Addressing Locking Issues in Reporting Server Post MariaDB Upgrade

After recently upgrading our production environment from Percona 5.7 to MariaDB 10.11, we've encountered a critical issues affecting the performance and stability of our "Reporting Server" service.

This database, now exhibits locking issues in one of the data centers (DCs), which we did not experience prior to the migration. Notably, no changes were made to the application layer during this transition.

Some data about the running transactions (from the provided logs):

• maria-backup was running, and was executing "BACKUP STAGE BLOCK_DDL". It had waited for the lock for 1341 seconds.
• ALTER TABLE aggregated_usage_value_history ADD PARTITION ( PARTITION p20240324 VALUES LESS THAN (739335) ) was waiting for a table metadata lock. Had waited for 4985 seconds.
• DELETE FROM aggregated_usage_value_history WHERE aggregated_usage_value_history.date = '2023-02-13' was watiting for a table metadata lock. Had waited for 4950 seconds.
• Several SELECT using aggregated_usage_value_history was waiting for table metadata locks (all waiting less than 4387 seconds)

The ALTER TABLE was the longest running transaction, but I could not see what was blocking this one.
It should not be related to maria-backup as the BLOCK_DDL was executed much later.

Killing the ALTER TABLE resolved the issue. There is no information what was blocking the ALTER TABLE