When using the dumpling tool to export data, is it by default reading data from the follower nodes of TiKV?

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

Original topic: 使用dumpling工具导数据时,默认是从TIKV的follower节点上读取数据吗?

| username: Hacker_7b2KWuuo

When using the dumpling tool to export data, does it read data from the follower nodes of TiKV by default, or does it force reading from the leader nodes? I didn’t see this point mentioned in the command help parameters.

| username: 像风一样的男子 | Original post link

Dumpling is a logical export tool and does not read data directly from KV. If you need to read data directly from KV, you can use BR for backup.

| username: Kamner | Original post link

Since it is a logical export, it is all possible, and data consistency is controlled by the parameter --consistency.

| username: zhaokede | Original post link

Logical data export, data consistency

| username: 啦啦啦啦啦 | Original post link

The main thing to check is whether Follower Read is enabled.

| username: Hacker_7b2KWuuo | Original post link

Thank you, this is exactly the answer I wanted.

| username: TiDBer_JUi6UvZm | Original post link

Indeed, I haven’t seen the information. Connecting to the TiDB Server operation can ensure data consistency. As for how to control whether to use the Leader or Follower, for now, pay attention to what was mentioned above and control it through Follower Read. Mark.

| username: dba远航 | Original post link

Enable Follower Read

| username: terry0219 | Original post link

I have a question: follower read is disabled by default. Is it recommended to enable it in a production environment?

| username: 啦啦啦啦啦 | Original post link

There are use cases, you can refer to this post:

| username: terry0219 | Original post link

If there is no read hotspot, use the default value. If there is a read hotspot, enabling the follower read feature can be understood this way, right?

| username: 啦啦啦啦啦 | Original post link

Personally, I think if the read requests haven’t reached the cluster’s bottleneck, there’s no need to enable it.

| username: terry0219 | Original post link

Got it, thank you.

| username: 舞动梦灵 | Original post link

Dumpling is a logical export, similar to using mysqldump for export. It reads the SQL and writes it to a specified local file. Logical export accesses TiDB, with PD designating TiKV to transfer the data.

| username: wangkk2024 | Original post link

Learn a bit.

| username: system | Original post link

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