Note:
This topic has been translated from a Chinese forum by GPT and might contain errors.
Original topic: BR备份系统表兼容问题
[TiDB Usage Environment] Production Environment / Testing / PoC
[TiDB Version]
[Reproduction Path] What operations were performed when the issue occurred
[Encountered Issue: Problem Phenomenon and Impact]
System table incompatibility during same version BR backup migration
[Resource Configuration] Go to TiDB Dashboard - Cluster Info - Hosts and take a screenshot of this page
[Attachments: Screenshots/Logs/Monitoring]


Does the table structure upstream and downstream look the same?
This is from the source database. The above is from the target database.
It seems to be different. So the error is understandable.
My cluster is 713, and it’s like this. Is the source database upgraded from a lower version? It feels like there might be something wrong with the initialization of this metadata table. 
Did the source database start upgrading from version 4.0 and encountered issues during the upgrade?
It seems like it could be an upgrade bug. After upgrading, the tidb-server will initialize the system tables, and it feels like the attribute of the initialized system table got lost.
I think you can try manually modifying the table structure of the original table. However, I’m not sure about the risks.
It’s better not to back up the system tables with br. Use a script to migrate the account passwords instead.
I wrote one before: PingCAP/out_user.py at master · WalterWj/PingCAP · GitHub
OK, after removing the system table, it returned to normal.
Okay. I saw the code. You input the source address, account, and password, then write the account to a file, right? And then execute it on the target, right?
Military boss.
For database names like this, how to adjust the script to output
database
?
Just add `` to escape in batches. Handle it manually.