and latency. Controls how many mutations are batched into each query to the target. Increase for higher throughput at the cost of higher latency.
	Control memory usage during operation. Increase to allow higher throughput at the expense of memory; decrease to apply backpressure and limit memory consumption.
	Set larger than by a safety factor to avoid exhausting target pool connections. Replicator enforces setting parallelism to 80% of this value.
	Reduce the number of queries to the target by combining multiple mutations on the same primary key within each batch. Disable only if exact mutation order matters more than end state.
	Improve apply throughput for independent tables and table groups that share foreign key dependencies. Increases memory and target connection usage, so ensure you increase or reduce .
	Set to the maximum allowable time between flushes (for example, 10s if data must be applied within 10 seconds). Works with to control when buffered mutations are committed to the target.
	Lower this value if constraint violations resolve quickly on your workload to make retries more frequent and reduce latency. Do not lower if constraint violations take time to resolve.
	Applies to (replicator start) scenarios only. Balance memory usage and throughput. Increase to read more rows at once from the CockroachDB staging cluster for higher throughput, at the cost of memory pressure. Decrease to reduce memory pressure and increase stability.

Replicator metrics

MOLT Replicator metrics are not enabled by default. Enable Replicator metrics by specifying the flag with a port (or host:port) when you start Replicator. This exposes Replicator metrics at http://{host}:{port}/_/varz. For example, the following flag exposes metrics on port 30005:

--metricsAddr :30005

For guidelines on using and interpreting replication metrics, refer to .

Start MOLT Replicator (forward replication)

With initial load complete, start replication of ongoing changes on the source to CockroachDB using .

MOLT Fetch captures a consistent point-in-time checkpoint at the start of the data load (shown as cdc_cursor in the fetch output). Starting replication from this checkpoint ensures that all changes made during and after the data load are replicated to CockroachDB, preventing data loss or duplication. The following steps use the checkpoint values from the fetch output to start replication at the correct position.

Run the replicator command, using the same slot name that you specified with --pglogical-replication-slot-name and the publication name created by --pglogical-publication-and-slot-drop-and-recreate in the Fetch command. Use --stagingSchema to specify a unique name for the staging database, and include --stagingCreateSchema to have MOLT Replicator automatically create the staging database:

replicator pglogical \
--sourceConn $SOURCE \
--targetConn $TARGET \
--targetSchema defaultdb.migration_schema \
--slotName molt_slot \
--publicationName molt_fetch \
--stagingSchema defaultdb._replicator \
--stagingCreateSchema \
--metricsAddr :30005 \
-v

Check that replication is working

Verify that Replicator is processing changes successfully. To do so, check the MOLT Replicator logs. Since you enabled debug logging with -v, you should see connection and row processing messages:

You should see periodic primary keepalive messages:

DEBUG  [2025-08-25T14:38:10-05:00] primary keepalive received                    ReplyRequested=false ServerTime="2025-08-25 14:38:09.556773 -0500 CDT" ServerWALEnd=0/49913A58
DEBUG  [2025-08-25T14:38:15-05:00] primary keepalive received                    ReplyRequested=false ServerTime="2025-08-25 14:38:14.556836 -0500 CDT" ServerWALEnd=0/49913E60

When rows are successfully replicated, you should see debug output like the following:

DEBUG  [2025-08-25T14:40:02-05:00] upserted rows                                 conflicts=0 duration=7.855333ms proposed=1 target="\"molt\".\"public\".\"tbl1\"" upserted=1
DEBUG  [2025-08-25T14:40:02-05:00] progressed to LSN: 0/49915DD0

These messages confirm successful replication. You can disable verbose logging after verifying the connection.

Continue MOLT Replicator after an interruption (forward replication)

Run the pglogical command using the same --stagingSchema value from your initial replication command.

Be sure to specify the same --slotName value that you used during your initial replication command. The replication slot on the source PostgreSQL database automatically tracks the LSN (Log Sequence Number) checkpoint, so replication will resume from where it left off.

replicator pglogical \
--sourceConn $SOURCE \
--targetConn $TARGET \
--targetSchema defaultdb.migration_schema \
--slotName molt_slot \
--stagingSchema defaultdb._replicator \
--metricsAddr :30005 \
-v

Replication resumes from the last checkpoint without performing a fresh load. Monitor the metrics endpoint at http://localhost:30005/_/varz to track replication progress.

Step 7: Stop application traffic

Once the inital data load has been verified and the target schema has been finalized, it’s time to begin the cutover process. First, stop application traffic to the source for this particular region.

If the Kubernetes cluster that deploys the application has pre-region deployments (for example, app-us, app-eu, app-apac), you can scale down only the deployment for that region.

kubectl scale deploy/app-eu --replicas=0

Or this can be handled by the NGINX Ingress Controller, by including the following to your NGINX configuration, ensuring that the conditional statement is suitable for your deployment:

apiVersion: networking.k8s.io/v1
kind: Ingress
metadata:
  name: app
  annotations:
    nginx.ingress.kubernetes.io/server-snippet: |
      if ($http_x_region = "eu") {
        return 503;
      }
spec:
  ingressClassName: nginx
  rules:
  - host: api.example.com
    http:
      paths:
      - path: /
        pathType: Prefix
        backend:
          service:
            name: app
            port:
              number: 80

Application downtime begins now, for users in the given region.

It is strongly recommended that you perform a dry run of this migration in a test environment. This will allow you to practice using the MOLT tools in real time, and it will give you an accurate sense of how long application downtime might last.

Step 8: Stop forward replication

Before you can cut over traffic to the target, the changes to the source database need to finish being written to the target. Once the source is no longer receiving write traffic, MOLT Replicator will take some seconds to finish replicating the final changes. This is known as drainage.

1. Wait for replication to drain, which means that all transactions that occurred on the source database have been fully processed and replicated. There are several ways to determine that replication has fully drained:

   • When replication is caught up, you will not see new upserted rows logs.

   • If you set up the replication metrics endpoint with in the preceding steps, metrics are available at:

     http://{host}:{port}/_/varz
     

     Use the following Prometheus alert expression to observe when the combined rate of upserts and deletes is 0 for each schema:

     sum by (schema) (rate(apply_upserts_total[$__rate_interval]) + rate(apply_deletes_total[$__rate_interval]))
     

   • You can also check Prometheus metrics associated with replication lag, including , , and .

2. Cancel replication by entering ctrl-c to issue a SIGTERM signal. This returns an exit code 0.

Step 9: Verify the replicated data

Repeat Step 4 to verify the updated data.

Step 10: Begin failback replication

In this step, you will:

• Prepare both databases for failback replication
• Configure MOLT Replicator with the flags needed for your migration.
• Start MOLT Replicator.
• Understand how to continue replication after an interruption.

Prepare your source and target databases for failback replication

Prepare the CockroachDB cluster

For details on enabling CockroachDB changefeeds, refer to .

If you are migrating to a CockroachDB self-hosted cluster, on the cluster:

SET CLUSTER SETTING kv.rangefeed.enabled = true;

Use the following optional settings to increase changefeed throughput.

The following settings can impact source cluster performance and stability, especially SQL foreground latency during writes. For details, refer to .

To lower changefeed emission latency, but increase SQL foreground latency:

SET CLUSTER SETTING kv.rangefeed.closed_timestamp_refresh_interval = '250ms';

To lower the lag duration:

SET CLUSTER SETTING kv.closed_timestamp.target_duration = '1s';

To improve catchup speeds but increase cluster CPU usage:

SET CLUSTER SETTING kv.rangefeed.concurrent_catchup_iterators = 64;

Grant target database user permissions

You should have already created a migration user on the target database (your original source database) with the necessary privileges. Refer to Create migration user on source database.

For failback replication, grant the user additional privileges to write data back to the target database:

-- Grant INSERT and UPDATE on tables to fail back to
GRANT INSERT, UPDATE ON ALL TABLES IN SCHEMA migration_schema TO migration_user;
ALTER DEFAULT PRIVILEGES IN SCHEMA migration_schema GRANT INSERT, UPDATE ON TABLES TO migration_user;

Create a CockroachDB changefeed

On the target cluster, create a CockroachDB changefeed to send changes to MOLT Replicator.

1. Get the current logical timestamp from CockroachDB, after ensuring that forward replication has fully drained:

   SELECT cluster_logical_timestamp();
   
       cluster_logical_timestamp
   ----------------------------------
     1759246920563173000.0000000000
   

2. Create the CockroachDB changefeed pointing to the MOLT Replicator webhook endpoint. Use cursor to specify the logical timestamp from the preceding step. For details on the webhook sink URI, refer to .

   Explicitly set a default 10s value in the CREATE CHANGEFEED statement. This value ensures that the webhook can report failures in inconsistent networking situations and make crash loops more visible.

   The target schema is specified in the webhook URL path in the fully-qualified format /database/schema. The path specifies the database and schema on the target PostgreSQL database. For example, /migration_db/migration_schema routes changes to the migration_schema schema in the migration_db database.

   CREATE CHANGEFEED FOR TABLE employees, payments, orders \
   INTO 'webhook-https://replicator-host:30004/migration_db/migration_schema?client_cert={base64_encoded_cert}&client_key={base64_encoded_key}&ca_cert={base64_encoded_ca}' \
   WITH updated, resolved = '250ms', min_checkpoint_frequency = '250ms', initial_scan = 'no', cursor = '1759246920563173000.0000000000', webhook_sink_config = '{"Flush":{"Bytes":1048576,"Frequency":"1s"}}', webhook_client_timeout = '10s';
   

   The output shows the job ID:

           job_id
   -----------------------
     1101234051444375553
   
   Ensure that only one changefeed points to MOLT Replicator at a time to avoid mixing streams of incoming data.

3. Monitor the changefeed status, specifying the job ID:

   SHOW CHANGEFEED JOB 1101234051444375553;
   
           job_id        | ... | status  |              running_status               | ...
   ----------------------+-----+---------+-------------------------------------------+----
     1101234051444375553 | ... | running | running: resolved=1759246920563173000,0 | ...
   

   To confirm the changefeed is active and replicating changes to the target database, check that status is running and running_status shows running: resolved={timestamp}.

   running: resolved may be reported even if data isn’t being sent properly. This typically indicates incorrect host/port configuration or network connectivity issues.

4. Verify that Replicator is reporting incoming HTTP requests from the changefeed. To do so, check the MOLT Replicator logs. Since you enabled debug logging with -v, you should see periodic HTTP request successes:

   DEBUG  [2025-08-25T11:52:47-05:00]  httpRequest="&{0x14000b068c0 45 200 3 9.770958ms   false false}"
   DEBUG  [2025-08-25T11:52:48-05:00]  httpRequest="&{0x14000d1a000 45 200 3 13.438125ms   false false}"
   

   These debug messages confirm successful changefeed connections to MOLT Replicator. You can disable verbose logging after verifying the connection.

Configure MOLT Replicator (failback replication)

When you run replicator, you can configure the following options for replication:

• Connection strings: Specify URL‑encoded source and target connections.

• TLS certificate and key: Configure secure TLS connections.

• Replicator flags: Specify required and optional flags to configure replicator behavior.

• Tuning parameters: Optimize failback performance and resource usage.

• Replicator metrics: Monitor failback replication performance.

Replication connection strings

MOLT Replicator uses --sourceConn and --targetConn to specify the source and target database connections.

For MOLT Replicator, the source is always the replication source, while the target is always the replication target. This is distinct from the migration source and target. In the case of this example migration, the new CockroachDB cluster is the migration target, but because failback replication moves data from the migration target back to the migration source, the replication target is the original source database. In essence, the --sourceConn and --targetConn strings should be reversed for failback replication.

--sourceConn specifies the connection string of the CockroachDB cluster:

--sourceConn 'postgresql://{username}:{password}@{host}:{port}/{database}'

--targetConn specifies the original source database:

--targetConn 'postgresql://{username}:{password}@{host}:{port}/{database}'

Follow best practices for securing connection strings. Refer to Secure connections.

Secure connections

• To keep your database credentials out of shell history and logs, follow these best practices when specifying your source and target connection strings:

  • Avoid plaintext connection strings.

  • Provide your connection strings as environment variables. For example:

    export SOURCE="postgres://migration_user:a%2452%26@localhost:5432/migration_db?sslmode=verify-full"
    export TARGET="postgres://root@localhost:26257/defaultdb?sslmode=verify-full"
    

    Afterward, reference the environment variables in MOLT commands:

    --source $SOURCE
    --target $TARGET
    

  • If possible, use an external secrets manager to load the environment variables from stored secrets.

• Use TLS-enabled connection strings to encrypt data in transit from MOLT to the database. When using TLS certificates, ensure certificate files are accessible to the MOLT binary on the same machine.

  For example, a PostgreSQL connection string with TLS certificates:

  postgresql://[e-mail]:5432/appdb?sslmode=verify-full&sslrootcert=/etc/migration_db/certs/ca.pem&sslcert=/etc/migration_db/certs/client.crt&sslkey=/etc/migration_db/certs/client.key
  

• URL-encode connection strings for the source database and so special characters in passwords are handled correctly.

  • Given a password a$52&, pass it to the molt escape-password command with single quotes:

    molt escape-password --password 'a$52&'
    

    Use the encoded password in your connection string. For example:

    postgres://migration_user:a%2452%26@localhost:5432/migration_db
    

• Remove sslmode=disable from production connection strings.

TLS certificate and key

Always use secure TLS connections for failback replication to protect data in transit. Do not use insecure configurations in production: avoid the --disableAuthentication and --tlsSelfSigned Replicator flags and insecure_tls_skip_verify=true query parameter in the changefeed webhook URI.

Generate self-signed TLS certificates or certificates from an external CA. Ensure the TLS server certificate and key are accessible on the MOLT Replicator host machine via a relative or absolute file path. When you start failback with Replicator, specify the paths with --tlsCertificate and --tlsPrivateKey. For example:

replicator start \
... \
--tlsCertificate ./certs/server.crt \
--tlsPrivateKey ./certs/server.key

The client certificates defined in the changefeed webhook URI must correspond to the server certificates specified in the replicator command. This ensures proper TLS handshake between the changefeed and MOLT Replicator. To include client certificates in the changefeed webhook URL, encode them with base64 and then URL-encode the output with jq:

base64 -i ./client.crt | jq -R -r '@uri'
base64 -i ./client.key | jq -R -r '@uri'
base64 -i ./ca.crt | jq -R -r '@uri'

When you create the changefeed, pass the encoded certificates in the changefeed URL, where client_cert, client_key, and ca_cert are :

CREATE CHANGEFEED FOR TABLE table1, table2
INTO 'webhook-https://host:port/database/schema?client_cert={base64_and_url_encoded_cert}&client_key={base64_and_url_encoded_key}&ca_cert={base64_and_url_encoded_ca}'
WITH ...;

For additional details on the webhook sink URI, refer to .

Replicator flags

Flag	Description
	Required. Staging schema name on CockroachDB for the changefeed checkpoint table. Schema name must be fully qualified in the format database.schema.
	Required. Network address to bind the webhook sink for the changefeed. For example, :30004.
	Path to the server TLS certificate for the webhook sink. Refer to TLS certificate and key.
	Path to the server TLS private key for the webhook sink. Refer to TLS certificate and key.Q
	Enable Prometheus metrics at a specified {host}:{port}. Metrics are served at http://{host}:{port}/_/varz.
	Path to a that enables data filtering, routing, or transformations. For examples, refer to .

• The staging schema is first created during with .

• When configuring a secure changefeed for failback, you must include and , which specify the paths to the server certificate and private key for the webhook sink connection.

Tuning parameters

Configure the following to optimize replication throughput and resource usage. Test different combinations in a pre-production environment to find the optimal balance of stability and performance for your workload.

The following parameters apply to PostgreSQL, Oracle, and CockroachDB (failback) sources.

Flag	Description
	Control the maximum number of concurrent target transactions. Higher values increase throughput but require more target connections. Start with a conservative value and increase based on target database capacity.
	Balance throughput and latency. Controls how many mutations are batched into each query to the target. Increase for higher throughput at the cost of higher latency.
	Control memory usage during operation. Increase to allow higher throughput at the expense of memory; decrease to apply backpressure and limit memory consumption.
	Set larger than by a safety factor to avoid exhausting target pool connections. Replicator enforces setting parallelism to 80% of this value.
	Reduce the number of queries to the target by combining multiple mutations on the same primary key within each batch. Disable only if exact mutation order matters more than end state.
	Improve apply throughput for independent tables and table groups that share foreign key dependencies. Increases memory and target connection usage, so ensure you increase or reduce .
	Set to the maximum allowable time between flushes (for example, 10s if data must be applied within 10 seconds). Works with to control when buffered mutations are committed to the target.
	Lower this value if constraint violations resolve quickly on your workload to make retries more frequent and reduce latency. Do not lower if constraint violations take time to resolve.
	Applies to (replicator start) scenarios only. Balance memory usage and throughput. Increase to read more rows at once from the CockroachDB staging cluster for higher throughput, at the cost of memory pressure. Decrease to reduce memory pressure and increase stability.

Replicator metrics

MOLT Replicator metrics are not enabled by default. Enable Replicator metrics by specifying the flag with a port (or host:port) when you start Replicator. This exposes Replicator metrics at http://{host}:{port}/_/varz. For example, the following flag exposes metrics on port 30005:

--metricsAddr :30005

For guidelines on using and interpreting replication metrics, refer to .

Start MOLT Replicator (failback replication)

With initial load complete, start replication of ongoing changes on the source to CockroachDB using .

MOLT Fetch captures a consistent point-in-time checkpoint at the start of the data load (shown as cdc_cursor in the fetch output). Starting replication from this checkpoint ensures that all changes made during and after the data load are replicated to CockroachDB, preventing data loss or duplication. The following steps use the checkpoint values from the fetch output to start replication at the correct position.

Run the start command to begin failback replication from CockroachDB to your source database. In this example, --metricsAddr :30005 enables a Prometheus endpoint for monitoring replication metrics, and --bindAddr :30004 sets up the webhook endpoint for the changefeed.

--stagingSchema specifies the staging database name (defaultdb._replicator in this example) used for replication checkpoints and metadata. This staging database was created during initial forward replication when you first ran MOLT Replicator with --stagingCreateSchema.

replicator start \
--targetConn $TARGET \
--stagingConn $STAGING \
--stagingSchema defaultdb._replicator \
--metricsAddr :30005 \
--bindAddr :30004 \
--tlsCertificate ./certs/server.crt \
--tlsPrivateKey ./certs/server.key \
-v

Continue MOLT Replicator after an interruption (failback replication)

Run the pglogical command using the same --stagingSchema value from your initial replication command.

Be sure to specify the same --slotName value that you used during your initial replication command. The replication slot on the source PostgreSQL database automatically tracks the LSN (Log Sequence Number) checkpoint, so replication will resume from where it left off.

replicator pglogical \
--sourceConn $SOURCE \
--targetConn $TARGET \
--targetSchema defaultdb.migration_schema \
--slotName molt_slot \
--stagingSchema defaultdb._replicator \
--metricsAddr :30005 \
-v

Replication resumes from the last checkpoint without performing a fresh load. Monitor the metrics endpoint at http://localhost:30005/_/varz to track replication progress.

Step 11: Cut over application traffic

With the target cluster verified and finalized, it’s time to resume application traffic for the current migration phase.

Modify application code

In the application back end, update the application to route traffic for this migration phase to the CockroachDB cluster. A simple example:

env:
  - name: DATABASE_URL_US_EAST
    value: postgres://[e-mail]-east:26257/defaultdb?sslmode=verify-full
  - name: DATABASE_URL_US_WEST
    value: postgres://legacy-db.us-west:5432/defaultdb  # Still on source

In your application code, route database connections based on the user’s region:

def get_db_connection(user_region):
    if user_region == "us-east":
        return os.getenv("DATABASE_URL_US_EAST")  # CockroachDB
    else:
        return os.getenv("DATABASE_URL_US_WEST")  # Source database

Resume application traffic

If you halted traffic by scaling down a regional Kubernetes deployment, scale it back up.

kubectl scale deploy/app-eu --replicas=3

Or if this was handled by the NGINX Controller, remove the 503 block that was written in step 3:

apiVersion: networking.k8s.io/v1
kind: Ingress
metadata:
  name: app
#   annotations:
#     nginx.ingress.kubernetes.io/server-snippet: |
#       if ($http_x_region = "eu") {
#         return 503;
#       }
spec:
  ingressClassName: nginx
  rules:
  - host: api.example.com
    http:
      paths:
      - path: /
        pathType: Prefix
        backend:
          service:
            name: app
            port:
              number: 80

This ends downtime for the current migration phase.

Step 12: Stop failback replication

After traffic has been cut over to the target, you can maintain failback replication indefinitely. Once you decide that you want to use the CockroachDB cluster as your sole data store going forward, you can end failback replication with the following steps.

1. Wait for replication to drain, which means that all transactions that occurred on the source database have been fully processed and replicated. There are several ways to determine that replication has fully drained:

   • When replication is caught up, you will not see new upserted rows logs.

   • If you set up the replication metrics endpoint with in the preceding steps, metrics are available at:

     http://{host}:{port}/_/varz
     

     Use the following Prometheus alert expression to observe when the combined rate of upserts and deletes is 0 for each schema:

     sum by (schema) (rate(apply_upserts_total[$__rate_interval]) + rate(apply_deletes_total[$__rate_interval]))
     

   • You can also check Prometheus metrics associated with replication lag, including , , and .

2. Cancel replication by entering ctrl-c to issue a SIGTERM signal. This returns an exit code 0.

Repeat for each phase

During the next scheduled migration phase, return to step 3 to migrate the next phase of data. Repeat steps 3-12 for each phase of data, until every region’s data has been migrated and all application traffic has been cut over to the target.

See also