Changing Task Parameters
DRS allows you to change task parameters based on service requirements.
Constraints
- You can change parameters of a synchronization task in the Configuration, Full, Full synchronization failed, Incremental, Incremental synchronization failed, or Paused state.
- After some parameters of a task are changed, the changes take effect only after you restart the task by referring to Restarting a Synchronization Task. In addition, operations for resetting a task, resuming a task, rebuilding a task, performing primary/standby switchover, or changing the flow control in the full synchronization phase will restart task processes for the parameter changes to take effect.
- You can change parameters only for the synchronization tasks of the following engines: For details about common task parameters, see Task Parameter Description.
- MySQL->MySQL
- MySQL->PostgreSQL
- MySQL -> GaussDB Distributed
- MySQL -> GaussDB Primary/Standby
- MySQL->GaussDB(DWS)
- MySQL->GaussDB(for MySQL)
- MySQL->Kafka
- MySQL->CSS/ES
- MySQL->Oracle
- Oracle->MySQL
- Oracle->PostgreSQL
- Oracle -> GaussDB Distributed
- Oracle -> GaussDB Primary/Standby
- Oracle->GaussDB(for MySQL)
- Oracle->DDM
- Oracle->Kafka
- Microsoft SQL Server->MySQL
- Microsoft SQL Server->PostgreSQL
- Microsoft SQL Server -> GaussDB Distributed
- Microsoft SQL Server -> GaussDB Primary/Standby
- Microsoft SQL Server->GaussDB(for MySQL)
- Microsoft SQL Server->Kafka
- Microsoft SQL Server->Microsoft SQL Server
- DB2 for LUW -> GaussDB Distributed
- DB2 for LUW -> GaussDB Primary/Standby
Procedure
- On the Data Synchronization Management page, locate the target synchronization task.
- Click the task name. The Basic Information page is displayed.
- Choose Parameters on the left pane. On the Parameters tab page, change the parameters of the current synchronization task.
Figure 1 Parameters
- Click Save or Cancel.
- To save your changes, click Save. In the displayed dialog box, click Yes.
- To cancel your changes, click Cancel. In the displayed dialog box, click Yes.
- Restart the task for the changes to take effect. You can click Restart in the upper right corner or restart the task by referring to Restarting a Synchronization Task.
In addition, operations for resetting a task, resuming a task, rebuilding a task, performing primary/standby switchover, or changing the flow control in the full synchronization phase will restart task processes for the parameter changes to take effect.
Task Parameter Description
|
Parameter |
Description |
Value Range |
Default Value |
Type |
Restart Required |
Applicable Data Flow |
|---|---|---|---|---|---|---|
|
applier_thread_num |
Number of threads for writing data to the destination database during the full synchronization phase. |
The value range depends on the task specifications.
|
The default value depends on the task specifications.
|
int |
Yes |
|
|
read_task_num |
Number of threads for reading data from the source database during the full synchronization phase. |
The value range depends on the task specifications.
|
The default value depends on the task specifications.
|
int |
Yes |
|
|
shard_task_num |
Number of sharding threads in the source database during the full synchronization phase. |
The value range depends on the task specifications.
|
The default value depends on the task specifications.
|
int |
Yes |
|
|
increment_writer_num |
Number of threads for writing data to the destination database during the incremental synchronization phase. |
The value range depends on the task specifications.
|
The default value depends on the task specifications.
|
int |
Yes |
|
|
relay_socket_timeout |
Timeout interval for obtaining logs from the source database during the incremental synchronization phase. |
30000-2147483647 ms |
30000 ms |
int |
Yes |
|
|
datamove_source_socket_timeout |
Timeout interval for obtaining data from the source database during the full synchronization phase. |
The value range depends on the DRS task engine.
|
The value range depends on the DRS task engine.
|
int |
Yes |
|
|
shard_length |
Rows per shard during full synchronization |
520000-1000000000 or 0 |
520000 |
int |
Yes |
|
|
datamove_fetchsize |
The size of the data reads from the source database during a full synchronization. |
10-10000 |
The value range depends on the DRS task engine. |
int |
Yes |
|
|
datamove_copy_mode |
The write mode used when writing data to the destination database during an incremental synchronization. The write performance of COPY is higher than that of INSERT. true indicates the COPY mode, and false indicates the INSERT mode. |
true/false |
true |
boolean |
Yes |
|
|
datamove_split_partition |
Whether to shard a partition table during full synchronization. true: Partition tables in the source database are sharded for parallel synchronization to improve the synchronization speed of large partitions. false: Partition tables are not sharded. |
true/false |
true |
boolean |
Yes |
|
|
read_log_num |
Number of threads for obtaining logs from the source database during incremental synchronization. |
1-16 |
2 |
int |
Yes |
|
|
logminer_fetchsize |
The size of the data reads when using LogMiner in scenarios where the source database is Oracle and incremental log read mode is set to LogMiner. |
10-10000 |
1000 |
int |
Yes |
|
|
deal_hot_table |
Whether to accelerate the replay of hot tables with frequent DML operations in the source database during incremental synchronization. true: The replay of hot tables is accelerated; false: The replay of hot tables is not accelerated. |
true/false |
fasle |
boolean |
Yes |
|
|
skip_all_ddl |
Whether to ignore all DDL statements during incremental synchronization. If there are a lot of irrelevant DDL statements, enable this function to improve the incremental synchronization performance. true: All DDL statements are ignored. false: All DDL statements are not ignored. |
true/false |
false |
boolean |
Yes |
|
|
replace_invalid_time |
Whether to replace the source time type with the default time when the source time type is invalid after being synchronized to the destination database during full+incremental synchronization. Default timestamp: 1970-01-01 00:00:00; default date: 1970-01-01; default time: 00:00:00. true: the source time type is replaced with the default time; false: the source time type is not replaced with the default time. |
true/false |
false |
boolean |
Yes |
|
|
replace_0x00 |
Whether to replace 0x00 characters that are not supported by the destination database with spaces during full+incremental synchronization. true: 0x00 characters are replaced with spaces; false: 0x00 characters are not replaced with spaces. |
true/false |
true |
boolean |
Yes |
|
|
replace_0xefbfbd |
Whether to replace abnormal characters with the default value '?' during full+incremental synchronization. DRS sets the character set of the JDBC client to UTF-8. During the query in the source database, abnormal bytes are converted to the UTF-8-encoded exception code EFBFBD. true: Abnormal code in UTF-8 encoding is replaced with '?'. false: Abnormal code is not replaced. |
true/false |
false |
boolean |
Yes |
|
Feedback
Was this page helpful?
Provide feedbackThank you very much for your feedback. We will continue working to improve the documentation.See the reply and handling status in My Cloud VOC.
For any further questions, feel free to contact us through the chatbot.
Chatbot
