When using the JDBC backend and a lot of messages are stored in the table LARGE_MESSAGES (in my case > 1GB and > 1000 messages) and the broker is restarted, it is no longer able to complete the startup.

On startup, the broker tries to load all the messages one by one, but because there is no corresponding index on the table, this results in a timeout exception.

The query used is:

select-file-by-filename=SELECT ID, FILENAME, EXTENSION, DATA FROM %s WHERE fileName=?

but fileName does not have an index.