Hopefully, quick questions for the experts :): 1. should hibernate and queryexecute connections automatically be closed when used inside runAsync? 2. Same question for use of cfthread. Asking because seems as if database connections / connection pools are being left open / abandoned resulting in “max connections reached, additional reserved for super user” errors since upgrading from lucee 5 to lucee 6 (all stable + RC) and change of one cfthread to runAsync equivalent.

cfquery borrows and returns from the pool each time it's used, queryExecute is just a cfml wrapper around cfquery

If used in runAsync for a very long running query , any chance the connection wouldn’t get released back? Eg if request timeout is less than length of query

possibly....

And or https://dev.lucee.org/t/db-connection-idle-timeout-isnt-honored-in-lucee-6/13841/8

I've created https://luceeserver.atlassian.net/browse/LDEV-5538

Will do. Did some more debugging. Seems related to entityload (or ormexecutequery) in a thread that’s called by ORM event handler. The result set is a collection of empty structs and , although no error is thrown, seems as if the connection isn’t released when the thread finishes

Which Orm Extension are you using?

Hibernate ORM 3.5.5.89

Have you tried the Ortus ext? It's had a lot of updates

Briefly but I guess I didn’t set it up correctly because got a “no ORM extension” found.

Oh that's a strange bug, with 6.2 and 3.5.5?

6.2 for sure. Got ortus ORM working and same thing happens (both with regard to abandoned connections as well as the /string as timestamp issue)

please file a bug! would you be able to adapt one of this existing test to repo the issue and file a PR? here's a good template testcase https://github.com/lucee/Lucee/blob/6.2/test/tickets/LDEV4461.cfc https://github.com/lucee/Lucee/tree/6.2/test/tickets/LDEV4461

I’ll try to set up test case next few days when I have some free time

thanks for getting back to me on this