> From time to time i got the following execption in my logifile (with
> Jaybird 1.5.5).
>
> If this error occurs once it happens again and again.
> I think it's because of the connection pool that reuses this bad
> connection. Would it be possible that the pool detects this kind of errors
> and discards the connection so that this error will not happen again ?

Pool executes periodically a ping statement for the connection (controlled
by "pingInterval" and "pingStatement" properties). In your case such
connection should be removed from the pool automatically. If this does not
work - then it's JayBird bug.

You can check this by starting the ClassicServer with few connections in the
pool and multiple threads that take connections, execute some queries and
return connections back to the pool. Then, if you simply kill one of the CS
instances, one thread should show exceptions, that should disappear after
the "pingInterval" milliseconds.

Roman