Perform an initial bulk load of the source data using MOLT Fetch, then use MOLT Replicator to replicate ongoing changes to the target.
For CockroachDB-to-CockroachDB migrations, use backup and restore with MOLT Replicator. Contact your account team for guidance.
Before you begin
- Create a CockroachDB Cloud or self-hosted cluster.
- Install the MOLT (Migrate Off Legacy Technology) tools.
- Review the Fetch and Replicator best practices.
- Review Migration Strategy.
Prerequisites
Oracle Instant Client
Install Oracle Instant Client on the machine that will run molt
and replicator
. If using the MOLT Replicator binary (instead of Docker), the Oracle Instant Client libraries must be accessible at /usr/lib
.
On macOS ARM machines, download the Oracle Instant Client. After installation, you should have a new directory at
/Users/$USER/Downloads/instantclient_23_3
containing.dylib
files. Set theLD_LIBRARY_PATH
environment variable to this directory:export LD_LIBRARY_PATH=/Users/$USER/Downloads/instantclient_23_3
On Linux machines, install the Oracle Instant Client dependencies and set the
LD_LIBRARY_PATH
to the client library path:sudo apt-get install -yqq --no-install-recommends libaio1t64 sudo ln -s /usr/lib/x86_64-linux-gnu/libaio.so.1t64 /usr/lib/x86_64-linux-gnu/libaio.so.1 curl -o /tmp/ora-libs.zip https://replicator.cockroachdb.com/third_party/instantclient-basiclite-linux-amd64.zip unzip -d /tmp /tmp/ora-libs.zip sudo mv /tmp/instantclient_21_13/* /usr/lib export LD_LIBRARY_PATH=/usr/lib
Enable ARCHIVELOG
Enable ARCHIVELOG
mode on the Oracle database. This is required for Oracle LogMiner, Oracle's built-in changefeed tool that captures DML events for replication.
SELECT log_mode FROM v$database;
SHUTDOWN IMMEDIATE;
STARTUP MOUNT;
ALTER DATABASE ARCHIVELOG;
ALTER DATABASE OPEN;
SELECT log_mode FROM v$database;
LOG_MODE
--------
ARCHIVELOG
1 row selected.
Enable supplemental primary key logging for logical replication:
ALTER DATABASE ADD SUPPLEMENTAL LOG DATA (PRIMARY KEY) COLUMNS;
SELECT supplemental_log_data_min, supplemental_log_data_pk FROM v$database;
SUPPLEMENTAL_LOG_DATA_MIN SUPPLEMENTAL_LOG_DATA_PK
------------------------- ------------------------
IMPLICIT YES
1 row selected.
Enable FORCE_LOGGING
to ensure that all data changes are captured for the tables to migrate:
ALTER DATABASE FORCE LOGGING;
Limitations
Fetch limitations
OID LOB
types in PostgreSQL are not supported, although similar types likeBYTEA
are supported.
- Migrations must be performed from a single Oracle schema. You must include
--schema-filter
so that MOLT Fetch only loads data from the specified schema. Refer to Schema and table filtering.- Specifying
--table-filter
is also strongly recommended to ensure that only necessary tables are migrated from the Oracle schema.
- Specifying
- Oracle advises against
LONG RAW
columns and recommends converting them toBLOB
.LONG RAW
can only store binary values up to 2GB, and only oneLONG RAW
column per table is supported.
- Only tables with primary key types of
INT
,FLOAT
, orUUID
can be sharded with--export-concurrency
.
Replicator limitations
- Replication modes require write access to the PostgreSQL primary instance. MOLT cannot create replication slots or run replication against a read replica.
- MySQL replication is supported only with GTID-based configurations. Binlog-based features that do not use GTID are not supported.
- Replication will not work for tables or column names exceeding 30 characters. This is a limitation of Oracle LogMiner.
- The following data types are not supported for replication:
- User-defined types (UDTs)
- Nested tables
VARRAY
LONGBLOB
/CLOB
columns (over 4000 characters)
- If your Oracle workload executes
UPDATE
statements that modify only LOB columns, theseUPDATE
statements are not supported by Oracle LogMiner and will not be replicated. - If you are using Oracle 11 and execute
UPDATE
statements onXMLTYPE
or LOB columns, those changes are not supported by Oracle LogMiner and will be excluded from ongoing replication. - If you are migrating LOB columns from Oracle 12c, use AWS DMS Binary Reader instead of LogMiner. Oracle LogMiner does not support LOB replication in 12c.
- Running DDL on the source or target while replication is in progress can cause replication failures.
TRUNCATE
operations on the source are not captured. OnlyINSERT
,UPDATE
,UPSERT
, andDELETE
events are replicated.- Changes to virtual columns are not replicated automatically. To migrate these columns, you must define them explicitly with transformation rules.
Prepare the source database
Create migration user on source database
Create a dedicated migration user (e.g., 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 source_database 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;
CREATE USER 'migration_user'@'%' IDENTIFIED BY 'password';
Grant the user privileges to select only the tables you migrate:
GRANT SELECT ON source_database.* TO MIGRATION_USER@'%';
FLUSH PRIVILEGES;
CREATE USER MIGRATION_USER IDENTIFIED BY 'password';
When migrating from Oracle Multitenant (PDB/CDB), this should be a common user. Prefix the username with C##
(e.g., C##MIGRATION_USER
).
Grant the user privileges to connect, read metadata, and SELECT
and FLASHBACK
the tables you plan to migrate. The tables should all reside in a single schema (e.g., migration_schema
). For details, refer to Schema and table filtering.
Oracle Multitenant (PDB/CDB) user privileges
Connect to the Oracle CDB as a DBA and grant the following:
-- Basic access
GRANT CONNECT TO C##MIGRATION_USER;
GRANT CREATE SESSION TO C##MIGRATION_USER;
-- General metadata access
GRANT EXECUTE_CATALOG_ROLE TO C##MIGRATION_USER;
GRANT SELECT_CATALOG_ROLE TO C##MIGRATION_USER;
-- Access to necessary V$ views
GRANT SELECT ON V_$LOG TO C##MIGRATION_USER;
GRANT SELECT ON V_$LOGFILE TO C##MIGRATION_USER;
GRANT SELECT ON V_$LOGMNR_CONTENTS TO C##MIGRATION_USER;
GRANT SELECT ON V_$ARCHIVED_LOG TO C##MIGRATION_USER;
GRANT SELECT ON V_$DATABASE TO C##MIGRATION_USER;
GRANT SELECT ON V_$LOG_HISTORY TO C##MIGRATION_USER;
-- Direct grants to specific DBA views
GRANT SELECT ON ALL_USERS TO C##MIGRATION_USER;
GRANT SELECT ON DBA_USERS TO C##MIGRATION_USER;
GRANT SELECT ON DBA_OBJECTS TO C##MIGRATION_USER;
GRANT SELECT ON DBA_SYNONYMS TO C##MIGRATION_USER;
GRANT SELECT ON DBA_TABLES TO C##MIGRATION_USER;
Connect to the Oracle PDB (not the CDB) as a DBA and grant the following:
-- Allow C##MIGRATION_USER to connect to the PDB and see active transaction metadata
GRANT CONNECT TO C##MIGRATION_USER;
GRANT CREATE SESSION TO C##MIGRATION_USER;
-- General metadata access
GRANT SELECT_CATALOG_ROLE TO C##MIGRATION_USER;
-- Access to necessary V$ views
GRANT SELECT ON V_$SESSION TO C##MIGRATION_USER;
GRANT SELECT ON V_$TRANSACTION TO C##MIGRATION_USER;
-- Grant these two for every table to migrate in the migration_schema
GRANT SELECT, FLASHBACK ON migration_schema.tbl TO C##MIGRATION_USER;
Single-tenant Oracle user privileges
Connect to the Oracle database as a DBA and grant the following:
-- Basic access
GRANT CONNECT TO MIGRATION_USER;
GRANT CREATE SESSION TO MIGRATION_USER;
-- General metadata access
GRANT SELECT_CATALOG_ROLE TO MIGRATION_USER;
GRANT EXECUTE_CATALOG_ROLE TO MIGRATION_USER;
-- Access to necessary V$ views
GRANT SELECT ON V_$DATABASE TO MIGRATION_USER;
GRANT SELECT ON V_$SESSION TO MIGRATION_USER;
GRANT SELECT ON V_$TRANSACTION TO MIGRATION_USER;
-- Direct grants to specific DBA views
GRANT SELECT ON ALL_USERS TO MIGRATION_USER;
GRANT SELECT ON DBA_USERS TO MIGRATION_USER;
GRANT SELECT ON DBA_OBJECTS TO MIGRATION_USER;
GRANT SELECT ON DBA_SYNONYMS TO MIGRATION_USER;
GRANT SELECT ON DBA_TABLES TO MIGRATION_USER;
-- Grant these two for every table to migrate in the migration_schema
GRANT SELECT, FLASHBACK ON migration_schema.tbl TO MIGRATION_USER;
Configure source database for replication
Connect to the primary PostgreSQL instance, not a read replica. Read replicas cannot create or manage logical replication slots. Verify that you are connected to the primary server by running SELECT pg_is_in_recovery();
and getting a false
result.
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';
Create a publication for the tables you want to replicate. Do this before creating the replication slot.
To create a publication for all tables:
CREATE PUBLICATION molt_publication FOR ALL TABLES;
To create a publication for specific tables:
CREATE PUBLICATION molt_publication FOR TABLE employees, payments, orders;
Create a logical replication slot:
SELECT pg_create_logical_replication_slot('molt_slot', 'pgoutput');
Verify logical replication setup
Verify the publication was created successfully:
SELECT * FROM pg_publication;
oid | pubname | pubowner | puballtables | pubinsert | pubupdate | pubdelete | pubtruncate | pubviaroot
-------+------------------+----------+--------------+-----------+-----------+-----------+-------------+------------
59084 | molt_publication | 10 | t | t | t | t | t | f
Verify the replication slot was created:
SELECT * FROM pg_replication_slots;
slot_name | plugin | slot_type | datoid | database | temporary | active | active_pid | xmin | catalog_xmin | restart_lsn | confirmed_flush_lsn | wal_status | safe_wal_size | two_phase
-----------+----------+-----------+--------+----------+-----------+--------+------------+------+--------------+-------------+---------------------+------------+---------------+-----------
molt_slot | pgoutput | logical | 16385 | molt | f | f | | | 2261 | 0/49913A20 | 0/49913A58 | reserved | | f
Enable global transaction identifiers (GTID) and configure binary logging. Set binlog-row-metadata
or binlog-row-image
to full
to provide complete metadata for replication.
GTID replication sends all database changes to Replicator. To limit replication to specific tables or schemas, use the --table-filter
and --schema-filter
flags in the replicator
command.
Version | Configuration |
---|---|
MySQL 5.6 | --gtid-mode=on --enforce-gtid-consistency=on --server-id={unique_id} --log-bin=mysql-binlog --binlog-format=row --binlog-row-image=full --log-slave-updates=ON |
MySQL 5.7 | --gtid-mode=on --enforce-gtid-consistency=on --binlog-row-image=full --server-id={unique_id} --log-bin=log-bin |
MySQL 8.0+ | --gtid-mode=on --enforce-gtid-consistency=on --binlog-row-metadata=full |
MariaDB | --log-bin --server_id={unique_id} --log-basename=master1 --binlog-format=row --binlog-row-metadata=full |
Verify MySQL GTID setup
Get the current GTID set to use as the starting point for replication:
-- For MySQL < 8.0:
SHOW MASTER STATUS;
-- For MySQL 8.0+:
SHOW BINARY LOG STATUS;
+---------------+----------+--------------+------------------+-------------------------------------------+
| File | Position | Binlog_Do_DB | Binlog_Ignore_DB | Executed_Gtid_Set |
+---------------+----------+--------------+------------------+-------------------------------------------+
| binlog.000005 | 197 | | | 77263736-7899-11f0-81a5-0242ac120002:1-38 |
+---------------+----------+--------------+------------------+-------------------------------------------+
Use the Executed_Gtid_Set
value for the --defaultGTIDSet
flag in MOLT Replicator.
To verify that a GTID set is valid and not purged, use the following queries:
-- Verify the GTID set is in the executed set
SELECT GTID_SUBSET('77263736-7899-11f0-81a5-0242ac120002:1-38', @@GLOBAL.gtid_executed) AS in_executed;
-- Verify the GTID set is not in the purged set
SELECT GTID_SUBSET('77263736-7899-11f0-81a5-0242ac120002:1-38', @@GLOBAL.gtid_purged) AS in_purged;
If in_executed
returns 1
and in_purged
returns 0
, the GTID set is valid for replication.
Enable ARCHIVELOG and FORCE LOGGING
Enable ARCHIVELOG
mode for LogMiner to access archived redo logs:
-- Check current log mode
SELECT log_mode FROM v$database;
-- Enable ARCHIVELOG (requires database restart)
SHUTDOWN IMMEDIATE;
STARTUP MOUNT;
ALTER DATABASE ARCHIVELOG;
ALTER DATABASE OPEN;
-- Verify ARCHIVELOG is enabled
SELECT log_mode FROM v$database; -- Expected: ARCHIVELOG
Enable supplemental logging for primary keys:
ALTER DATABASE ADD SUPPLEMENTAL LOG DATA (PRIMARY KEY) COLUMNS;
-- Verify supplemental logging
SELECT supplemental_log_data_min, supplemental_log_data_pk FROM v$database;
-- Expected: SUPPLEMENTAL_LOG_DATA_MIN: IMPLICIT (or YES), SUPPLEMENTAL_LOG_DATA_PK: YES
Enable FORCE LOGGING
to ensure all changes are logged:
ALTER DATABASE FORCE LOGGING;
-- Verify FORCE LOGGING is enabled
SELECT force_logging FROM v$database; -- Expected: YES
Create source sentinel table
Create a checkpoint table called _replicator_sentinel
in the Oracle schema you will migrate:
CREATE TABLE migration_schema."_replicator_sentinel" (
keycol NUMBER PRIMARY KEY,
lastSCN NUMBER
);
Grant privileges to modify the checkpoint table. In Oracle Multitenant, grant the privileges on the PDB:
GRANT SELECT, INSERT, UPDATE ON migration_schema."_replicator_sentinel" TO C##MIGRATION_USER;
Grant LogMiner privileges
Grant LogMiner privileges. In Oracle Multitenant, grant the permissions on the CDB:
-- Access to necessary V$ views
GRANT SELECT ON V_$LOG TO C##MIGRATION_USER;
GRANT SELECT ON V_$LOGFILE TO C##MIGRATION_USER;
GRANT SELECT ON V_$LOGMNR_CONTENTS TO C##MIGRATION_USER;
GRANT SELECT ON V_$ARCHIVED_LOG TO C##MIGRATION_USER;
GRANT SELECT ON V_$LOG_HISTORY TO C##MIGRATION_USER;
-- SYS-prefixed views (for full dictionary access)
GRANT SELECT ON SYS.V_$LOGMNR_DICTIONARY TO C##MIGRATION_USER;
GRANT SELECT ON SYS.V_$LOGMNR_LOGS TO C##MIGRATION_USER;
GRANT SELECT ON SYS.V_$LOGMNR_PARAMETERS TO C##MIGRATION_USER;
GRANT SELECT ON SYS.V_$LOGMNR_SESSION TO C##MIGRATION_USER;
-- Access to LogMiner views and controls
GRANT LOGMINING TO C##MIGRATION_USER;
GRANT EXECUTE ON DBMS_LOGMNR TO C##MIGRATION_USER;
The user must:
- Query redo logs from LogMiner.
- Retrieve active transaction information to determine the starting point for ongoing replication.
- Update the internal
_replicator_sentinel
table created on the Oracle source schema by the DBA.
Verify LogMiner privileges
Query the locations of redo files in the Oracle database:
SELECT
l.GROUP#,
lf.MEMBER,
l.FIRST_CHANGE# AS START_SCN,
l.NEXT_CHANGE# AS END_SCN
FROM
V$LOG l
JOIN
V$LOGFILE lf
ON
l.GROUP# = lf.GROUP#;
GROUP# MEMBER START_SCN END_SCN
_________ _________________________________________ ____________ ______________________
3 /opt/oracle/oradata/ORCLCDB/redo03.log 1232896 9295429630892703743
2 /opt/oracle/oradata/ORCLCDB/redo02.log 1155042 1232896
1 /opt/oracle/oradata/ORCLCDB/redo01.log 1141934 1155042
3 rows selected.
Get the current snapshot System Change Number (SCN):
SELECT CURRENT_SCN FROM V$DATABASE;
CURRENT_SCN
-----------
2358840
1 row selected.
Get SCNs for replication startup
If you plan to use initial data load followed by replication, obtain the correct SCNs before starting the initial data load to ensure no active transactions are missed. Run the following queries on the PDB in the order shown:
-- Query the current SCN from Oracle
SELECT CURRENT_SCN FROM V$DATABASE;
-- Query the starting SCN of the earliest active transaction
SELECT MIN(t.START_SCNB) FROM V$TRANSACTION t;
Use the results as follows:
--scn
: Use the result from the first query (current SCN)--backfillFromSCN
: Use the result from the second query (earliest active transaction SCN). If the second query returns no results, use the result from the first query instead.
Add the redo log files to LogMiner, using the redo log file paths you queried:
EXEC DBMS_LOGMNR.ADD_LOGFILE(LOGFILENAME => '/opt/oracle/oradata/ORCLCDB/redo01.log', OPTIONS => DBMS_LOGMNR.NEW);
EXEC DBMS_LOGMNR.ADD_LOGFILE(LOGFILENAME => '/opt/oracle/oradata/ORCLCDB/redo02.log', OPTIONS => DBMS_LOGMNR.ADDFILE);
EXEC DBMS_LOGMNR.ADD_LOGFILE(LOGFILENAME => '/opt/oracle/oradata/ORCLCDB/redo03.log', OPTIONS => DBMS_LOGMNR.ADDFILE);
Start LogMiner, specifying the SCN you queried:
EXEC DBMS_LOGMNR.START_LOGMNR(
STARTSCN => 2358840,
OPTIONS => DBMS_LOGMNR.DICT_FROM_ONLINE_CATALOG
);
PL/SQL procedure successfully completed.
If you receive ORA-01435: user does not exist
, the Oracle user lacks sufficient LogMiner privileges. Refer to Grant LogMiner privileges.
Prepare the target database
Create the target schema
Convert the source schema into a CockroachDB-compatible schema. CockroachDB supports the PostgreSQL wire protocol and is largely compatible with PostgreSQL syntax.
The source and target schemas must match. Review Type mapping to understand which source types can be mapped to CockroachDB types.
For example, a 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 create a matching CockroachDB schema using the
drop-on-target-and-recreate
option. - If you create the target schema manually, review how MOLT Fetch handles type mismatches. You can use the MOLT Schema Conversion Tool to create a matching schema.
By default, table and column names are case-insensitive in MOLT Fetch. If using the
--case-sensitive
flag, schema, table, and column names must match Oracle's default uppercase identifiers. Use quoted names on the target to preserve case. For example, the following CockroachDB SQL statement will error:CREATE TABLE co.stores (... store_id ...);
It should be written as:
CREATE TABLE "CO"."STORES" (... "STORE_ID" ...);
When using
--case-sensitive
, quote all identifiers and match the case exactly (for example, use"CO"."STORES"
and"STORE_ID"
).
- MOLT Fetch can automatically create a matching CockroachDB schema using the
Every table must have an explicit primary key. For more information, refer to Primary key best practices.
Review Transformations 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 forINT8
, 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 to4
. For more information, refer to Considerations for 64-bit signed integers.
Schema Conversion Tool
The MOLT Schema Conversion Tool (SCT) automates target schema creation. It requires a free CockroachDB Cloud account.
Upload a source
.sql
file to convert the syntax and identify unimplemented features and syntax incompatibilities in the schema.Import the converted schema to a CockroachDB cluster:
- When migrating to CockroachDB Cloud, the Schema Conversion Tool automatically applies the converted schema to a new Cloud database.
- When migrating to a self-hosted CockroachDB cluster, export a converted schema file and pipe the data definition language (DDL) directly into
cockroach sql
.
Syntax that cannot automatically be converted will be displayed in the Summary Report. These may include the following:
String case sensitivity
Strings are case-insensitive in MySQL and case-sensitive in CockroachDB. You may need to edit your MySQL data to get the results you expect from CockroachDB. For example, you may have been doing string comparisons in MySQL that will need to be changed to work with CockroachDB.
For more information about the case sensitivity of strings in MySQL, refer to Case Sensitivity in String Searches from the MySQL documentation. For more information about CockroachDB strings, refer to STRING
.
Identifier case sensitivity
Identifiers are case-sensitive in MySQL and case-insensitive in CockroachDB. When using the Schema Conversion Tool, you can either keep case sensitivity by enclosing identifiers in double quotes, or make identifiers case-insensitive by converting them to lowercase.
AUTO_INCREMENT
attribute
The MySQL AUTO_INCREMENT
attribute, which creates sequential column values, is not supported in CockroachDB. When using the Schema Conversion Tool, columns with AUTO_INCREMENT
can be converted to use sequences, UUID
values with gen_random_uuid()
, or unique INT8
values using unique_rowid()
. Cockroach Labs does not recommend using a sequence to define a primary key column. For more information, refer to Unique ID best practices.
Changing a column type during schema conversion will cause MOLT Verify to identify a type mismatch during data validation. This is expected behavior.
ENUM
type
MySQL ENUM
types are defined in table columns. On CockroachDB, ENUM
is a standalone type. When using the Schema Conversion Tool, you can either deduplicate the ENUM
definitions or create a separate type for each column.
TINYINT
type
TINYINT
data types are not supported in CockroachDB. The Schema Conversion Tool automatically converts TINYINT
columns to INT2
(SMALLINT
).
Geospatial types
MySQL geometry types are not converted to CockroachDB geospatial types by the Schema Conversion Tool. They should be manually converted to the corresponding types in CockroachDB.
FIELD
function
The MYSQL FIELD
function is not supported in CockroachDB. Instead, you can use the array_position
function, which returns the index of the first occurrence of element in the array.
Example usage:
SELECT array_position(ARRAY[4,1,3,2],1);
array_position
------------------
2
(1 row)
While MySQL returns 0 when the element is not found, CockroachDB returns NULL
. So if you are using the ORDER BY
clause in a statement with the array_position
function, the caveat is that sort is applied even when the element is not found. As a workaround, you can use the COALESCE
operator.
SELECT * FROM table_a ORDER BY COALESCE(array_position(ARRAY[4,1,3,2],5),999);
Drop constraints and indexes
To optimize data load performance, drop all non-PRIMARY KEY
constraints and indexes on the target CockroachDB database before migrating:
FOREIGN KEY
UNIQUE
- Secondary indexes
CHECK
DEFAULT
NOT NULL
(you do not need to drop this constraint when usingdrop-on-target-and-recreate
for table handling)
Do not drop PRIMARY KEY
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 user privileges to create internal MOLT tables like _molt_fetch_exceptions
in the public schema:
Ensure that you are connected to the target database.
GRANT CREATE ON SCHEMA public TO crdb_user;
If you manually created the target schema (i.e., 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:
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;
Depending on the MOLT Fetch data load mode 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 in the target schema:
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
system-level privilege:
GRANT EXTERNALIOIMPLICITACCESS TO crdb_user;
COPY FROM
privileges
Grant admin
privileges to the user:
GRANT admin TO crdb_user;
Replication privileges
Grant permissions to create the staging schema for replication:
ALTER USER crdb_user CREATEDB;
Configure Fetch
When you run molt fetch
, you can configure the following options for data load:
- Connection strings: Specify URL‑encoded source and target connections.
- Intermediate file storage: Export data to cloud storage or a local file server.
- Table handling mode: Determine how existing target tables are initialized before load.
- Schema and table filtering: Specify schema and table names to migrate.
- Data load mode: Choose between
IMPORT INTO
andCOPY FROM
. - Fetch metrics: Configure metrics collection during the load.
Connection strings
Define the connection strings for the source and target databases, and keep them secure.
Source connection string
The --source
flag specifies the connection string for the source database:
--source 'postgres://{username}:{password}@{host}:{port}/{database}?sslmode=verify-full'
For example:
--source 'postgres://migration_user:password@localhost:5432/molt?sslmode=verify-full'
The source connection must point to the PostgreSQL primary instance, not a read replica.
--source 'mysql://{username}:{password}@{host}:{port}/{database}?sslmode=verify-full&sslcert={path_to_client_crt}&sslkey={path_to_client_key}&sslrootcert={path_to_ca_crt}'
For example:
--source 'mysql://migration_user:password@localhost/molt?sslcert=.%2fsource_certs%2fclient.root.crt&sslkey=.%2fsource_certs%2fclient.root.key&sslmode=verify-full&sslrootcert=.%2fsource_certs%2fca.crt'
--source 'oracle://{username}:{password}@{host}:{port}/{service_name}'
In Oracle Multitenant, --source
specifies the connection string for the PDB. --source-cdb
specifies the connection string for the CDB. The username specified in both --source
and --source-cdb
must be a common user with the privileges described in Create migration user on source database.
--source 'oracle://{username}:{password}@{host}:{port}/{PDB_service_name}'
--source-cdb 'oracle://{username}:{password}@{host}:{port}/{CDB_service_name}'
Escape the C##
prefix in the Oracle Multitenant username. For example, write C##MIGRATION_USER
as C%23%23
:
--source 'oracle://C%23%23MIGRATION_USER:password@host:1521/ORCLPDB1'
--source-cdb 'oracle://C%23%23MIGRATION_USER:password@host:1521/ORCLCDB'
Target connection string
The --target
flag specifies the connection string for the target CockroachDB database:
--target 'postgres://{username}:{password}@{host}:{port}/{database}?sslmode=verify-full'
For example:
--target 'postgres://crdb_user:password@localhost:26257/defaultdb?sslmode=verify-full'
For details, refer to Connect using a URL.
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.
URL-encode connection strings for the source database and CockroachDB so special characters in passwords are handled correctly.
Given a password
a$52&
, pass it to themolt escape-password
command with single quotes:molt escape-password --password 'a$52&'
Use the encoded password in your
--source
connection string. For example:--source 'postgres://migration_user:a%2452%26@localhost:5432/replicationload'
Provide your connection strings as environment variables. For example:
export SOURCE="postgres://migration_user:a%2452%26@localhost:5432/molt?sslmode=verify-full" export TARGET="postgres://root@localhost:26257/molt?sslmode=verify-full"
Afterward, reference the environment variables as follows:
--source $SOURCE --target $TARGET
If possible, use an external secrets manager to load the environment variables from stored secrets.
Intermediate file storage
MOLT Fetch can write intermediate files to either a cloud storage bucket or a local file server:
Destination | MOLT Fetch flag(s) | Address and authentication |
---|---|---|
Cloud storage | --bucket-path |
Specify a s3://bucket/path , gs://bucket/path , or azure-blob://bucket/path URL.
|
Local file server | --local-path --local-path-listen-addr --local-path-crdb-access-addr |
Write to --local-path on a local file server at --local-path-listen-addr ; if the target CockroachDB cluster cannot reach this address, specify a publicly accessible address with --local-path-crdb-access-addr . No additional authentication is required. |
Cloud storage is often preferred over a local file server, which may require significant disk space.
Cloud storage authentication
Ensure that access control is properly configured for Amazon S3, Google Cloud Storage, or Azure Blob Storage.
Amazon S3
Set the following environment variables in the terminal running
molt fetch
:export AWS_REGION='us-east-1' export AWS_SECRET_ACCESS_KEY='key' export AWS_ACCESS_KEY_ID='id'
To run
molt fetch
in a containerized environment (e.g., Docker), pass the required environment variables using-e
. If your authentication method relies on local credential files, you may also need to volume map the host path to the appropriate location inside the container using-v
. For example:docker run \ -e AWS_ACCESS_KEY_ID='your-access-key' \ -e AWS_SECRET_ACCESS_KEY='your-secret-key' \ -v ~/.aws:/root/.aws \ -it \ cockroachdb/molt fetch \ --bucket-path 's3://migration/data/cockroach' ...
Alternatively, set
--use-implicit-auth
to use implicit authentication. When using assume role authentication, specify the service account with--assume-role
. For example:--bucket-path 's3://migration/data/cockroach' --assume-role 'arn:aws:iam::123456789012:role/MyMigrationRole' --use-implicit-auth
Set
--import-region
to specify anAWS_REGION
(e.g.,--import-region 'ap-south-1'
).Ensure the S3 bucket is created and accessible by authorized roles and users only.
Google Cloud Storage
Authenticate your local environment with Application Default Credentials:
Using
gcloud
:gcloud init gcloud auth application-default login
Using the environment variable:
export GOOGLE_APPLICATION_CREDENTIALS={path_to_cred_json}
To run
molt fetch
in a containerized environment (e.g., Docker), pass the required environment variables using-e
. If your authentication method relies on local credential files, you may also need to volume map the host path to the appropriate location inside the container using-v
. For example:docker run \ -e GOOGLE_APPLICATION_CREDENTIALS='/root/.config/gcloud/application_default_credentials.json' \ -v ~/.config/gcloud:/root/.config/gcloud \ -it \ cockroachdb/molt fetch \ --bucket-path 'gs://migration/data/cockroach' ...
Alternatively, set
--use-implicit-auth
to use implicit authentication. When using assume role authentication, specify the service account with--assume-role
. For example:--bucket-path 'gs://migration/data/cockroach --use-implicit-auth --assume-role 'user-test@cluster-ephemeral.iam.gserviceaccount.com'
Ensure the Google Cloud Storage bucket is created and accessible by authorized roles and users only.
Azure Blob Storage
Set the following environment variables in the terminal running
molt fetch
:export AZURE_ACCOUNT_NAME='account' export AZURE_ACCOUNT_KEY='key'
You can also speicfy client and tenant credentials as environment variables:
export AZURE_CLIENT_SECRET='secret' export AZURE_TENANT_ID='id'
To run
molt fetch
in a containerized environment (e.g., Docker), pass the required environment variables using-e
. If your authentication method relies on local credential files, you may also need to volume map the host path to the appropriate location inside the container using-v
. For example:docker run \ -e AZURE_ACCOUNT_NAME='account' \ -e AZURE_ACCOUNT_KEY='key' \ -e AZURE_CLIENT_SECRET='secret' \ -e AZURE_TENANT_ID='id' \ -v ~/.azure:/root/.azure \ -it \ cockroachdb/molt fetch \ --bucket-path 'azure-blob://migration/data/cockroach' ...
Alternatively, set
--use-implicit-auth
to use implicit authentication: For example:--bucket-path 'azure-blob://migration/data/cockroach' --use-implicit-auth
This mode supports Azure managed identities and workload identities.
Ensure the Azure Blob Storage container is created and accessible by authorized roles and users only.
Table handling mode
MOLT Fetch can initialize target tables on the CockroachDB database in one of three modes using --table-handling
:
Mode | MOLT Fetch flag | Description |
---|---|---|
none |
Default mode |
|
truncate-if-exists |
--table-handling truncate-if-exists |
|
drop-on-target-and-recreate |
--table-handling drop-on-target-and-recreate |
|
- Use
none
when you need to retain existing data and schema. - Use
--table-handling truncate-if-exists
to clear existing data while preserving schema definitions. - Use
--table-handling drop-on-target-and-recreate
for initial imports or when source and target schemas differ, letting MOLT Fetch generate compatible tables automatically.
When using the drop-on-target-and-recreate
option, only PRIMARY KEY
and NOT NULL
constraints are preserved on the target tables. Other constraints, such as FOREIGN KEY
references, UNIQUE
, or DEFAULT
value expressions, are not retained.
To guide schema creation with drop-on-target-and-recreate
, you can explicitly map source types to CockroachDB types. Refer to Type mapping.
Schema and table filtering
MOLT Fetch can restrict which schemas (or users) and tables are migrated by using the following filter flags:
Filter type | Flag | Description |
---|---|---|
Schema filter | --schema-filter |
POSIX regex matching schema names to include; all matching schemas and their tables are moved. |
Table filter | --table-filter |
POSIX regex matching table names to include across all selected schemas. |
Table exclusion filter | --table-exclusion-filter |
POSIX regex matching table names to exclude across all selected schemas. |
Use --schema-filter
to migrate only the specified schemas, and refine which tables are moved using --table-filter
or --table-exclusion-filter
.
When migrating from Oracle, you must include --schema-filter
to name an Oracle schema to migrate. This prevents Fetch from attempting to load tables owned by other users. For example:
--schema-filter 'migration_schema'
Table filter userscript
When loading a subset of tables using --table-filter
, you must provide a TypeScript userscript to specify which tables to replicate.
For example, the following table_filter.ts
userscript filters change events to the specified source tables:
import * as api from "replicator@v1";
// List the source tables (matching source names and casing) to include in replication
const allowedTables = ["EMPLOYEES", "PAYMENTS", "ORDERS"];
// Update this to your target CockroachDB database and schema name
api.configureSource("defaultdb.migration_schema", {
dispatch: (doc: Document, meta: Document): Record<Table, Document[]> | null => {
// Replicate only if the table matches one of the allowed tables
if (allowedTables.includes(meta.table)) {
let ret: Record<Table, Document[]> = {};
ret[meta.table] = [doc];
return ret;
}
// Ignore all other tables
return null;
},
deletesTo: (doc: Document, meta: Document): Record<Table, Document[]> | null => {
// Optionally filter deletes the same way
if (allowedTables.includes(meta.table)) {
let ret: Record<Table, Document[]> = {};
ret[meta.table] = [doc];
return ret;
}
return null;
},
});
Pass the userscript to MOLT Replicator with the --userscript
flag:
--userscript table_filter.ts
Table filter userscript
When loading a subset of tables using --table-filter
, you must provide a TypeScript userscript to specify which tables to replicate.
For example, the following table_filter.ts
userscript filters change events to the specified source tables:
import * as api from "replicator@v1";
// List the source tables (matching source names and casing) to include in replication
const allowedTables = ["EMPLOYEES", "PAYMENTS", "ORDERS"];
// Update this to your target CockroachDB database and schema name
api.configureSource("defaultdb.migration_schema", {
dispatch: (doc: Document, meta: Document): Record<Table, Document[]> | null => {
// Replicate only if the table matches one of the allowed tables
if (allowedTables.includes(meta.table)) {
let ret: Record<Table, Document[]> = {};
ret[meta.table] = [doc];
return ret;
}
// Ignore all other tables
return null;
},
deletesTo: (doc: Document, meta: Document): Record<Table, Document[]> | null => {
// Optionally filter deletes the same way
if (allowedTables.includes(meta.table)) {
let ret: Record<Table, Document[]> = {};
ret[meta.table] = [doc];
return ret;
}
return null;
},
});
Pass the userscript to MOLT Replicator with the --userscript
flag:
--userscript table_filter.ts
Data load mode
MOLT Fetch can use either IMPORT INTO
or COPY FROM
to load data into CockroachDB:
Statement | MOLT Fetch flag | Description |
---|---|---|
IMPORT INTO |
Default mode |
|
COPY FROM |
--use-copy or --direct-copy |
|
- Use
IMPORT INTO
(the default mode) for large datasets, wide rows, or partitioned tables. - Use
--use-copy
when tables must remain online during data load. - Use
--direct-copy
only when you cannot move data to a public cloud, or want to perform local testing without intermediate storage. In this case, no intermediate file storage is used.
Fetch metrics
By default, MOLT Fetch exports Prometheus metrics at http://127.0.0.1:3030/metrics
. You can override the address with --metrics-listen-addr '{host}:{port}'
, where the endpoint will be http://{host}:{port}/metrics
.
Cockroach Labs recommends monitoring the following metrics during data load:
Metric Name | Description |
---|---|
molt_fetch_num_tables |
Number of tables that will be moved from the source. |
molt_fetch_num_task_errors |
Number of errors encountered by the fetch task. |
molt_fetch_overall_duration |
Duration (in seconds) of the fetch task. |
molt_fetch_rows_exported |
Number of rows that have been exported from a table. For example:molt_fetch_rows_exported{table="public.users"} |
molt_fetch_rows_imported |
Number of rows that have been imported from a table. For example:molt_fetch_rows_imported{table="public.users"} |
molt_fetch_table_export_duration_ms |
Duration (in milliseconds) of a table's export. For example:molt_fetch_table_export_duration_ms{table="public.users"} |
molt_fetch_table_import_duration_ms |
Duration (in milliseconds) of a table's import. For example:molt_fetch_table_import_duration_ms{table="public.users"} |
You can also use the sample Grafana dashboard to view the preceding metrics.
Metrics from the replicator
process are enabled by setting the --metricsAddr
replication flag, and are served at http://{host}:{port}/_/varz
. replicator
, import this Grafana dashboard.
Start Fetch
Perform the initial load of the source data.
Issue the MOLT Fetch 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 data load mode defaults toIMPORT INTO
.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
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 \ ~~~
The command assumes an Oracle Multitenant (CDB/PDB) source.
--source-cdb
specifies the container database (CDB) connection string.molt fetch \ --source $SOURCE \ --source-cdb $SOURCE_CDB \ --target $TARGET \ --schema-filter 'migration_schema' \ --table-filter 'employees|payments|orders' \ --bucket-path 's3://migration/data/cockroach' \ --table-handling truncate-if-exists \ ~~~
Check the output to observe
fetch
progress.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"}
{"level":"info","type":"summary","num_tables":3,"cdc_cursor":"4c658ae6-e8ad-11ef-8449-0242ac140006:1-28","time":"2025-02-10T14:28:11-05:00","message":"starting fetch"}
{"level":"info","type":"summary","num_tables":3,"cdc_cursor":"2358840","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"}
{"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":"4c658ae6-e8ad-11ef-8449-0242ac140006:1-29","time":"2025-02-10T14:28:13-05:00","message":"data import on target for table complete"}
{"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":"2358840","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"],"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"}
{"level":"info","type":"summary","fetch_id":"f5cb422f-4bb4-4bbd-b2ae-08c4d00d1e7c","num_tables":3,"tables":["migration_schema.employees","migration_schema.payments","migration_schema.payments"],"cdc_cursor":"4c658ae6-e8ad-11ef-8449-0242ac140006:1-29","net_duration_ms":6752.847625,"net_duration":"000h 00m 06s","time":"2024-03-18T12:30:37-04:00","message":"fetch complete"}
This message includes a
cdc_cursor
value. You must set the--defaultGTIDSet
replication flag to this value when starting Replicator:--defaultGTIDSet 4c658ae6-e8ad-11ef-8449-0242ac140006:1-29
{"level":"info","type":"summary","fetch_id":"f5cb422f-4bb4-4bbd-b2ae-08c4d00d1e7c","num_tables":3,"tables":["migration_schema.employees","migration_schema.payments","migration_schema.payments"],"cdc_cursor":"2358840","net_duration_ms":6752.847625,"net_duration":"000h 00m 06s","time":"2024-03-18T12:30:37-04:00","message":"fetch complete"}
The following message shows the appropriate values for the
--backfillFromSCN
and--scn
replication flags to use when starting Replicator:replication-only mode should include the following replicator flags: --backfillFromSCN 26685444 --scn 26685786
Verify the data load
Use MOLT Verify to confirm that the source and target data is consistent. This ensures that the data load was successful.
Run the MOLT Verify command, specifying the source and target connection strings and the tables to validate.
molt verify \ --source $SOURCE \ --target $TARGET \ --table-filter 'employees|payments|orders'
molt verify \ --source $SOURCE \ --target $TARGET \ --table-filter 'employees|payments|orders'
molt verify \ --source $SOURCE \ --target $TARGET \ --table-filter 'employees|payments|orders'
Note:With Oracle Multitenant deployments, while
--source-cdb
is required forfetch
, it is not necessary forverify
.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 containing a summary is written after each table is compared. For details on the summary fields, refer to the MOLT Verify 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"}
Configure Replicator
Replication flags
The following MOLT Replicator flags are required for continuous replication. For details on all available flags, refer to the MOLT Replicator documentation.
Flag | Description |
---|---|
--targetSchema |
Required. Target schema name on CockroachDB where tables will be replicated. |
--metricsAddr |
Enable Prometheus metrics at a specified {host}:{port} . Metrics are served at http://{host}:{port}/_/varz . |
Flag | Description |
---|---|
--targetSchema |
Required. Target schema name on CockroachDB where tables will be replicated. |
--defaultGTIDSet |
Required. Default GTID set for changefeed. |
--fetchMetadata |
Explicitly fetch column metadata for MySQL versions that do not support binlog_row_metadata . Requires SELECT permissions on the source database or PROCESS privileges. |
--metricsAddr |
Enable Prometheus metrics at a specified {host}:{port} . Metrics are served at http://{host}:{port}/_/varz . |
--userscript |
Path to a userscript that enables table filtering from MySQL sources. Refer to Table filter userscript. |
Replication from MySQL requires --defaultGTIDSet
, which sets the starting GTID for replication. You can find this value in the cdc_cursor
field of the fetch complete
message after the initial data load completes.
Flag | Description |
---|---|
--targetSchema |
Required. Target schema name on CockroachDB where tables will be replicated. |
--sourceSchema |
Required. Source schema name on Oracle where tables will be replicated from. |
--scn |
Required. Snapshot System Change Number (SCN) for the initial changefeed starting point. |
--backfillFromSCN |
Required. SCN of the earliest active transaction at the time of the snapshot. Ensures no transactions are skipped. |
--metricsAddr |
Enable Prometheus metrics at a specified {host}:{port} . Metrics are served at http://{host}:{port}/_/varz . |
--userscript |
Path to a userscript that enables table filtering from Oracle sources. Refer to Table filter userscript. |
Replication from Oracle requires --sourceSchema
, --scn
, and --backfillFromSCN
. The --sourceSchema
specifies the Oracle schema to replicate from, while --scn
and --backfillFromSCN
specify the snapshot SCN and the earliest active transaction SCN, respectively. You can find the SCN values in the message replication-only mode should include the following replicator flags
after the initial data load completes.
Tuning parameters
Configure the following replicator
flags to optimize replication throughput and resource usage. Test different combinations in a staging 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 |
---|---|
--parallelism |
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. |
--flushSize |
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. |
--targetApplyQueueSize |
Control memory usage during operation. Increase to allow higher throughput at the expense of memory; decrease to apply backpressure and limit memory consumption. |
--targetMaxPoolSize |
Set larger than --parallelism by a safety factor to avoid exhausting target pool connections. Cockroach Labs recommends setting parallelism to 80% of this value. |
--collapseMutations |
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. |
--enableParallelApplies |
Improve apply throughput for independent tables and table groups that share foreign key dependencies. Increases memory and target connection usage, so ensure you increase --targetMaxPoolSize or reduce --parallelism . |
--flushPeriod |
Set to the maximum allowable time between flushes (for example, 10s if data must be applied within 10 seconds). Works with --flushSize to control when buffered mutations are committed to the target. |
--quiescentPeriod |
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. |
--scanSize |
Applies to failback (replicator start ) scenarios only. Balance memory usage and throughput. Increase to read more rows at once for higher throughput, at the cost of memory pressure. Decrease to reduce memory pressure and increase stability. |
Replicator metrics
By default, MOLT Replicator exports Prometheus metrics at the address specified by --metricsAddr
(default :30005
) at the path /_/varz
. For example: http://localhost:30005/_/varz
.
Cockroach Labs recommends monitoring the following metrics during replication:
Metric Name | Description |
---|---|
source_lag_seconds_histogram |
Time between when a source transaction is committed and when its COMMIT transaction log arrives at Replicator. |
target_lag_seconds_histogram |
End-to-end lag from when a source transaction is committed to when its changes are fully written to the target CockroachDB. |
replicator_applier_mutations_staged |
Number of mutations that have been staged for application to the target database. |
replicator_applier_mutations_applied |
Number of mutations that have been successfully applied to the target database. |
You can use the Replicator Grafana dashboard to visualize these metrics.
To check MOLT Replicator health, run curl http://localhost:30005/_/healthz
. This returns a status code of 200
if Replicator is running.
Start Replicator
With initial load complete, start replication of ongoing changes on the source to CockroachDB using MOLT Replicator.
Run the
replicator
command, specifying the same--slotName
value that you configured during source database setup. 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.public \ --slotName molt_slot \ --stagingSchema _replicator \ --stagingCreateSchema \ --metricsAddr :30005 \ --verbose
Get the executed GTID set from the MySQL source, which shows what transactions have been applied on the source database. Use the
Executed_Gtid_Set
value as your--defaultGTIDSet
:-- For MySQL < 8.0: SHOW MASTER STATUS; -- For MySQL 8.0+: SHOW BINARY LOG STATUS;
+---------------+----------+--------------+------------------+-------------------------------------------+ | File | Position | Binlog_Do_DB | Binlog_Ignore_DB | Executed_Gtid_Set | +---------------+----------+--------------+------------------+-------------------------------------------+ | binlog.000005 | 197 | | | 77263736-7899-11f0-81a5-0242ac120002:1-38 | +---------------+----------+--------------+------------------+-------------------------------------------+
Run the
replicator
command, specifying the GTID set from the previous step. Use--stagingSchema
to specify a unique name for the staging database, and include--stagingCreateSchema
to have MOLT Replicator automatically create the staging database:replicator mylogical \ --sourceConn $SOURCE \ --targetConn $TARGET \ --targetSchema defaultdb.public \ --defaultGTIDSet 77263736-7899-11f0-81a5-0242ac120002:1-38 \ --stagingSchema _replicator \ --stagingCreateSchema \ --metricsAddr :30005 \ --userscript table_filter.ts \ --verbose
Tip:For MySQL versions that do not support
binlog_row_metadata
, include--fetchMetadata
to explicitly fetch column metadata. This requires additional permissions on the source MySQL database. GrantSELECT
permissions withGRANT SELECT ON source_database.* TO 'migration_user'@'localhost';
. If that is insufficient for your deployment, useGRANT PROCESS ON *.* TO 'migration_user'@'localhost';
, though this is more permissive and allows seeing processes and server status.
Run the
replicator
command, specifying the backfill and starting SCN from the checkpoint recorded during data load. Use--stagingSchema
to specify a unique name for the staging database, and include--stagingCreateSchema
to have MOLT Replicator automatically create the staging database:replicator oraclelogminer \ --sourceConn $SOURCE \ --sourcePDBConn $SOURCE_PDB \ --targetConn $TARGET \ --sourceSchema migration_schema \ --targetSchema defaultdb.public \ --backfillFromSCN 26685444 \ --scn 26685786 \ --stagingSchema _replicator \ --stagingCreateSchema \ --metricsAddr :30005 \ --userscript table_filter.ts \ --verbose
Note:When filtering out tables in a schema with a userscript, replication performance may decrease because filtered tables are still included in LogMiner queries and processed before being discarded.
Verify replication
Verify that Replicator is processing changes successfully. To do so, check the MOLT Replicator logs. Since you enabled debug logging with
--verbose
, you should see connection and row processing messages:You should see periodic primary keepalive messages:
DEBUG [Aug 25 14:38:10] primary keepalive received ReplyRequested=false ServerTime="2025-08-25 14:38:09.556773 -0500 CDT" ServerWALEnd=0/49913A58 DEBUG [Aug 25 14:38:15] 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 [Aug 25 14:40:02] upserted rows conflicts=0 duration=7.855333ms proposed=1 target="\"molt\".\"public\".\"tbl1\"" upserted=1 DEBUG [Aug 25 14:40:02] progressed to LSN: 0/49915DD0
You should see binlog syncer connection and row processing:
[2025/08/25 15:29:09] [info] binlogsyncer.go:463 begin to sync binlog from GTID set 77263736-7899-11f0-81a5-0242ac120002:1-38 [2025/08/25 15:29:09] [info] binlogsyncer.go:409 Connected to mysql 8.0.43 server INFO [Aug 25 15:29:09] connected to MySQL version 8.0.43
When rows are successfully replicated, you should see debug output like the following:
DEBUG [Aug 25 15:29:38] upserted rows conflicts=0 duration=1.801ms proposed=1 target="\"molt\".\"public\".\"tbl1\"" upserted=1 <!-- DEBUG [Aug 25 15:29:38] progressed to consistent point: 77263736-7899-11f0-81a5-0242ac120002:1-39 -->
TODO: Oracle verification logs
These messages confirm successful replication. You can disable verbose logging after verifying the connection.
Stop replication and verify data
Stop application traffic to your source database. This begins downtime.
Wait for replication to drain, which means that all transactions that occurred on the source database have been fully processed and replicated to CockroachDB. There are two 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
--metricsAddr
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]))
- When replication is caught up, you will not see new
Cancel replication to CockroachDB by entering
ctrl-c
to issue aSIGTERM
signal. This returns an exit code0
.Repeat Verify the data load to verify the updated data.
Modify the CockroachDB schema
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 PRIMARY KEY
and NOT NULL
constraints are preserved. You must manually recreate all other constraints and indexes.
For the appropriate SQL syntax, refer to ALTER TABLE ... ADD CONSTRAINT
and CREATE INDEX
. Review the best practices for creating secondary indexes on CockroachDB.
Cutover
Perform a cutover by resuming application traffic, now to CockroachDB.
Troubleshooting
Fetch issues
Fetch exits early due to mismatches
molt fetch
exits early in the following cases, and will output a log with a corresponding mismatch_tag
and failable_mismatch
set to true
:
- A source table is missing a primary key.
- A source primary key and target primary key have mismatching types.
- A
STRING
primary key has a different collation on the source and target. - A source and target column have mismatching types that are not allowable mappings.
- A target table is missing a column that is in the corresponding source table.
- A source column is nullable, but the corresponding target column is not nullable (i.e., the constraint is more strict on the target).
molt fetch
can continue in the following cases, and will output a log with a corresponding mismatch_tag
and failable_mismatch
set to false
:
- A target table has a column that is not in the corresponding source table.
- A source column has a
NOT NULL
constraint, and the corresponding target column is nullable (i.e., the constraint is less strict on the target). - A
DEFAULT
,CHECK
,FOREIGN KEY
, orUNIQUE
constraint is specified on a target column and not on the source column.
ORA-01950: no privileges on tablespace
If you receive ORA-01950: no privileges on tablespace 'USERS'
, it means the Oracle table owner (migration_schema
in the preceding examples) does not have sufficient quota on the tablespace used to store its data. By default, this tablespace is USERS
, but it can vary. To resolve this issue, grant a quota to the table owner. For example:
-- change UNLIMITED to a suitable limit for the table owner
ALTER USER migration_schema QUOTA UNLIMITED ON USERS;
No tables to drop and recreate on target
When expecting a bulk load but seeing no tables to drop and recreate on the target
, ensure the migration user has SELECT
and FLASHBACK
privileges on each table to be migrated. For example:
GRANT SELECT, FLASHBACK ON migration_schema.employees TO C##MIGRATION_USER;
GRANT SELECT, FLASHBACK ON migration_schema.payments TO C##MIGRATION_USER;
GRANT SELECT, FLASHBACK ON migration_schema.orders TO C##MIGRATION_USER;
Table or view does not exist
If the Oracle migration user lacks privileges on certain tables, you may receive errors stating that the table or view does not exist. Either use --table-filter
to limit the tables to be migrated, or grant the migration user SELECT
privileges on all objects in the schema. Refer to Create migration user on source database.
Missing redo logs or unavailable SCN
If the Oracle redo log files are too small or do not retain enough history, you may get errors indicating that required log files are missing for a given SCN range, or that a specific SCN is unavailable.
Increase the number and size of online redo log files, and verify that archived log files are being generated and retained correctly in your Oracle environment.
Missing replicator flags
If required --replicator-flags
are missing, ensure that the necessary flags for your mode are included. For details, refer to Replication flags.
Replicator lag
If the replicator
process is lagging significantly behind the current Oracle SCN, you may see log messages like: replicator is catching up to the current SCN at 5000 from 1000…
. This indicates that replication is progressing but is still behind the most recent changes on the source database.
Oracle sessions remain open after forcefully stopping molt
or replicator
If you shut down molt
or replicator
unexpectedly (e.g., with kill -9
or a system crash), Oracle sessions opened by these tools may remain active.
- Check your operating system for any running
molt
orreplicator
processes and terminate them manually. After confirming that both processes have stopped, ask a DBA to check for active Oracle sessions using:
SELECT sid, serial#, username, status, osuser, machine, program FROM v$session WHERE username = 'C##MIGRATION_USER';
Wait until any remaining sessions display an
INACTIVE
status, then terminate them using:ALTER SYSTEM KILL SESSION 'sid,serial#' IMMEDIATE;
Replace
sid
andserial#
in the preceding statement with the values returned by theSELECT
query.
Replicator issues
Unable to create publication or slot
This error occurs when the source database does not support logical replication.
Resolution: Verify that the source database supports logical replication by checking the wal_level
parameter on PostgreSQL:
SHOW wal_level;
If wal_level
is not set to logical
, update it and restart PostgreSQL:
ALTER SYSTEM SET wal_level = 'logical';
Replication slot already exists
ERROR: replication slot "molt_slot" already exists
Resolution: Either create a new slot with a different name, or drop the existing slot to start fresh:
SELECT pg_drop_replication_slot('molt_slot');
Dropping a replication slot can be destructive and delete data that is not yet replicated. Only use this if you want to restart replication from the current position.
Publication does not exist
run CREATE PUBLICATION molt_fetch FOR ALL TABLES;
Resolution: Create the publication on the source database. Ensure you also create the replication slot:
CREATE PUBLICATION molt_publication FOR ALL TABLES;
SELECT pg_create_logical_replication_slot('molt_slot', 'pgoutput');
Could not connect to PostgreSQL
could not connect to source database: failed to connect to `user=migration_user database=source_database`
Resolution: Verify the connection details including user, host, port, and database name. Ensure the database name in your --sourceConn
connection string matches exactly where you created the publication and slot. Verify you're connecting to the same host and port where you ran the CREATE PUBLICATION
and SELECT pg_create_logical_replication_slot()
commands. Check if TLS certificates need to be included in the connection URI.
Wrong replication slot name
run SELECT pg_create_logical_replication_slot('molt_slot', 'pgoutput'); in source database
Resolution: Create the replication slot or verify the correct slot name:
SELECT pg_create_logical_replication_slot('molt_slot', 'pgoutput');
Repeated binlog syncing restarts
If Replicator repeatedly restarts binlog syncing, this indicates an invalid or purged GTID.
Resolution: Verify the GTID set is valid and not purged:
-- Check if GTID is in executed set
SELECT GTID_SUBSET('your-gtid-set', @@GLOBAL.gtid_executed) AS in_executed;
-- Check if GTID is purged
SELECT GTID_SUBSET('your-gtid-set', @@GLOBAL.gtid_purged) AS in_purged;
Interpret the results as follows:
- If
in_executed
returns1
andin_purged
returns0
, the GTID is valid for replication. - If
in_purged
returns1
, the GTID has been purged and you must find a newer consistent point. - If both return
0
, the GTID doesn't exist in the records and is invalid.
If the GTID is purged or invalid, follow these steps:
Increase binlog retention by configuring
binlog_expire_logs_seconds
in MySQL or through your cloud provider:-- Increase binlog retention (example: 7 days = 604800 seconds) SET GLOBAL binlog_expire_logs_seconds = 604800;
Get a current GTID set to restart replication:
-- For MySQL < 8.0: SHOW MASTER STATUS; -- For MySQL 8.0+: SHOW BINARY LOG STATUS;
Invalid GTID format
Invalid GTIDs can occur when GTIDs are purged due to insufficient binlog retention, when connecting to a replica instead of the primary host, or when passing a GTID that has valid format but doesn't exist in the binlog history.
Resolution: Use a valid GTID from SHOW MASTER STATUS
(MySQL < 8.0) or SHOW BINARY LOG STATUS
(MySQL 8.0+) and ensure you're connecting to the primary host. If GTIDs are being purged, increase binlog retention.
Table/column names exceed 30 characters
Oracle LogMiner excludes tables and columns with names longer than 30 characters from redo logs.
Resolution: Rename tables and columns to 30 characters or fewer before migration.
Unsupported data types
LogMiner and replication do not support: - Long BLOB/CLOBs (4000+ characters) - User-defined types (UDTs) - Nested tables - Varrays
Resolution: Convert unsupported data types or exclude affected tables from replication.
LOB column UPDATE statements
UPDATE statements that only modify LOB columns are not supported by Oracle LogMiner.
Resolution: Avoid LOB-only updates during replication, or use Binary Reader for Oracle 12c.
JSONB null handling
SQL NULL and JSON null values are not distinguishable in JSON payloads during replication.
Resolution: Avoid using nullable JSONB columns where the distinction between SQL NULL and JSON null is important.
Schema drift errors
Indicates source and target schemas are mismatched:
WARNING: schema drift detected in "database"."table" at payload object offset 0: unexpected columns: column_name
Resolution: Align schemas or use userscripts to transform data.
Apply flow failures
Apply flow failures occur when the target database encounters error conditions such as unique constraint violations, target database being unavailable, or incorrect data (missing or extraneous columns) during apply operations:
WARNING: warning during tryCommit: ERROR: duplicate key value violates unique constraint
ERROR: maximum number of retries (10) exceeded
Resolution: Check target database constraints and connection stability. MOLT Replicator will log warnings for each retry attempt and surface a final error after exhausting all retry attempts, then restart the apply loop to continue processing.
CockroachDB changefeed connection issues
Connection errors when setting up changefeeds:
transient error: Post "https://localhost:30004/molt/public": dial tcp [::1]:30004: connect: connection refused
Resolution: Verify MOLT Replicator is running on the specified port and the webhook URL is correct.
Incorrect schema path errors
Schema path mismatches in changefeed URLs:
transient error: 400 Bad Request: unknown schema:
Resolution: Verify the webhook path matches your target database schema. Use /database/schema
for CockroachDB/PostgreSQL targets and /DATABASE
for MySQL/Oracle targets.
Performance troubleshooting
If MOLT Replicator appears hung or performs poorly:
Enable trace logging with
-vv
to get more visibility into the replicator's state and behavior.If MOLT Replicator is in an unknown, hung, or erroneous state, collect performance profiles to include with support tickets:
curl 'localhost:30005/debug/pprof/trace?seconds=15' > trace.out curl 'localhost:30005/debug/pprof/profile?seconds=15' > profile.out curl 'localhost:30005/debug/pprof/goroutine?seconds=15' > gr.out curl 'localhost:30005/debug/pprof/heap?seconds=15' > heap.out
Monitor lag metrics and adjust performance parameters as needed.