Hi Tom, Glad to head the streams are performing better. I think we are making progress sorting out the deadlocks. I will post some info on what people can do to help us solve these problems in a moment in another thread.
-- Luke On 5/7/07, Tom Krcha <[EMAIL PROTECTED]> wrote: > Hi, > > I have just tested trunk r1976 on my videochat and I have two things: > 1) streams were much more reliable and quicker > 2) application deadlocked (freezed) after 20 minutes - peak was about 7 mbps > and 34% CPU od Intel Core 2 Duo > > I am using 2 shared objects - non-persistent - but this is quite weird - I > set persistancy to false and it still writes objects to file system - so > they are persistent, even if you set them to non-persistent. > > Should I stop using SharedObjects and point on messages sending and > distributing architecture using nc.call? > > I really would love to put the app on net. > > Thanks > > Tom > > 2007/5/5, Dan Rossi <[EMAIL PROTECTED]>: > > LOL somehow this got into my junk filters damn thunderbird ! Yes i know > > its annoying I actually made a blog about it, they force you to do OS > > upgrades to get the next major java version which is totally wrong, and > > its kinda freaky they call it OSX tiger when Java 1.5 of the same name > > was being released in it. So yeh I was on panther and had to upgrade > > specifically to build the red5 code, now Ill have to get the next > > release to get 6 then ? The major problems that ive experienced is that > > all the apple supplied software in Tiger dont work properly, ie Mac Mail > > couldnt handle all my email and started crashing so i moved to > > thunderbird which is just as buggy and absolutely crap support, you have > > to pay to tell them of bugs. > > > > I have a Shuttle Mini PC running Kubuntu which I compile the Java 6 > > source on after doing most the dev on osx, I then also run a windows 2k3 > > server inside vmware server on kubuntu to do the final testing inside > > the service wrapper, and then make the changes live onto the windows > > machines. So a bit of handling of many machines there ;) > > > > Luke Hubbard wrote: > > > Apple are slow at releasing Java updates, god knows why they dont let > > > sun do it. > > > Jobs is saving Java 6 for Leopard me thinks. Will make a nice bullet > > > point. > > > > > > - Luke > > > > > > On 5/2/07, *Dan Rossi* <[EMAIL PROTECTED] > > > <mailto:[EMAIL PROTECTED]>> wrote: > > > > > > Joachim Bauch wrote: > > > > To quote Luke: "Can people have a test and see if these changes, > > > > a) work for you, b) fix any the freezing or memory issue. > > > > Have not tested with Java 6 yet. Let me know if it works ok." > > > > > > I just noticed that thankyou. You dont use JDK6 then ? Is 5 more > > > stable > > > then ? > > > > > > Ive just been going a few things, I was hoping to see if this will > > > work > > > to stress test rtmp on red5 ? At least work out how many > > > connections the > > > servers can do. > > > > > > > http://labs.adobe.com/wiki/index.php/Flex_Stress_Testing_Framework > > > > > > > > > > > > > > > > > > -------- Original-Nachricht -------- > > > > Betreff: [Red5devs] Changed to Mina 1.1, and other changes. > > > > Datum: Wed, 2 May 2007 22:40:58 +0700 > > > > Von: Luke Hubbard < [EMAIL PROTECTED] > <mailto:[EMAIL PROTECTED]>> > > > > Antwort an: [EMAIL PROTECTED] <mailto:[EMAIL PROTECTED]> > > > > An: [EMAIL PROTECTED] <mailto:[EMAIL PROTECTED] > > > > > > > > > > > > > Hi All, > > > > > > > > I did some digging into the memory issue and it does seem like > > > we were > > > > leaking some buffers.. The really strange thing is its not our > > > code not > > > > releasing them, it seems its a new thing happening with mina > > > 1.0.1 and > > > > 1.1. Anyway, I hacked a solution to the problem, flushing after > > > each > > > > write and releasing the buffers manually. Only this causes another > > > > issue. Mina does release some of the buffers so those already > > > released > > > > thow an error. Ok so we could just ignore those exceptions, but it > > > > leaves a bad feeling. > > > > > > > > So.. I decided lets see what AsyncWeb does. I went and had a > > > look at the > > > > code, ported the http mina tansport to rtmp mina transport, > > > upgraded to > > > > mina 1.1 and changed to heap buffers. Heap buffers dont need to be > > > > released.. Oh joy!. The config for the setup of rtmp has changed a > > > > little, no more wiring it all together with spring, thats done > > > by the > > > > transport bean. IMHO this simplifies the config since we just > > > have a > > > > list of properties people can change. I've added these to > > > > red5.properties . Internally the tranport uses (#cpu +1) io > > > threads and > > > > 16 (is the default) event threads. Event threads can be set from > > > > red5.properties. > > > > > > > > Here is the new red5.properties > > > > > > > > # HTTP > > > > http.host=0.0.0.0 > > > > http.port=5080 > > > > # RTMP > > > > rtmp.host=0.0.0.0 > > > > rtmp.port=1935 > > > > rtmp.event_threads=16 > > > > rtmp.send_buffer_size=271360 > > > > rtmp.receive_buffer_size=65536 > > > > rtmp.ping_interval=5000 > > > > rtmp.max_inactivity=60000 > > > > # RTMPT > > > > rtmpt.host=0.0.0.0 > > > > rtmpt.port=8088 > > > > rtmpt.ping_interval=5000 > > > > rtmpt.max_inactivity=60000 > > > > > > > > Can people have a test and see if these changes, a) work for > > > you, b) fix > > > > any the freezing or memory issue. > > > > Have not tested with Java 6 yet. Let me know if it works ok. > > > > > > > > - Luke > > > > > > > > > > > > ------------------------------------------------------------------------ > > > > > > > > _______________________________________________ > > > > Red5 mailing list > > > > [email protected] <mailto:[email protected]> > > > > > http://osflash.org/mailman/listinfo/red5_osflash.org > > > > > > > > > _______________________________________________ > > > Red5 mailing list > > > [email protected] <mailto: [email protected]> > > > > http://osflash.org/mailman/listinfo/red5_osflash.org > > > > > > > > > > > > > > > -- > > > Luke Hubbard > > > codegent | coding for the people > > > http://www.codegent.com > > > > > > NMA Top 100 Interactive Agencies - Ones to watch! > > > http://www.codegent.com/top100/ > > > > > > want to know more? > > > http://www.codegent.com/showreel/ > > > > > > This e-mail may contain information which is privileged, confidential > > > and protected from disclosure. If you are not the intended recipient > > > of this e-mail, or any part of it, please delete this email and any > > > attachments immediately on receipt. You should not disclose the > > > contents to any other person or take copies. Any views expressed in > > > this message are those of the individual sender, except where the > > > sender specifically states them to be the views of codegent limited. > > > > ------------------------------------------------------------------------ > > > > > > _______________________________________________ > > > Red5 mailing list > > > [email protected] > > > http://osflash.org/mailman/listinfo/red5_osflash.org > > > > > > > > > _______________________________________________ > > Red5 mailing list > > [email protected] > > http://osflash.org/mailman/listinfo/red5_osflash.org > > > > > _______________________________________________ > Red5 mailing list > [email protected] > http://osflash.org/mailman/listinfo/red5_osflash.org > > -- Luke Hubbard codegent | coding for the people http://www.codegent.com NMA Top 100 Interactive Agencies - Ones to watch! http://www.codegent.com/top100/ want to know more? http://www.codegent.com/showreel/ This e-mail may contain information which is privileged, confidential and protected from disclosure. If you are not the intended recipient of this e-mail, or any part of it, please delete this email and any attachments immediately on receipt. You should not disclose the contents to any other person or take copies. Any views expressed in this message are those of the individual sender, except where the sender specifically states them to be the views of codegent limited. _______________________________________________ Red5 mailing list [email protected] http://osflash.org/mailman/listinfo/red5_osflash.org
