Doc Menu

TiCDC OpenAPI

Warning:

TiCDC OpenAPI is still an experimental feature. It is not recommended to use it in a production environment.

TiCDC provides the OpenAPI feature for querying and operating the TiCDC cluster, which is similar to the feature of cdc cli tool.

You can use the APIs to perform the following maintenance operations on the TiCDC cluster:

The request body and returned value of all APIs are in JSON format. The following sections describe the specific usage of the APIs.

In the following examples, the listening IP address of the TiCDC server is 127.0.0.1 and the port is 8300. You can bind a specified IP and port via --addr=ip:port when starting the TiCDC server.

API error message template

After sending an API request, if an error occurs, the returned error message is in the following format:

{
    "error_msg": "",
    "error_code": ""
}

From the above JSON output, error_msg describes the error message and error_code is the corresponding error code.

Get the status information of a TiCDC node

This API is a synchronous interface. If the request is successful, the status information of the corresponding node is returned.

Request URI

GET /api/v1/status

Example

The following request gets the status information of the TiCDC node whose IP address is 127.0.0.1 and port number is 8300.

curl -X GET http://127.0.0.1:8300/api/v1/status
{
    "version": "v5.2.0-master-dirty",
    "git_hash": "f191cd00c53fdf7a2b1c9308a355092f9bf8824e",
    "id": "c6a43c16-0717-45af-afd6-8b3e01e44f5d",
    "pid": 25432,
    "is_owner": true
}

The fields of the above output are described as follows:

  • version: The current TiCDC version number.
  • git_hash: The Git hash value.
  • id: The capture ID of the node.
  • pid: The capture process PID of the node.
  • is_owner: Indicates whether the node is an owner.

Check the health status of a TiCDC cluster

This API is a synchronous interface. If the cluster is healthy, 200 OK is returned.

Request URI

GET /api/v1/health

Example

curl -X GET http://127.0.0.1:8300/api/v1/health

Create a replication task

This API is an asynchronous interface. If the request is successful, 202 Accepted is returned. The returned result only means that the server agrees to run the command but does not guarantee that the command will be run successfully.

Request URI

POST /api/v1/changefeeds

Parameter description

Compared to the optional parameters for creating a replication task using the cli command, the optional parameters for creating such task using the API are not as complete. This API supports the following parameters.

Parameters for the request body

Parameter nameDescription
changefeed_idSTRING type. The ID of the replication task. (Optional)
start_tsUINT64 type. Specifies the start TSO of the changefeed. (Optional)
target_tsUINT64 type. Specifies the target TSO of the changefeed. (Optional)
sink_uriSTRING type. The downstream address of the replication task. (Required)
force_replicateBOOLEAN type. Determines whether to forcibly replicate the tables without unique indexes. (Optional)
ignore_ineligible_tableBOOLEAN type. Determines whether to ignore the tables that cannot be replicated. (Optional)
filter_rulesSTRING type array. The rules for table schema filtering. (Optional)
ignore_txn_start_tsUINT64 type array. Ignores the transaction of a specified start_ts. (Optional)
mounter_worker_numINT type. The mounter thread number. (Optional)
sink_configThe configuration parameters of sink. (Optional)

The meaning and format of changefeed_id, start_ts, target_ts, and sink_uri are the same as those described in the Use cdc cli to create a replication task document. For the detailed description of these parameters, see this document. Note that when you specify the certificate path in sink_uri, make sure you have uploaded the corresponding certificate to the corresponding TiCDC server.

Some other parameters in the above table are described further as follows.

force_replicate: This parameter defaults to false. When it is specified as true, TiCDC tries to forcibly replicate tables that do not have a unique index.

ignore_ineligible_table: This parameter defaults to false. When it is specified as true, TiCDC ignores tables that cannot be replicated.

filter_rules: The rules for table schema filtering, such as filter_rules = ['foo*.*','bar*.*']. For details, see the Table Filter document.

ignore_txn_start_ts: When this parameter is specified, the specified start_ts is ignored. For example, ignore-txn-start-ts = [1, 2].

mounter_worker_num: The thread number of mounter. Mounter is used to decode the data output from TiKV. The default value is 16.

The configuration parameters of sink are as follows:

{
  "dispatchers":[
    {"matcher":["test1.*", "test2.*"], "dispatcher":"ts"},
    {"matcher":["test3.*", "test4.*"], "dispatcher":"rowid"},
  ],
  "protocal":"default",
}

dispatchers: For the sink of MQ type, you can use dispatchers to configure the event dispatcher. Four dispatchers are supported: default, ts, rowid, and table. The dispatcher rules are as follows:

  • default: When multiple unique indexes (including the primary key) exist or the Old Value feature is enabled, events are dispatched in the table mode. When only one unique index (or the primary key) exists, events are dispatched in the rowid mode.
  • ts: Uses the commitTs of the row change to create the hash value and dispatch events.
  • rowid: Uses the name and value of the selected HandleKey column to create the hash value and dispatch events.
  • table: Uses the schema name of the table and the table name to create the hash value and dispatch events.

matcher: The matching syntax of matcher is the same as the filter rule syntax.

protocal: For the sink of MQ type, you can specify the protocol format of the message. Currently four protocols are supported: default, canal, avro, and maxwell. The default protocol is the TiCDC Open Protocol.

Example

The following request creates a replication task with an ID of test5 and a sink_uri of blackhome://.

curl -X POST -H "'Content-type':'application/json'" http://127.0.0.1:8300/api/v1/changefeeds -d '{"changefeed_id":"test5","sink_uri":"blackhole://"}'

If the request is successful, 202 Accepted is returned. If the request fails, an error message and error code are returned.

Remove a replication task

This API is an asynchronous interface. If the request is successful, 202 Accepted is returned. The returned result only means that the server agrees to run the command but does not guarantee that the command will be run successfully.

Request URI

DELETE /api/v1/changefeeds/{changefeed_id}

Parameter description

Path parameters

Parameter nameDescription
changefeed_idThe ID of the replication task (changefeed) to be removed.

Example

The following request removes the replication task with the ID test1.

curl -X DELETE http://127.0.0.1:8300/api/v1/changefeeds/test1

If the request is successful, 202 Accepted is returned. If the request fails, an error message and error code are returned.

Update the replication configuration

This API is an asynchronous interface. If the request is successful, 202 Accepted is returned. The returned result only means that the server agrees to run the command but does not guarantee that the command will be run successfully.

To modify the changefeed configuration, follow the steps of pause the replication task -> modify the configuration -> resume the replication task.

Request URI

PUT /api/v1/changefeeds/{changefeed_id}

Parameter description

Path parameters

Parameter nameDescription
changefeed_idThe ID of the replication task (changefeed) to be updated.

Parameters for the request body

Currently, only the following configuration can be modified via the API.

Parameter nameDescription
target_tsUINT64 type. Specifies the target TSO of the changefeed. (Optional)
sink_uriSTRING type. The downstream address of the replication task. (Optional)
filter_rulesSTRING type array. The rules for table schema filtering. (Optional)
ignore_txn_start_tsUINT64 type array. Ignores the transaction of a specified start_ts. (Optional)
mounter_worker_numINT type. The mounter thread number. (Optional)
sink_configThe configuration parameters of sink. (Optional)

The meanings of the above parameters are the same as those in the Create a replication task section. See that section for details.

Example

The following request updates the mounter_worker_num of the replication task with the ID test1 to 32.

 curl -X PUT -H "'Content-type':'application/json'" http://127.0.0.1:8300/api/v1/changefeeds/test1 -d '{"mounter_worker_num":32}'

If the request is successful, 202 Accepted is returned. If the request fails, an error message and error code are returned.

Query the replication task list

This API is a synchronous interface. If the request is successful, the basic information of all nodes in the TiCDC cluster is returned.

Request URI

GET /api/v1/changefeeds

Parameter description

Query parameters

Parameter nameDescription
stateWhen this parameter is specified, the replication status information only of this state is returned.(Optional)

The value options for state are all, normal, stopped, error, failed, and finished.

If this parameter is not specified, the basic information of replication tasks whose state is normal, stopped, or failed is returned by default.

Example

The following request queries the basic information of all replication tasks whose state is normal.

curl -X GET http://127.0.0.1:8300/api/v1/changefeeds?state=normal
[
    {
        "id": "test1",
        "state": "normal",
        "checkpoint_tso": 426921294362574849,
        "checkpoint_time": "2021-08-10 14:04:54.242",
        "error": null
    },
    {
        "id": "test2",
        "state": "normal",
        "checkpoint_tso": 426921294362574849,
        "checkpoint_time": "2021-08-10 14:04:54.242",
        "error": null
    }
]

The fields in the returned result above are described as follows:

  • id: The ID of the replication task.
  • state: The current state of the replication task.
  • checkpoint_tso: The TSO representation of the current checkpoint of the replication task.
  • checkpoint_tso: The formatted time representation of the current checkpoint of the replication task.
  • error: The error information of the replication task.

Query a specific replication task

This API is a synchronous interface. If the request is successful, the detailed information of the specified replication task is returned.

Request URI

GET /api/v1/changefeeds/{changefeed_id}

Parameter description

Path parameters

Parameter nameDescription
changefeed_idThe ID of the replication task (changefeed) to be queried.

Example

The following request queries the detailed information of the replication task with the ID test1.

curl -X GET http://127.0.0.1:8300/api/v1/changefeeds/test1
{
    "id": "test1",
    "sink_uri": "blackhole://",
    "create_time": "2021-08-10 11:41:30.642",
    "start_ts": 426919038970232833,
    "target_ts": 0,
    "checkpoint_tso": 426921014615867393,
    "checkpoint_time": "2021-08-10 13:47:07.093",
    "sort_engine": "unified",
    "state": "normal",
    "error": null,
    "error_history": null,
    "creator_version": "",
    "task_status": [
        {
            "capture_id": "d8924259-f52f-4dfb-97a9-c48d26395945",
            "table_ids": [
                63,
                65,
            ],
            "table_operations": {}
        }
    ]
}

Pause a replication task

This API is an asynchronous interface. If the request is successful, 202 Accepted is returned. The returned result only means that the server agrees to run the command but does not guarantee that the command will be run successfully.

Request URI

POST /api/v1/changefeeds/{changefeed_id}/pause

Parameter description

Path parameters

Parameter nameDescription
changefeed_idThe ID of the replication task (changefeed) to be paused.

Example

The following request pauses the replication task with the ID test1.

curl -X POST http://127.0.0.1:8300/api/v1/changefeeds/test1/pause

If the request is successful, 202 Accepted is returned. If the request fails, an error message and error code are returned.

Resume a replication task

This API is an asynchronous interface. If the request is successful, 202 Accepted is returned. The returned result only means that the server agrees to run the command but does not guarantee that the command will be run successfully.

Request URI

POST /api/v1/changefeeds/{changefeed_id}/resume

Parameter description

Path parameters

Parameter nameDescription
changefeed_idThe ID of the replication task (changefeed) to be resumed.

Example

The following request resumes the replication task with the ID test1.

curl -X POST http://127.0.0.1:8300/api/v1/changefeeds/test1/resume

If the request is successful, 202 Accepted is returned. If the request fails, an error message and error code are returned.

Query the replication subtask list

This API is a synchronous interface. If the request is successful, the basic information of all replication subtasks (processor) is returned.

Request URI

GET /api/v1/processors

Example

curl -X GET http://127.0.0.1:8300/api/v1/processors
[
    {
        "changefeed_id": "test1",
        "capture_id": "561c3784-77f0-4863-ad52-65a3436db6af"
    }
]

Query a specific replication subtask

This API is a synchronous interface. If the request is successful, the detailed information of the specified replication subtask (processor) is returned.

Request URI

GET /api/v1/processors/{changefeed_id}/{capture_id}

Parameter description

Path parameters

Parameter nameDescription
changefeed_idThe changefeed ID of the replication subtask to be queried.
capture_idThe capture ID of the replication subtask to be queried.

Example

The following request queries the detailed information of a subtask whose changefeed_id is test and capture_id is 561c3784-77f0-4863-ad52-65a3436db6af. A subtask can be indentifed by changefeed_id and capture_id.

curl -X GET http://127.0.0.1:8300/api/v1/processors/test1/561c3784-77f0-4863-ad52-65a3436db6af
{
    "checkpoint_ts": 426919123303006208,
    "resolved_ts": 426919123369066496,
    "table_ids": [
        63,
        65,
    ],
    "error": null
}

Query the TiCDC service process list

This API is a synchronous interface. If the request is successful, the basic information of all replication processes (capture) is returned.

Request URI

GET /api/v1/captures

Example

curl -X GET http://127.0.0.1:8300/api/v1/captures
[
    {
        "id": "561c3784-77f0-4863-ad52-65a3436db6af",
        "is_owner": true,
        "address": "127.0.0.1:8300"
    }
]

Evict an owner node

This API is an asynchronous interface. If the request is successful, 202 Accepted is returned. The returned result only means that the server agrees to run the command but does not guarantee that the command will be run successfully.

Request URI

POST /api/v1/owner/resign

Example

The following request evicts the current owner node of TiCDC and triggers a new round of elections to generate a new owner node.

curl -X POST http://127.0.0.1:8300/api/v1/owner/resign

If the request is successful, 202 Accepted is returned. If the request fails, an error message and error code are returned.

Manually trigger the load balancing of all tables in a replication task

This API is an asynchronous interface. If the request is successful, 202 Accepted is returned. The returned result only means that the server agrees to run the command but does not guarantee that the command will be run successfully.

Request URI

POST /api/v1/changefeeds/{changefeed_id}/tables/rebalance_table

Parameter description

Path parameters

Parameter nameDescription
changefeed_idThe ID of the replication task (changefeed) to be scheduled.

Example

The following request triggers the load balancing of all tables in the changefeed with the ID test1.

 curl -X POST http://127.0.0.1:8300/api/v1/changefeeds/test1/tables/rebalance_table

If the request is successful, 202 Accepted is returned. If the request fails, an error message and error code are returned.

Manually schedule a table to another node

This API is an asynchronous interface. If the request is successful, 202 Accepted is returned. The returned result only means that the server agrees to run the command but does not guarantee that the command will be run successfully.

Request URI

POST /api/v1/changefeeds/{changefeed_id}/tables/move_table

Parameter description

Path parameters

Parameter nameDescription
changefeed_idThe ID of the replication task (changefeed) to be scheduled.

Parameters for the request body

Parameter nameDescription
target_capture_idThe ID of the target capture.
table_idThe ID of the table to be scheduled.

Example

The following request schedules the table with the ID 49 in the changefeed with the ID test1 to the capture with the ID 6f19a6d9-0f8c-4dc9-b299-3ba7c0f216f5.

curl -X POST -H "'Content-type':'application/json'" http://127.0.0.1:8300/api/v1/changefeeds/changefeed-test1/tables/move_table -d '{"capture_id":"6f19a6d9-0f8c-4dc9-b299-3ba7c0f216f5","table_id":49}'

If the request is successful, 202 Accepted is returned. If the request fails, an error message and error code are returned.

Dynamically adjust the log level of the TiCDC server

This API is a synchronous interface. If the request is successful, 202 OK is returned.

Request URI

POST /api/v1/log

Request parameters

Parameters for the request body

Parameter nameDescription
log_levelThe log level you want to set.

log_level supports the log levels provided by zap: "debug", "info", "warn", "error", "dpanic" , "panic", and "fatal".

Example

curl -X POST -H "'Content-type':'application/json'" http://127.0.0.1:8300/api/v1/log -d '{"log_level":"debug"}'

If the request is successful, 202 OK is returned. If the request fails, an error message and error code are returned.