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.
1. Connect to MySQL server. 2. Expand server node, rightclick some DB and choose "Connect". Result: Connection node is added, but connection can't be established with error message about too many errors. Look at screenshot for full message. By the way, I was able to connect to MySQL server itself. I reproduced this on Mac OS X.
Created attachment 70437 [details] Screenshot of error message.
I can't reproduce. It looks like MySQL has blocked any new connections because there were too many connection errors from that host - likely caused by a flaky network. The error you provide suggests running mysqladmin flush-hosts? If you can reproduce consistently even after flush-hosts, please specify exactly how you reproduce. I am pretty sure this is an issue specific to your environment.
I can reproduce this issue constantly, but everything was fine before. So it is a regression. Try easier steps: 1. Add MySQL server. 2. Connect to it. 3. Rightclick server's node and choose "Delete". 4. Add MySQL server and try to connect to it. Result: Error message appears. After flush-hosts everything starts to work fine, but only until next delete/add server cycle.
If this issue is rarely reproducible (only for me, for example), then I don't mind to remove fix-in-65 status whiteboard.
Works fine for me on Mac. I'll try on XP.
Works fine for me on XP too. I suspect there's something messed up with your MySQL server in that there are two issues that are only happening to you. Please try rebooting your machine and see if that helps.
Actually issue about stopping MySQL is reproducible on Windows XP with local standalone MySQL server, but this issue is reproducible on Mac OS X with remote standalone MySQL installed on Solaris. And main point that everything was fine before with this bundle. And I'm constantly able to reproduce this issue, with reboot too. So if it helps, we can try to setup VNC session.
Hi, Andrei, can you please look at this, I can't reproduce and Roman and I have very little time to collaborate. Many thanks! David
I can reproduce the issue with rave-opteron.
The cause of the problem seems to be the MySQLDatabaseServer.checkRunning() method. It uses a socket to connect to and immediately disconnect from the server, so the handshake at the MySQL protocol level that the JDBC driver normally performs doesn't take place, and the server considers the connection attempt an error. Perhaps a way to fix this would be to use the JDBC driver instead of a socket, and use the socketTimeout property. But this looks like a major change, and I think David, as the module owner, should make the call.
Thanks for reproducing and figuring this out. If someone can point me to a MySQL server that reproduces this error I would much appreciate it.
http://www.netbeans.org/nonav/issues/showattachment.cgi/70437/message.png
e18773e6403d Redid how we check for aliveness of the server, by using Andrei's suggestion of just opening a connection. Also, reduced assumptions about whether the server is "up" or "down." All we really know for sure is we are connected or not connected. Thus Start and Stop are always enabled, and Connect is always enabled except when you're connected already. Also should be handling error conditions like bad port better now - please test and let me know. I am running a fetch/merge/build/test cycle now, should be able to push in the morning (well, later in the morning :))
Integrated into 'main-golden', will be available in build *200810050201* on http://bits.netbeans.org/dev/nightly/ (upload may still be in progress) Changeset: http://hg.netbeans.org/main/rev/e18773e6403d User: David Van Couvering <davidvc@netbeans.org> Log: #148753: Can't start MySQL server from Netbeans in the latest build #149150: [65cat] cannot click the submenu properties on server due to error dialog box #148197: Connection can't be established because of too many errors
Verified with build 081005. Thanks Andrei and David that spent time on this issue. I knew that my magic machine is really magic, not lying. :)