You are backing up onto a sharedfile system. Even though the backup might have ended in the source, it might not have flushed entirely to disk and thus the 2nd server might not be seeing the complete and correct backup. https://www.percona.com/blog/using-xtrabackup-on-nfs-for-mysql-backups/
Latest logs from Percona Server says “ready for connection” and no startup error. Previous messages show that the restore was recognized and an upgrade was performed between 8.0.42 and 8.0.45.
The error is only showing on some application logs. Can you please confirm if you can manually (not through application, intermediate layers nor third party tools) connect to MySQL , check schemas and execute some queries?
Only once did I try with this, & it worked after that. For recent backup & restore, it’s not working: mysqlcheck -u root -p --all-databases --auto-repair
Please execute some “select count(*) from tablename” for a couple tables and check that the data is there. Also check the restore log from xtrabackup and logs from DB to see if any errors pops up. If none then the problem is the application and not the DB
Data is available, and both application logs are attached. The application only requires DB IP configuration, but it is still unable to connect to the database.
The application only requires the DB IP for connectivity. It works fine with rsync and mysqldump-based setups, but when using XtraBackup, the application is unable to recognize the database, even though the configuration remains unchanged.
If you can manually connect to the DB but not through the application then I would suggest you take a look at existing users. Maybe as part of the restore you are modifying those
Consider creating a new user and re configuring that new user from the app side
Given that the database is up, accessible and the data is there then the problem might be something else related to the application or the way you are connecting from the app server.
I would suggest checking with the application team or testing on a different system, checking firewalls, users or anything else as the database seems to be fine
I completely agree with your observation. I checked with the application team, and as per their confirmation, both Rsync and Mysqldump are working fine from the application side, which indicates there is no issue with the application connectivity or configuration.
The issue is occurring specifically after backup and restore using XtraBackup. From the application perspective, the only required change is updating the database IP, and no other application-side modifications are needed.
This suggests the problem is likely related to the XtraBackup restore process or the restored database environment rather than the application itself.
I do not know what else to tell you since the backup finishes successful and the database is up and accessible.
I would suggest you make sure that the user/permission is correctly set and you are not doing anything different. For example users can be overwritten with xtrabackup while mysqldump does not modify the mysql.users
Else I would suggest you make a simpler test case (no data) and compare what is being changed in the database when using xtrabackup vs mysqldump
Could you please support us by taking a remote session? It would help us clearly understand if there are any additional steps required from our side, as we have performed the same procedure on different hardware, and the results were successful there.