This Bugzilla instance is a read-only archive of historic NetBeans bug reports. To report a bug in NetBeans please follow the project's instructions for reporting issues.
Sometimes I work from home and VPN into work. In NB, I'd open one of my database connections (Services->Databases->one-of-my-db-connections) and do some DB related work. Later, I'd close my laptop - without explicitly closing the VPN connection - and take it to work. At work, I'd open up the laptop and see that NB still shows the database connection as open. Knowing that trying anything with that connection is futile (and has, on occasion caused its own hangs), I would right-click "Disconnect". The icon for the connection now shows disconnected as it should. But if I try to reconnect right away, NB hangs forever (i.e. it no longer responds to mouse/keyboard input, but screen still refreshes, so the EDT is not hung). I think I've noticed issues like this in earlier releases of NB, so I don't think this is a regression. I'm running Mac OS X 10.9.2 with JDK 7u51.
I forgot to mention that the database connection is to an Oracle 11 database using the "Oracle Thin" DB driver.
Hello, thank you for the report. Please attach the IDE log - http://wiki.netbeans.org/FaqLogMessagesFile
Created attachment 145973 [details] Messages.log file I'm including the messages.log file at the time of the "hang". I killed NB (since I couldn't do anything with it) and restarted it before writing this bug. Hopefully it has the info you need. If not, let me know and I'll try and reproduce tonight from home.
Why do you guys mark bugs as "Resolved"/Incomplete when it isn't actually resolved? Sure, I can understand that you'd like additional details - like the message log - but shouldn't you at least wait a few days before marking it as Resolved/incomplete? I just opened the bug a couple hours ago! Secondly, what is your definition for P1 vs. P3.? I thought a hanging NB qualifies as a P1, but you changed it to a P3 - why? The reason I'm asking is simple: not everyone who submits bugs will notice (or expect!) that you changed the status - I almost missed it too and I've used NB for 10+ years! I attached the evidence you requested and almost just left it at that. Then I saw the changed status. This is definitely a new process. And I don't think it's good.
(In reply to twolf2919 from comment #4) > Why do you guys mark bugs as "Resolved"/Incomplete when it isn't actually The status Incomplete is quite new one, it is intended to be set always while the issue is waiting for additional information from reporter. > resolved? Sure, I can understand that you'd like additional details - like > the message log - but shouldn't you at least wait a few days before marking > it as Resolved/incomplete? I just opened the bug a couple hours ago! > > Secondly, what is your definition for P1 vs. P3.? I thought a hanging NB > qualifies as a P1, but you changed it to a P3 - why? There is http://wiki.netbeans.org/BugPriorityGuidelines for setting the initial priority, later the priotiry could be changed on the basis of problem evaluation. In my point of view, this issue has the priority somewhere between P2-P3, however the problem would be in Oracle JDBC driver rather than in NetBeans DB support. Best regards, Jiri
Jiri, There is "Status" and there is "Resolution". I can understand that the "Resolution" field be set to "Incomplete" while the issue is awaiting more information from the reporter, but why should the "Status" be changed? The problem is not resolved.
(In reply to twolf2919 from comment #6) > Jiri, > There is "Status" and there is "Resolution". I can understand that the > "Resolution" field be set to "Incomplete" while the issue is awaiting more > information from the reporter, but why should the "Status" be changed? The > problem is not resolved. I see but it's just nature of current version of issue tracking system on NetBeans site - it changes the status anytime when some resolution was chosen. Each resolution is mapped to some status.
======= With regard to the problem at hand: ======= It looks like the oracle driver is the problem - see this post: http://www.techper.net/2008/01/28/oracle-readtimeout-a-really-useful-oracle-driver-property/ It is written for IDEA, but the solution should be appliable to netbeans as well (you can edit the connection properties). ======= Towards closing a bug with NEEDINFO ======= I went through several DB bugs (no I'm not a netbeans developer) and after doing that, I think directly closing an incomplete bug is better, than keeping it open indefinitly. Incomplete bugs are just noise in the tracker, that distracts from the real bugs. I agree that it looks strange, but as long as bugzilla not automaticly closes bugs by a timeout, this is a good way to keep the tracker clean. Even if the reporter does not change the status after adding needed info, a message is still triggered, that informs about new facts, so I'd asume, that the responsible developer will take another look at the bug.
Thanks Matthias for the link and your opinion about closing bugs. Also thanks to Jirka for the explanation.