Issue of CDC Blocking TiKV GC

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

Original topic: cdc 阻塞tikv gc的问题

| username: Raymond

Experts, I encountered a problem.
A TiDB cluster with CDC has two synchronization tasks enabled. One of the synchronization tasks was interrupted (unable to connect to the downstream TiDB cluster), but the other synchronization task did not stop. At this time, I noticed that the GC of this TiDB cluster kept moving forward and was not blocked. So I want to ask, does GC blocking only get triggered when both synchronization tasks encounter issues? Will the interruption of only one synchronization task not affect the progress of GC?

| username: zhanggame1 | Original post link

How long does it take for the GC retention time setting to take effect? Has the time not yet arrived?

| username: weixiaobing | Original post link

CDC has its own GC-TTL settings. When a CDC task encounters an exception, it will block the cluster’s GC to prevent data from being garbage collected, which would cause the task’s synchronized data to be redone.

| username: kkpeter | Original post link

Looking at this phenomenon, it should be.

| username: redgame | Original post link

Yes, you guessed it right.