Resolved TS advance slowly when creating table every 5s

[fubinzh]

What did you do?

1. Create Kafka changefeed with redo on, and flush-interval = 2s

# changefeed config
[consistent]
level = "eventual"
storage = xxx
flush-interval = 2000

2. Create 100 tables, sleep 5s between 2 create table DDLs.
3. Update changefeed to disable redo
4. Create another 100 tables, sleep 5s between 2 create table DDLs.

What did you expect to see?

Resolved TS and checkpoint ts lag should be normal.

What did you see instead?

Resoved ts advances slowly.

Versions of the cluster

CDC version:
Release Version: v7.6.0-alpha
Git Commit Hash: 684d117

[fubinzh]

/severity moderate