TiCDC FAQs

This document introduces the common questions that you might encounter when using TiCDC.

Note

In this document, the server address specified in cdc cli commands is --server=http://127.0.0.1:8300. When you use the command, replace the address with your actual PD address.

How do I choose `start-ts` when creating a task in TiCDC?

The start-ts of a replication task corresponds to a Timestamp Oracle (TSO) in the upstream TiDB cluster. TiCDC requests data from this TSO in a replication task. Therefore, the start-ts of the replication task must meet the following requirements:

The value of start-ts is larger than the tikv_gc_safe_point value of the current TiDB cluster. Otherwise, an error occurs when you create a task.
Before starting a task, ensure that the downstream has all data before start-ts. For scenarios such as replicating data to message queues, if the data consistency between upstream and downstream is not required, you can relax this requirement according to your application need.

If you do not specify start-ts, or specify start-ts as 0, when a replication task is started, TiCDC gets a current TSO and starts the task from this TSO.

Why can't some tables be replicated when I create a task in TiCDC?

When you execute cdc cli changefeed create to create a replication task, TiCDC checks whether the upstream tables meet the replication requirements. If some tables do not meet the requirements, some tables are not eligible to replicate is returned with a list of ineligible tables. You can choose Y or y to continue creating the task, and all updates on these tables are automatically ignored during the replication. If you choose an input other than Y or y, the replication task is not created.

How do I view the state of TiCDC replication tasks?

To view the status of TiCDC replication tasks, use cdc cli. For example:

cdc cli changefeed list --server=http://127.0.0.1:8300

The expected output is as follows:

[{
    "id": "4e24dde6-53c1-40b6-badf-63620e4940dc",
    "summary": {
      "state": "normal",
      "tso": 417886179132964865,
      "checkpoint": "2020-07-07 16:07:44.881",
      "error": null
    }
}]

checkpoint: TiCDC has replicated all data before this timestamp to downstream.
state: The state of this replication task. For more information about each state and its meaning, see Changefeed states.

Note

This feature is introduced in TiCDC 4.0.3.

How to verify if TiCDC has replicated all updates after upstream stops updating?

After the upstream TiDB cluster stops updating, you can verify if replication is complete by comparing the latest TSO timestamp of the upstream TiDB cluster with the replication progress in TiCDC. If the TiCDC replication progress timestamp is greater than or equal to the upstream TiDB cluster's TSO, then all updates have been replicated. To verify replication completeness, perform the following steps:

Get the latest TSO timestamp from the upstream TiDB cluster.
Note
Use the TIDB_CURRENT_TSO() function to get the current TSO, instead of using functions like NOW() that return the current time.
The following example uses TIDB_PARSE_TSO() to convert the TSO to a readable time format for further comparison:
```
BEGIN;
SELECT TIDB_PARSE_TSO(TIDB_CURRENT_TSO());
ROLLBACK;
```
The output is as follows:
```
+------------------------------------+
| TIDB_PARSE_TSO(TIDB_CURRENT_TSO()) |
+------------------------------------+
| 2024-11-12 20:35:34.848000         |
+------------------------------------+
```
Get the replication progress in TiCDC.
You can check the replication progress in TiCDC using one of the following methods:
- Method 1: query the checkpoint of the changefeed (recommended).
  Use the TiCDC command-line tool cdc cli to view the checkpoint for all replication tasks:
```
cdc cli changefeed list --server=http://127.0.0.1:8300
```
  The output is as follows:
```
[
  {
    "id": "syncpoint",
    "namespace": "default",
    "summary": {
      "state": "normal",
      "tso": 453880043653562372,
      "checkpoint": "2024-11-12 20:36:01.447",
      "error": null
    }
  }
]
```
  In the output, "checkpoint": "2024-11-12 20:36:01.447" indicates that TiCDC has replicated all upstream TiDB changes before this time. If this timestamp is greater than or equal to the upstream TiDB cluster's TSO obtained in step 1, then all updates have been replicated downstream.
- Method 2: query Syncpoint from the downstream TiDB.
  If the downstream is a TiDB cluster and the TiCDC Syncpoint feature is enabled, you can get the replication progress by querying the Syncpoint in the downstream TiDB.
  Note
  The Syncpoint update interval is controlled by the sync-point-interval configuration item. For the most up-to-date replication progress, use method 1.
  Execute the following SQL statement in the downstream TiDB to get the upstream TSO (primary_ts) and downstream TSO (secondary_ts):
```
SELECT * FROM tidb_cdc.syncpoint_v1;
```
  The output is as follows:
```
+------------------+------------+--------------------+--------------------+---------------------+
| ticdc_cluster_id | changefeed | primary_ts         | secondary_ts       | created_at          |
+------------------+------------+--------------------+--------------------+---------------------+
| default          | syncpoint  | 453879870259200000 | 453879870545461257 | 2024-11-12 20:25:01 |
| default          | syncpoint  | 453879948902400000 | 453879949214351361 | 2024-11-12 20:30:01 |
| default          | syncpoint  | 453880027545600000 | 453880027751907329 | 2024-11-12 20:35:00 |
+------------------+------------+--------------------+--------------------+---------------------+
```
  In the output, each row shows the upstream TiDB snapshot at primary_ts matches the downstream TiDB snapshot at secondary_ts.
  To view the replication progress, convert the latest primary_ts to a readable time format:
```
SELECT TIDB_PARSE_TSO(453880027545600000);
```
  The output is as follows:
```
+------------------------------------+
| TIDB_PARSE_TSO(453880027545600000) |
+------------------------------------+
| 2024-11-12 20:35:00                |
+------------------------------------+
```
  If the time corresponding to the latest primary_ts is greater than or equal to the upstream TiDB cluster's TSO obtained in step 1, then TiCDC has replicated all updates downstream.

What is `gc-ttl` in TiCDC?

Since v4.0.0-rc.1, PD supports external services in setting the service-level GC safepoint. Any service can register and update its GC safepoint. PD ensures that the key-value data later than this GC safepoint is not cleaned by GC.

When the replication task is unavailable or interrupted, this feature ensures that the data to be consumed by TiCDC is retained in TiKV without being cleaned by GC.

When starting the TiCDC server, you can specify the Time To Live (TTL) duration of GC safepoint by configuring gc-ttl. You can also use TiUP to modify gc-ttl. The default value is 24 hours. In TiCDC, this value means:

The maximum time the GC safepoint is retained at the PD after the TiCDC service is stopped.
When TiKV's GC is blocked by TiCDC's GC safepoint, gc-ttl indicates the maximum replication delay of a TiCDC replication task. If the delay of the replication task exceeds the value set by gc-ttl, the replication task enters into the failed state and reports the ErrGCTTLExceeded error. It cannot be recovered, and no longer blocks GC safepoint to advance.

The second behavior above is introduced in TiCDC v4.0.13 and later versions. The purpose is to prevent a replication task in TiCDC from suspending for too long, causing the GC safepoint of the upstream TiKV cluster not to continue for a long time and retaining too many outdated data versions, thus affecting the performance of the upstream cluster.

Note

In some scenarios, for example, when you use TiCDC for incremental replication after full replication with Dumpling/BR, the default 24 hours of gc-ttl may not be sufficient. You need to specify an appropriate value for gc-ttl when you start the TiCDC server.

What is the complete behavior of TiCDC garbage collection (GC) safepoint?

If a replication task starts after the TiCDC service starts, the TiCDC owner updates the PD service GC safepoint with the smallest value of checkpoint-ts among all replication tasks. The service GC safepoint ensures that TiCDC does not delete data generated at that time and after that time. If the replication task is interrupted, or manually stopped, the checkpoint-ts of this task does not change. Meanwhile, PD's corresponding service GC safepoint is not updated either.

If the replication task is suspended longer than the time specified by gc-ttl, the replication task enters the failed status and cannot be resumed. The PD corresponding service GC safepoint will continue.

The default Time-To-Live (TTL) that TiCDC sets for a service GC safepoint is 24 hours, which means that the GC mechanism does not delete the data required by TiCDC for continuing replication if the TiCDC service can be recovered within 24 hours after it is interrupted.

How to recover a replication task after it fails?

Use cdc cli changefeed query to query the error information of the replication task and fix the error as soon as possible.
Increase the value of gc-ttl to allow more time to fix the error, ensuring that the replication task does not enter the failed status due to the replication delay exceeding gc-ttl after the error is fixed.
After evaluating the impact on the system, increase the value of tidb_gc_life_time in TiDB to block GC and retain data, ensuring that the replication task does not enter the failed status due to GC cleaning data after the error is fixed.

How to understand the relationship between the TiCDC time zone and the time zones of the upstream/downstream databases?

	Upstream time zone	TiCDC time zone	Downstream time zone
Configuration method	See Time Zone Support	Configured using the `--tz` parameter when you start the TiCDC server	Configured using the `time-zone` parameter in `sink-uri`
Description	The time zone of the upstream TiDB, which affects DML operations of the timestamp type and DDL operations related to timestamp type columns.	TiCDC assumes that the upstream TiDB's time zone is the same as the TiCDC time zone configuration, and performs related operations on the timestamp column.	The downstream MySQL processes the timestamp in the DML and DDL operations according to the downstream time zone setting.

Note

Be careful when you set the time zone of the TiCDC server, because this time zone is used for converting the time type. Keep the upstream time zone, TiCDC time zone, and the downstream time zone consistent. The TiCDC server chooses its time zone in the following priority:

TiCDC first uses the time zone specified using --tz.
When --tz is not available, TiCDC tries to read the time zone set using the TZ environment variable.
When the TZ environment variable is not available, TiCDC uses the default time zone of the machine.

What is the default behavior of TiCDC if I create a replication task without specifying the configuration file in `--config`?

If you use the cdc cli changefeed create command without specifying the -config parameter, TiCDC creates the replication task in the following default behaviors:

Replicates all tables except system tables
Only replicates tables that contain valid indexes

Does TiCDC support outputting data changes in the Canal protocol?

Yes. Note that for the Canal protocol, TiCDC only supports the JSON output format, while the protobuf format is not officially supported yet. To enable Canal output, specify protocol as canal-json in the --sink-uri configuration. For example:

cdc cli changefeed create --server=http://127.0.0.1:8300 --sink-uri="kafka://127.0.0.1:9092/cdc-test?kafka-version=2.4.0&protocol=canal-json" --config changefeed.toml

Note

This feature is introduced in TiCDC 4.0.2.
TiCDC currently supports outputting data changes in the Canal-JSON format only to MQ sinks such as Kafka.

For more information, refer to TiCDC changefeed configurations.

Why does the latency from TiCDC to Kafka become higher and higher?

Check how do I view the state of TiCDC replication tasks.
Adjust the following parameters of Kafka:
- Increase the message.max.bytes value in server.properties to 1073741824 (1 GB).
- Increase the replica.fetch.max.bytes value in server.properties to 1073741824 (1 GB).
- Increase the fetch.message.max.bytes value in consumer.properties to make it larger than the message.max.bytes value.

When TiCDC replicates data to Kafka, can I control the maximum size of a single message in TiDB?

When protocol is set to avro or canal-json, messages are sent per row change. A single Kafka message contains only one row change and is generally no larger than Kafka's limit. Therefore, there is no need to limit the size of a single message. If the size of a single Kafka message does exceed Kafka's limit, refer to Why does the latency from TiCDC to Kafka become higher and higher?.

When protocol is set to open-protocol, messages are sent in batches. Therefore, one Kafka message might be excessively large. To avoid this situation, you can configure the max-message-bytes parameter to control the maximum size of data sent to the Kafka broker each time (optional, 10MB by default). You can also configure the max-batch-size parameter (optional, 16 by default) to specify the maximum number of change records in each Kafka message.

If I modify a row multiple times in a transaction, will TiCDC output multiple row change events?

No. When you modify the same row in one transaction multiple times, TiDB only sends the latest modification to TiKV. Therefore, TiCDC can only obtain the result of the latest modification.

When TiCDC replicates data to Kafka, does a message contain multiple types of data changes?

Yes. A single message might contain multiple updates or deletes, and update and delete might co-exist.

When TiCDC replicates data to Kafka, how do I view the timestamp, table name, and schema name in the output of TiCDC Open Protocol?

The information is included in the key of Kafka messages. For example:

{
    "ts":<TS>,
    "scm":<Schema Name>,
    "tbl":<Table Name>,
    "t":1
}

For more information, refer to TiCDC Open Protocol event format.

When TiCDC replicates data to Kafka, how do I know the timestamp of the data changes in a message?

You can get the unix timestamp by moving ts in the key of the Kafka message by 18 bits to the right.

How does TiCDC Open Protocol represent `null`?

In TiCDC Open Protocol, the type code 6 represents null.

Type	Code	Output Example	Note
Null	6	`{"t":6,"v":null}`

For more information, refer to TiCDC Open Protocol column type code.

How can I tell if a Row Changed Event of TiCDC Open Protocol is an `INSERT` event or an `UPDATE` event?

UPDATE event contains both "p" and "u" fields
INSERT event only contains the "u" field
DELETE event only contains the "d" field

For more information, refer to Open protocol Row Changed Event format.

How much PD storage does TiCDC use?

TiCDC uses etcd in PD to store and regularly update the metadata. Because the time interval between the MVCC of etcd and PD's default compaction is one hour, the amount of PD storage that TiCDC uses is proportional to the amount of metadata versions generated within this hour. However, in v4.0.5, v4.0.6, and v4.0.7, TiCDC has a problem of frequent writing, so if there are 1000 tables created or scheduled in an hour, it then takes up all the etcd storage and returns the etcdserver: mvcc: database space exceeded error. You need to clean up the etcd storage after getting this error. See etcd maintenance space-quota for details. It is recommended to upgrade your cluster to v4.0.9 or later versions.

Does TiCDC support replicating large transactions? Is there any risk?

TiCDC provides partial support for large transactions (more than 5 GB in size). Depending on different scenarios, the following risks might exist:

The latency of primary-secondary replication might greatly increase.
When TiCDC's internal processing capacity is insufficient, the replication task error ErrBufferReachLimit might occur.
When TiCDC's internal processing capacity is insufficient or the throughput capacity of TiCDC's downstream is insufficient, out of memory (OOM) might occur.

Since v6.2, TiCDC supports splitting a single-table transaction into multiple transactions. This can greatly reduce the latency and memory consumption of replicating large transactions. Therefore, if your application does not have a high requirement on transaction atomicity, it is recommended to enable the splitting of large transactions to avoid possible replication latency and OOM. To enable the splitting, set the value of the sink uri parameter transaction-atomicity to none.

If you still encounter an error above, it is recommended to use BR to restore the incremental data of large transactions. The detailed operations are as follows:

Record the checkpoint-ts of the changefeed that is terminated due to large transactions, use this TSO as the --lastbackupts of the BR incremental backup, and execute incremental data backup.
After backing up the incremental data, you can find a log record similar to ["Full backup Failed summary : total backup ranges: 0, total success: 0, total failed: 0"] [BackupTS=421758868510212097] in the BR log output. Record the BackupTS in this log.
Restore the incremental data.
Create a new changefeed and start the replication task from BackupTS.
Delete the old changefeed.

Does TiCDC replicate data changes caused by lossy DDL operations to the downstream?

Lossy DDL refers to DDL that might cause data changes when executed in TiDB. Some common lossy DDL operations include:

Modifying the type of a column, for example, INT -> VARCHAR
Modifying the length of a column, for example, VARCHAR(20) -> VARCHAR(10)
Modifying the precision of a column, for example, DECIMAL(10, 3) -> DECIMAL(10, 2)
Modifying the UNSIGNED or SIGNED attribute of a column, for example, INT UNSIGNED -> INT SIGNED

Before TiDB v7.1.0, TiCDC replicates DML events with identical old and new data to the downstream. When the downstream is MySQL, these DML events do not cause any data changes until the downstream receives and executes the DDL statement. However, when the downstream is Kafka or a cloud storage service, TiCDC writes a row of redundant data to the downstream.

Starting from TiDB v7.1.0, TiCDC eliminates these redundant DML events and no longer replicates them to downstream.

The default value of the time type field is inconsistent when replicating a DDL statement to the downstream MySQL 5.7. What can I do?

Suppose that the create table test (id int primary key, ts timestamp) statement is executed in the upstream TiDB. When TiCDC replicates this statement to the downstream MySQL 5.7, MySQL uses the default configuration. The table schema after the replication is as follows. The default value of the timestamp field becomes CURRENT_TIMESTAMP:

mysql root@127.0.0.1:test> show create table test;
+-------+----------------------------------------------------------------------------------+
| Table | Create Table                                                                     |
+-------+----------------------------------------------------------------------------------+
| test  | CREATE TABLE `test` (                                                            |
|       |   `id` int(11) NOT NULL,                                                         |
|       |   `ts` timestamp NOT NULL DEFAULT CURRENT_TIMESTAMP ON UPDATE CURRENT_TIMESTAMP, |
|       |   PRIMARY KEY (`id`)                                                             |
|       | ) ENGINE=InnoDB DEFAULT CHARSET=latin1                                           |
+-------+----------------------------------------------------------------------------------+
1 row in set

From the result, you can see that the table schema before and after the replication is inconsistent. This is because the default value of explicit_defaults_for_timestamp in TiDB is different from that in MySQL. See MySQL Compatibility for details.

Since v5.0.1 or v4.0.13, for each replication to MySQL, TiCDC automatically sets explicit_defaults_for_timestamp = ON to ensure that the time type is consistent between the upstream and downstream. For versions earlier than v5.0.1 or v4.0.13, pay attention to the compatibility issue caused by the inconsistent explicit_defaults_for_timestamp value when using TiCDC to replicate the time type data.

Why do `INSERT`/`UPDATE` statements from the upstream become `REPLACE INTO` after being replicated to the downstream if I set `safe-mode` to `true` when I create a TiCDC replication task?

TiCDC guarantees that all data is replicated at least once. When there is duplicate data in the downstream, write conflicts occur. To avoid this problem, TiCDC converts INSERT and UPDATE statements into REPLACE INTO statements. This behavior is controlled by the safe-mode parameter.

In versions earlier than v6.1.3, safe-mode defaults to true, which means all INSERT and UPDATE statements are converted into REPLACE INTO statements. In v6.1.3 and later versions, TiCDC can automatically determine whether the downstream has duplicate data, and the default value of safe-mode changes to false. If no duplicate data is detected, TiCDC replicates INSERT and UPDATE statements without conversion.

When the sink of the replication downstream is TiDB or MySQL, what permissions do users of the downstream database need?

When the sink is TiDB or MySQL, the users of the downstream database need the following permissions:

Select
Index
Insert
Update
Delete
Create
Drop
Alter
Create View

If you need to replicate recover table to the downstream TiDB, you should have the Super permission.

Why does TiCDC use disks? When does TiCDC write to disks? Does TiCDC use memory buffer to improve replication performance?

When upstream write traffic is at peak hours, the downstream may fail to consume all data in a timely manner, resulting in data pile-up. TiCDC uses disks to process the data that is piled up. TiCDC needs to write data to disks during normal operation. However, this is not usually the bottleneck for replication throughput and replication latency, given that writing to disks only results in latency within a hundred milliseconds. TiCDC also uses memory to accelerate reading data from disks to improve replication performance.

Why does replication using TiCDC stall or even stop after data restore using TiDB Lightning physical import mode and BR from upstream?

Currently, TiCDC is not yet fully compatible with TiDB Lightning physical import mode and BR. Therefore, avoid using TiDB Lightning physical import mode and BR on tables that are replicated by TiCDC. Otherwise, unknown errors might occur, such as TiCDC replication getting stuck, a significant spike in replication latency, or data loss.

If you need to use TiDB Lightning physical import mode or BR to restore data for some tables replicated by TiCDC, take these steps:

Remove the TiCDC replication task related to these tables.
Use TiDB Lightning physical import mode or BR to restore data separately in the upstream and downstream clusters of TiCDC.
After the restoration is complete and data consistency between the upstream and downstream clusters is verified, create a new TiCDC replication task for incremental replication, with the timestamp (TSO) from the upstream backup as the start-ts for the task. For example, assuming the snapshot timestamp of the BR backup in the upstream cluster is 431434047157698561, you can create a new TiCDC replication task using the following command:
```
cdc cli changefeed create -c "upstream-to-downstream-some-tables" --start-ts=431434047157698561 --sink-uri="mysql://root@127.0.0.1:4000? time-zone="
```

After a changefeed resumes from pause, its replication latency gets higher and higher and returns to normal only after a few minutes. Why?

When a changefeed is resumed, TiCDC needs to scan the historical versions of data in TiKV to catch up with the incremental data logs generated during the pause. The replication process proceeds only after the scan is completed. The scan process might take several to tens of minutes.

How should I deploy TiCDC to replicate data between two TiDB cluster located in different regions?

For TiCDC versions earlier than v6.5.2, it is recommended that you deploy TiCDC in the downstream TiDB cluster. If the network latency between the upstream and downstream is high, for example, more than 100 ms, the latency produced when TiCDC executes SQL statements to the downstream might increase dramatically due to the MySQL transmission protocol issues. This results in a decrease in system throughput. However, deploying TiCDC in the downstream can greatly ease this problem. After optimization, starting from TiCDC v6.5.2, it is recommended that you deploy TiCDC in the upstream TiDB cluster.

What is the order of executing DML and DDL statements?

Currently, TiCDC adopts the following order:

TiCDC blocks the replication progress of the tables affected by DDL statements until the DDL commitTS. This ensures that DML statements executed before DDL commitTS can be successfully replicated to the downstream first.
TiCDC continues with the replication of DDL statements. If there are multiple DDL statements, TiCDC replicates them in a serial manner.
After the DDL statements are executed in the downstream, TiCDC will continue with the replication of DML statements executed after DDL commitTS.

How should I check whether the upstream and downstream data is consistent?

If the downstream is a TiDB cluster or MySQL instance, it is recommended that you compare the data using sync-diff-inspector.

Replication of a single table can only be run on a single TiCDC node. Will it be possible to use multiple TiCDC nodes to replicate data of multiple tables?

Starting from v7.1.0, TiCDC supports the MQ sink to replicate data change logs at the granularity of TiKV Regions, which achieves scalable processing capability and allows TiCDC to replicate a single table with a large number of Regions. To enable this feature, you can configure the following parameter in the TiCDC configuration file:

[scheduler]
enable-table-across-nodes = true

Does TiCDC replication get stuck if the upstream has long-running uncommitted transactions?

TiDB has a transaction timeout mechanism. When a transaction runs for a period longer than max-txn-ttl, TiDB forcibly rolls it back. TiCDC waits for the transaction to be committed before proceeding with the replication, which causes replication delay.

Why can't I use the `cdc cli` command to operate a TiCDC cluster deployed by TiDB Operator?

This is because the default port number of the TiCDC cluster deployed by TiDB Operator is 8301, while the default port number of the cdc cli command to connect to the TiCDC server is 8300. When using the cdc cli command to operate the TiCDC cluster deployed by TiDB Operator, you need to explicitly specify the --server parameter, as follows:

./cdc cli changefeed list --server "127.0.0.1:8301"
[
  {
    "id": "4k-table",
    "namespace": "default",
    "summary": {
      "state": "stopped",
      "tso": 441832628003799353,
      "checkpoint": "2023-05-30 22:41:57.910",
      "error": null
    }
  },
  {
    "id": "big-table",
    "namespace": "default",
    "summary": {
      "state": "normal",
      "tso": 441872834546892882,
      "checkpoint": "2023-06-01 17:18:13.700",
      "error": null
    }
  }
]

Was this page helpful?

Download PDF Request docs changes Ask the community

TiCDC FAQs

How do I choose start-ts when creating a task in TiCDC?

Why can't some tables be replicated when I create a task in TiCDC?

How do I view the state of TiCDC replication tasks?

How to verify if TiCDC has replicated all updates after upstream stops updating?

What is gc-ttl in TiCDC?

What is the complete behavior of TiCDC garbage collection (GC) safepoint?

How to recover a replication task after it fails?

How to understand the relationship between the TiCDC time zone and the time zones of the upstream/downstream databases?

What is the default behavior of TiCDC if I create a replication task without specifying the configuration file in --config?

Does TiCDC support outputting data changes in the Canal protocol?

Why does the latency from TiCDC to Kafka become higher and higher?

When TiCDC replicates data to Kafka, can I control the maximum size of a single message in TiDB?

If I modify a row multiple times in a transaction, will TiCDC output multiple row change events?

When TiCDC replicates data to Kafka, does a message contain multiple types of data changes?

When TiCDC replicates data to Kafka, how do I view the timestamp, table name, and schema name in the output of TiCDC Open Protocol?

When TiCDC replicates data to Kafka, how do I know the timestamp of the data changes in a message?

How does TiCDC Open Protocol represent null?

How can I tell if a Row Changed Event of TiCDC Open Protocol is an INSERT event or an UPDATE event?

How much PD storage does TiCDC use?

Does TiCDC support replicating large transactions? Is there any risk?

Does TiCDC replicate data changes caused by lossy DDL operations to the downstream?

The default value of the time type field is inconsistent when replicating a DDL statement to the downstream MySQL 5.7. What can I do?

Why do INSERT/UPDATE statements from the upstream become REPLACE INTO after being replicated to the downstream if I set safe-mode to true when I create a TiCDC replication task?

When the sink of the replication downstream is TiDB or MySQL, what permissions do users of the downstream database need?

Why does TiCDC use disks? When does TiCDC write to disks? Does TiCDC use memory buffer to improve replication performance?

Why does replication using TiCDC stall or even stop after data restore using TiDB Lightning physical import mode and BR from upstream?

After a changefeed resumes from pause, its replication latency gets higher and higher and returns to normal only after a few minutes. Why?

How should I deploy TiCDC to replicate data between two TiDB cluster located in different regions?

What is the order of executing DML and DDL statements?

How should I check whether the upstream and downstream data is consistent?

Replication of a single table can only be run on a single TiCDC node. Will it be possible to use multiple TiCDC nodes to replicate data of multiple tables?

Does TiCDC replication get stuck if the upstream has long-running uncommitted transactions?

Why can't I use the cdc cli command to operate a TiCDC cluster deployed by TiDB Operator?

Was this page helpful?

How do I choose `start-ts` when creating a task in TiCDC?

What is `gc-ttl` in TiCDC?

What is the default behavior of TiCDC if I create a replication task without specifying the configuration file in `--config`?

How does TiCDC Open Protocol represent `null`?

How can I tell if a Row Changed Event of TiCDC Open Protocol is an `INSERT` event or an `UPDATE` event?

Why do `INSERT`/`UPDATE` statements from the upstream become `REPLACE INTO` after being replicated to the downstream if I set `safe-mode` to `true` when I create a TiCDC replication task?

Why can't I use the `cdc cli` command to operate a TiCDC cluster deployed by TiDB Operator?