A redo log changefeed has too many connections(>151) while worker-count using default value

[Tammyxia]

What did you do?

• Create a changefeed with redo log enabled, worker-count is by default, redo storage is nfs.
• Run tpcc prepare 10 warehouse
• Umount nfs storage for 5min, then mount nfs as before.

What did you expect to see?

Changefeed works normally

What did you see instead?

• Changefeed checkpoint ts don't move forward after umount nfs, cdc log has ERROR: Error 1040: Too many connections. Check downstream mysql connection: "show process list" display there's 152 rows which is reached the default max_connections=151. --> Why cdc has so many connections with mysql? the expected connection is equal to worker-count (16 by default).
• cli changefeed pause or remove get stuck, cli command will never return.

Versions of the cluster

Upstream TiDB cluster version (execute SELECT tidb_version(); in a MySQL client):

(paste TiDB cluster version here)

Upstream TiKV version (execute tikv-server --version):

(paste TiKV version here)

TiCDC version (execute cdc version):

(paste TiCDC version here)
6.1.0

[fubinzh]

/severity Major

[nongfushanquan]

#6882.

[nongfushanquan]

/close

[ti-chi-bot]

@nongfushanquan: Closing this issue.

In response to this:

/close

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.