TiDB supports online schema change natively while MySQL often leverages tools like gh-ost. For task syncing from MySQL to TiDB, we'd better come up with a solution.

@xialuo1990 could you list ddl stmts gh-ost would execute during schema change?

Description

在全量迁移时 create table 阶段会执行 count操作，表创建完了，开始迁移数据时还会count。 所以create table阶段的 count操作可否省略掉。

Steps to reproduce the issue:
1.
2.
3.

Describe the results you received:

Describe the results you expected:

Output of ./dtle version:

(paste your output here)

**Additional information you deem important (e.g. issue happens only occasionally

• kafka.acks and kafka.transaction.id need to be explained more. Many deployments don't seem to need them by default and they can be confusing. Is this for reading your own writes? Only applies when producing and consuming in the same system.
• neo4j.batch.size is documented, but is confusingly similar and overlaps with kafka.batch.size. Clarify and probably remove neo4j.batch.size because i

Since input record offsets are stored in RocksDB, it can be difficult to debug issues which require knowing the input topic offset that corresponds to the last message consumed. This information is necessary to obtain the message that, for example, caused a crash. Some documentation on how this should be done would be very helpful.