DM synchronization: Can one server with one worker only support one data source? Can DM deployed on one server synchronize multiple data sources to one target data source?

Note:
This topic has been translated from a Chinese forum by GPT and might contain errors.

Original topic: DM同步,一台服务器一个worker,是否只能支持一个数据源?DM一台服务器部署能否同步多个数据源到一个目标端数据源?

| username: mm喝茶吧

[TiDB Usage Environment] Production Environment / Testing / PoC
[TiDB Version]
[Reproduction Path] What operations were performed when the issue occurred
[Encountered Issue: Issue Phenomenon and Impact]
[Resource Configuration]
[Attachments: Screenshots / Logs / Monitoring]
Currently, DM is deployed as a single node and only synchronizes one data source. Now, there are multiple data sources to be synchronized to the same target instance. Is there any way to achieve this?

| username: redgame | Original post link

It seems like you can have multiple sources, just configure multiple ones.

| username: mm喝茶吧 | Original post link

I currently have a worker, but when I configure a second source, it shows that the worker cannot be bound.

| username: okenJiang | Original post link

A worker can only match one source. If you want to configure multiple sources, you need to start multiple workers first.

| username: tidb狂热爱好者 | Original post link

Deploy several more virtual machines.

| username: zhanggame1 | Original post link

Add more workers.

| username: 天蓝色的小九 | Original post link

Just add a worker.

| username: Jellybean | Original post link

Q: Can one server support only one data source with one worker?
A: One server can deploy multiple workers, each worker being an independent process. A single worker can support only one data source. Internally, it integrates the old syncer binlog synchronization tool, which is used to synchronize a single MySQL data source to a downstream single MySQL or TiDB.

Q: Can DM deploy multiple data sources on one server to synchronize to one target data source?
A: One server can deploy multiple workers to synchronize multiple data sources to one target TiDB, as long as your server has sufficient memory, CPU, and bandwidth resources.

| username: dba-kit | Original post link

In fact, Li Wenjie has explained it very clearly. Each worker corresponds to a process. As long as there are no port conflicts and sufficient resources, it is entirely possible to run multiple workers on a single machine. However, it is best to deploy them on multiple machines. Even if one machine goes down, you can switch to the idle workers on other machines without any manual intervention.

| username: system | Original post link

This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.