On 3/25/2016 9:39 AM, Bergquist, Brett wrote:
Hey Kathey, that for taking the time to respond. Unfortunately we
had to bounce the network server and not it is going through the derby
recovery logs. 18000 of them at one every 10 seconds. So 5 hours
of down time L
Thanks Brett for looking at this issue and filing the issue.
Maybe next time try connecting to network server and shutting down the
database with shutdown=true first to minimize the downtime going through
the logs. I should have mentioned that in my first response.
A couple other things that came to mind.
I think recover shouldn't recover transactions that are not prepared.
I think maybe in your simple test case the session was still hanging
around and that is why you saw the transaction, but I could be wrong.
Best
Kathey
