NOTICE: This version of the NSF Unidata web site (archive.unidata.ucar.edu) is no longer being updated.
Current content can be found at unidata.ucar.edu.
To learn about what's going on, see About the Archive Site.
NOTICE: This version of the NSF Unidata web site (archive.unidata.ucar.edu) is no longer being updated.
Current content can be found at unidata.ucar.edu.
To learn about what's going on, see About the Archive Site.
Hi Doug, thanks for the pointer. It's no coincidence that that archived article was from a woman who used to be on the project I am now on ;) Hmm, I'm still a little worried. Saying "try sleeping a second or two" seems a little vague, as though there is a race condition lurking. I shall put a sleep in nevertheless , and maybe poke around the source code to hunt down any nasty condition. As a helpfull start, which classes generate new threads?? Many thanks Stuart On Wed, 17 Nov 2004, Doug Lindholm wrote: > Hi, > > There is something in the mail archive related to this. > (http://my.unidata.ucar.edu/cgi-bin/getfile?file=/content/support/help/MailArchives/visad-list/msg04606.html) > > It seems that putting a delay between the destroys and the > stopThreadPool() might help. > > Doug > > Bill Hibbard wrote: > > Hi Stuart, > > > > > >>Hi, I am new to visad but am enjoying using it, except for a lack of > >>understanding in the thread model. > >> > >>I am doing some offscreen image creation using DisplayImpl.getImage(true), > >>have added data for a number of FlatFields. The images come out great. > >>I should point out that I am using the same DisplayImpls over and over for > >>all images. I'm using just 2D rendering. > >> > >>However, when I want to shutdown the visad part of my app, sometimes it > >>hangs. I'm doing this: > >> > >>displayImplA.destroy(); > >>displayImplB.destroy(); > >>ActionImpl.stopThreadPool(); > >> > >>sometimes these three lines return OK. Sometimes I get an exception > >>mentioning a task queue (??) and sometimes it just hangs. This is bad, > >>since I run this job as a cron and the app must exit after it has done. > >> > >>Any help very gratefully appreciated. I apologise of this is FAQ, I > >>couldn't locate it. On a broader note, how many threads should I expect > >>if using two DisplayImpls. > > > > > > Its not a FAQ - we've never seen this problem before. > > > > We do need more information. Specifically, a stack dump with > > line numbers for the Exception. And if you can get it to hang > > while running under jdb, the output of 'where all' once it > > hangs (as I recall, you need to run 'suspend all' before you > > run 'where all'). > > > > Cheers, > > Bill > > -- > *----------------------------------------------------------------------* > | Doug Lindholm, Software Engineer | E-mail: lind@xxxxxxxx | > | Unidata Program Center | Phone: 303-497-8659 | > | UCAR Office of Programs | | > | P.O. Box 3000 | There's no place | > | Boulder, Colorado 80307-3000 | like $HOME | > *----------------------------------------------------------------------* >
visad archives: