michaelosthege commented on issue #24526: URL: https://github.com/apache/airflow/issues/24526#issuecomment-1190560680
Thanks, @potiuk from [the documentation on `sql_engine_collation_for_ids`](https://airflow.apache.org/docs/apache-airflow/stable/configurations-ref.html#sql-engine-collation-for-ids) I kew about `dag_id`, `task_id` and `key`, but it did not mention `external_executor_id` ⚠ Also thank you for clarifying the recommended CHARSET/COLLATION: 👉 `DEFAULT CHARSET=utf8 COLLATE=utf8_unicode_ci` (as described by the [database setup docs](https://airflow.apache.org/docs/apache-airflow/stable/howto/set-up-database.html#setting-up-a-mysql-database)) 👉 `CHARACTER SET utf8mb3 COLLATE utf8mb3_bin` for `dag_id`, `task_id`, `key`, `external_executor_id` fields I found the easiest way to fix this is to fix it right in my `backup.sql` file. When applying the modified `backup.sql` some `CREATE TABLE` failed because dependent tables only got updated later. I believe this is because `mysqldump` doesn't `DROP TABLE` everything at the beginning, but only one by one. Manually doing `DROP TABLE` for these few tables works. > My recommendation is as usual here: Switch to Postgres. First things first.. If there's a way to migrate the data I'll consider it once I recovered from todays incident.. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
