> I assume you are using one connection per thread (or at least: not share
> connections with multiple threads).
>
> This problem can be avoided by explicitly closing statements instead of
> letting the finalizer handle this. As too the cause of the problem, I
> have no idea. It looks like Jaybird is waiting for a response of the server.

Yes, you are correct - Jaybird is wating for a packet from the server. I
cannot explain why the server is not returning data for the
isc_dsql_free_statement call - in my understanding there should be no
issues there. In other cases that could be, for example, a transaction
that is waiting for another one to complete.

And, it can always be a bug... but at the moment I do not think it is a
bug...

Roman