Backup Exception

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

Original topic: br备份异常

| username: Hacker_UBVfs1Zv

[TiDB Usage Environment] Production Environment / Testing / Poc
[TiDB Version]
[Reproduction Path] What operations were performed when the issue occurred
[Encountered Issue: Issue Phenomenon and Impact]
[Resource Configuration]
[Attachments: Screenshots / Logs / Monitoring]

Using the br command resulted in an error: error: error happened in store 2 at ip:20160. File or directory not found on tikv node

| username: tidb菜鸟一只 | Original post link

Each TiKV node must create the /data/backup directory.

| username: Hacker_UBVfs1Zv | Original post link

That’s right, thank you. The official documentation is not clear on this point.

| username: 啦啦啦啦啦 | Original post link

It is best to use S3 or NFS for backups, otherwise, it will be very troublesome during restoration.

| username: redgame | Original post link

tikv create /data/backup

| username: zhanggame1 | Original post link

Backup requires each TiKV node to have the same backup path. Restoration is more complicated, as each TiKV node needs to access all backup files. Generally, NFS or other shared storage is used.

| username: linnana | Original post link

It is possible to create a backup path for TiKV nodes using Ansible.

| username: yulei7633 | Original post link

Using shared NFS for backup, there won’t be any errors during restoration. Otherwise, the restoration process can be quite troublesome.

| username: yulei7633 | Original post link

Of course, NFS is relatively slow. I have tested it before, and NFS is extremely slow.