The Software Grove

Same machine same message. Looking for an solution...

2010-03-12T13:30:21.009-05:00

Same machine same message. Looking for an solution to. Anyone?

I have a Packard Bell (Iextreme 6630)as well. When...

2010-03-03T08:06:16.108-05:00

I have a Packard Bell (Iextreme 6630)as well. When i start the media player, the screen blinks a few time, then a pop up message comes up, with a message that de video driver stopped working and has been restarted. Does anyone now how to fix this?

I was having the exact same issue as you except my...

2010-02-28T18:57:03.564-05:00

I was having the exact same issue as you except my monitor was connected to the VGA port. The only solution I found was to run the driver issued by my manufacturer Packard Bell. Unfortunately, not all the components such as the nvidia control centre had been installed.

Yeah, that could be caused by overheating. One wa...

2010-02-25T18:23:13.814-05:00

Yeah, that could be caused by overheating. One way to check the GPU temp is to download and run RealTemp. Its primary focus is CPU temperature, but it reports GPU numbers as well.

I currently have an issue with this card too, but ...

2010-02-25T18:15:05.685-05:00

I currently have an issue with this card too, but mine is really annoying. Boots every time and works great, then all of a sudden the screen freezes, restart works great for a while then the same. I feel this is a hardware problem with the chip overheating, as this plagues many nvidia cards. Would appreciate feedback if anyone has come across this on Win & 64bit...?

Ooh, that looks very bad! Good luck! Given how e...

2010-02-22T13:15:18.633-05:00

Ooh, that looks very bad! Good luck! Given how early in the boot process you're seeing problems, that may very well be a hardware problem with your card.

i just bought an nvidia GT220 and put it in. turne...

2010-02-22T13:02:16.151-05:00

i just bought an nvidia GT220 and put it in. turned on the computer and everything worked fine. installed drivers and restarted and when i got to the login screen the colors were messed up.
(http://i275.photobucket.com/albums/jj317/Neoethilan/0222000014.jpg)

running windows XP
installed driver 196.21
currently using VGA

Justin, glad to hear I could be of help. Thanks f...

2010-02-08T21:56:52.575-05:00

Justin, glad to hear I could be of help. Thanks for the offer to post about any solutions you find - I look forward to hearing about them. Good luck!

This is quite frustrating. At least (in part, than...

2010-02-08T21:53:20.643-05:00

This is quite frustrating. At least (in part, thanks to your article) I can be relatively sure that the screen-freezing is tied to the graphics driver. There seems to be no rhyme/reason as to when my pc decides to freeze.

Good to know I'm not alone in this... I'd rather not be relegated to VGA - gonna put some Google time in and try to find a solution. I'll be sure to report any good news here

Interesting... Thanks for mentioning your experie...

2010-01-20T16:08:51.366-05:00

Interesting... Thanks for mentioning your experiences with the HDMI support. I haven't tried that port on my card yet. As far as driver versions go, I'm currently on 195.81 beta. I hope they come out with something soon that fixes the problem.

I have a GT220 with the only output to the HDMI bi...

2010-01-19T20:41:01.165-05:00

I have a GT220 with the only output to the HDMI big screen TV. Took a while to get that working and I do not recall what I did specifically, but it works.
Now I switched from D-SUB/VGA connector on another new card GTS240 and bingo, DVI output does not work. Still working on the problem with the latest 196.21 driver...

I have had annoying cases happen like that as well...

2009-09-17T22:13:55.548-04:00

I have had annoying cases happen like that as well. What I do is export all of my changes to a temporary directory, blow away the working copy directory, re-import it and apply the changes I had set aside. It sucks, but it works.

Hi, not as pretty, but this one allows you to expa...

2009-08-31T12:39:36.564-04:00

Hi, not as pretty, but this one allows you to expand or collapse any element:
http://www.bodurov.com/JsonFormatter/

@Matt - I agree, you've hit the mark with your ana...

2009-08-29T20:15:09.104-04:00

@Matt - I agree, you've hit the mark with your analysis.

@Stefan - Alright. Since getting more sleep and l...

2009-08-29T00:04:47.556-04:00

@Stefan - Alright. Since getting more sleep and looking over all of your comments, I now understand how and where I got led astray.

My approach in the post relies on Statement.cancel() doing its job. The whole reason that we're pushing the JDBC operations to a worker thread is so that we can interrupt the main thread, since JDBC operations are not interruptible. Once the main thread has been interrupted, the hope is that the executing Statements are cancellable. In your situation, I gather that the Statement effectively isn't because it's blocked on I/O and doesn't respond to cancellation. That certainly would be a difficult problem to solve directly without altering the driver itself.

Although your poison pill approach may not be ideal, it may be the best available solution. At least since you involve the DB in the process, it has the chance to clean things up before severing the connection. That obviously wouldn't be the case with my hacky socket/InputStream idea (aside from whether it's even possible to access the socket). I hope updating the code for the C processes is a possibility. If they don't know how to speak your new poison pill protocol, that could be a real problem...

By the way, when I first put up the post, a colleague of mine let me know he was dealing with some issues similar to yours. I'll let him know about this comment thread and ask him to chime in with any ideas he's had.

@Matt - he, he, closing the socket from another th...

2009-08-27T17:00:29.254-04:00

@Matt - he, he, closing the socket from another thread, I actually considered that. Apart from being really ugly it wouldn't be that easy. As I said, it's the JDBC driver that does all these things, I do nothing more than calling an Oracle stored procdure (from the dbms_alert package) through a CallableStatement. How would I get access to the specific socket that is opened by the JDBC driver? I'm not aware of any way to do that.

However, it's not all bad. As the thread is blocking for a database event to occur, I simply take the "poison pill" approach and signal such an event (embedding a special "stop-message") from another thread. This way the listening thread unblocks, looks at the event message and shutdowns itself if it is the "stop-message". This seems to work cleanly so far.

What I don't like about this approach:
a) I need the help of (depend on) the DB to shutdown the thread
b) Other listeners (outside my WebApp) that may also listen on the same event must know my special "stop-message", so that they can ignore it (and we have lots of C processes in our app that may choose to listen on any event any day)

@Stefan - yeah, sorry. I guess I was responding to...

2009-08-27T12:40:41.097-04:00

@Stefan - yeah, sorry. I guess I was responding to a different scenario from the one you were describing. Out of curiosity, I experimented with closing a Socket and/or its InputStream from another thread while the main thread was blocked on a read. I had mixed success with causing an IOException to occur (unblocking the thread). It seemed to work if I was reading directly from Socket.getInputStream(), but if I wrapped that in a BufferedInputStream or BufferedReader, calling close() didn't seem to have any effect. Have you tried any approaches along those lines?

@Matt - hhm, maybe I'm dumb but I can't see how th...

2009-08-26T15:04:20.364-04:00

@Matt - hhm, maybe I'm dumb but I can't see how that would solve my problem. It's the JDBC driver who blocks on the socket waiting for either a database event to occur or to timeout after some specified timespan, 5 minutes say. It doesn't matter how cleverly the thread that runs this JDBC call is created because it won't respond to Thread.interrupt() calls. I want to take this thread down when I'm stopping my web app, that's the problem I attempted to solve cleanly with your technique. I see no way this could be done by layering Callables over Callables?

@Stefan - that's a shame. By the way, I didn't try...

2009-08-26T12:04:14.447-04:00

@Stefan - that's a shame. By the way, I didn't try to describe it in the post, but I'm also using a variant of this technique in an asynchronous setting. My main UI thread shouldn't block, so it creates/submits a Callable which in turn creates/submits the Callable that actually executes the query. It sounds complicated, but really isn't too bad.

I'd be happy to provide me more details. If you're interested, please contact me at mpassell (at) grovehillsoftware [dot] com.

@Matt - unfortunately it doesn't work in my case. ...

2009-08-26T08:38:32.922-04:00

@Matt - unfortunately it doesn't work in my case. The thread is blocked on a SocketInputStream, thus it is not interruptible :(

Thanks Stefan. I hope my approach turns out to be...

2009-08-22T20:05:10.049-04:00

Thanks Stefan. I hope my approach turns out to be as useful in your scenario as it was in mine. :)

That's really neat. I actually had almost the same...

2009-08-22T11:49:53.988-04:00

That's really neat. I actually had almost the same problem yesterday (JDBC statements blocking on a socket while waiting for a database event to occur). While I already have a solution, I'll try yours too - it looks much better than mine.

Thanks,
Stefan

@Luke - glad you liked it! @Claus - now that iBAT...

2009-08-21T10:02:05.919-04:00

@Luke - glad you liked it!

@Claus - now that iBATIS requires Java 5, that's an option without including the Doug Lea util.concurrent libraries. I haven't used iBATIS in years! How's it doing? (The product I refer to in the post might benefit from using it)

Cool, now get that into Apache iBatis 3 :)

2009-08-21T01:04:08.579-04:00

Cool, now get that into Apache iBatis 3 :)

Cute. I've wondered about how to do this nicely in...

2009-08-20T22:13:55.090-04:00

Cute. I've wondered about how to do this nicely in the past.