Richard Alam
|
68f94062c1
|
- merge scott's fixes for voip threads
Conflicts:
bbb-voice/src/main/java/org/bigbluebutton/voiceconf/red5/media/SipToFlashAudioStream.java
|
2010-12-02 16:07:44 -05:00 |
|
Scott Morris
|
e0a0e510e8
|
Added posioning support the the audioByteData class and check to see if a posioned packet has been added to the queue. If so then stop consuming packets from the queue. This fixes the left over audio threads.
|
2010-12-02 12:45:53 -05:00 |
|
Richard Alam
|
e5f1536ae2
|
- add more logging on why local UDP port for audio is hanging around
|
2010-12-01 12:12:44 -05:00 |
|
Richard Alam
|
75b6f10582
|
- increase delay check time as we notice we are dropping too many packets when Asterisk/FreeSWITCH is on a different server
|
2010-11-29 10:24:42 -05:00 |
|
Richard Alam
|
9ba6b2e878
|
- add logging when FS/Asterisk is the one telling us to hangup (e.g. being kicked from the conference)
|
2010-11-26 14:27:06 -05:00 |
|
Richard Alam
|
4665abd490
|
- change the logs so it's a little bit clear on what the user is doing
|
2010-11-26 13:46:17 -05:00 |
|
Richard Alam
|
db8ba6cd14
|
- fix log format
|
2010-11-25 16:32:49 -05:00 |
|
Richard Alam
|
095f532e35
|
- add more info on log so we can correlate with the red5 error.log if the client dropped because of connection problems
|
2010-11-25 15:59:07 -05:00 |
|
Richard Alam
|
b456c92822
|
- add meaningfull logging so we can track a user when joining/leaving conference.
|
2010-11-25 15:38:38 -05:00 |
|
Richard Alam
|
c28258d1f3
|
- format log a little bit better so as not to flood logging when we go thourhg a lot of ports and fail
|
2010-11-23 17:02:29 -05:00 |
|
Richard Alam
|
98e950ce13
|
- add more debugging info
|
2010-11-23 14:51:19 -05:00 |
|
Richard Alam
|
55a10e750a
|
- aggressively try to get a local audio port
|
2010-11-23 14:42:39 -05:00 |
|
Richard Alam
|
480c3e990d
|
- change license headers for bbb-voice
|
2010-11-06 11:30:32 -04:00 |
|
Richard Alam
|
c8aa90c790
|
- cleanup
|
2010-10-27 14:32:17 -04:00 |
|
Richard Alam
|
d3bc9fd29a
|
- fix problem where audio is silent because of how we set the fake metadata timestamp
|
2010-10-05 12:25:47 -04:00 |
|
Richard Alam
|
17b3f3bdae
|
- cleanup and add comment on possible reason why Asterisk sends RTCP
|
2010-09-25 11:14:08 -04:00 |
|
Richard Alam
|
00b2759bfd
|
- add comments and fix timestamps (should be incremented based on codec not based on clock)
|
2010-09-24 10:48:37 -04:00 |
|
Richard Alam
|
a44648a515
|
- drop delayed RTP packets
- add some comments
|
2010-09-24 10:47:47 -04:00 |
|
Richard Alam
|
b5b427298a
|
- remove debug logs
|
2010-09-24 10:46:52 -04:00 |
|
Richard Alam
|
7f29dfe3b0
|
- add some debug logs to determine how long Red5 is receiveing audio packets from the client
|
2010-09-24 10:43:10 -04:00 |
|
Richard Alam
|
1f9457395a
|
- make debug logging only when debug id enabled
|
2010-09-21 14:25:29 -04:00 |
|
Richard Alam
|
54fa14a809
|
- handle (discard) RTCP packets properly
|
2010-09-21 12:21:22 -04:00 |
|
Richard Alam
|
2ab97b4727
|
- just process received rtp packet without trying to figure out the sequence number to drop
delayed packets.
|
2010-09-20 15:25:27 -04:00 |
|
Richard Alam
|
fd1a87bea9
|
- handling the 52 byte packet doesn't work quite well...we keep on hearing a clicking sound.
Reverting to the old way of throwing away incorrect size packets.
|
2010-09-20 15:23:03 -04:00 |
|
Richard Alam
|
59023f6304
|
- see what happens when handling the 52 byte ulaw packet from Asterisk
|
2010-09-20 15:19:31 -04:00 |
|
Richard Alam
|
1134104118
|
- add log to display incoming frame length when not equal to expected frame length. Need to figure out
why we are getting incorrect frame length under heavy load
|
2010-09-20 11:11:06 -04:00 |
|
Richard Alam
|
e05555f2ad
|
- if we drop 3 consecutive rtp audio packets, reset the stream to handle the next incoming packets.
This way, we avoid dropping all remaining packets resulting in a silent stream for the unfortunate user.
|
2010-09-20 10:45:41 -04:00 |
|
Richard Alam
|
2018c74a81
|
- recognize market rtppacket to handle changes in rtp seq num and timestamp in the middle of the stream
|
2010-09-19 12:57:57 -04:00 |
|
Richard Alam
|
8b171a6055
|
- clean up some more
|
2010-09-10 10:46:40 -04:00 |
|
Richard Alam
|
3bed51ca1e
|
- minor cleanup
|
2010-09-10 10:37:58 -04:00 |
|
Richard Alam
|
42433ffccb
|
- handle rolling over of sequence number. Sequence is only 16-bits (65535) and can start at any number so it
can rollover if the audio stream is too long.
|
2010-09-09 16:15:28 -04:00 |
|
Richard Alam
|
6e961caab8
|
- change log to debug
|
2010-09-09 14:32:47 -04:00 |
|
Richard Alam
|
eef0c12091
|
- init last seq num and last timestamp from first packet received. Don't assume that seq num always start at 0.
|
2010-09-09 13:33:32 -04:00 |
|
Richard Alam
|
5100fff24b
|
- investigate why a user looses incoming audio (can't hear but can still talk)
|
2010-09-09 11:29:43 -04:00 |
|
Richard Alam
|
1fe65660fd
|
- throw away delayed rtp packets
|
2010-09-07 15:08:01 -04:00 |
|
Richard Alam
|
bdfd342159
|
- add comment on where the timestamp values came from so that others won't wonder
|
2010-08-27 11:58:40 -04:00 |
|
Richard Alam
|
78ca8120c8
|
- add a recording stream hook...used it to record ulaw and speex stream flv
|
2010-08-26 13:10:51 -04:00 |
|
Richard Alam
|
8aaccc024c
|
- cleanup of printlns
|
2010-08-24 15:48:11 -04:00 |
|
Richard Alam
|
b397c6130b
|
- increment timestamps
Ulaw: 160 (RTMP -> RTP)
32 (RTP -> RTMP)
Speex WB: 320 (RTMP -> RTP)
20 (RTP -> RTMP)
|
2010-08-24 15:39:55 -04:00 |
|
Richard Alam
|
905e47e8a7
|
- generate fake metadata to fix problem when upgrading from FP 10.0 to 10.1
|
2010-08-24 15:38:59 -04:00 |
|
Richard Alam
|
0d29c05b7f
|
- cleanup
- change start of timestamp from 0 to a random number from 0-1000
|
2010-08-23 13:22:03 -04:00 |
|
Richard Alam
|
7a87fc2a53
|
- change timestamp for ulaw to 180 increments
|
2010-08-23 12:04:17 -04:00 |
|
Richard Alam
|
b574413d1e
|
- cleanup and change timestamp for speex flash to sip to 320ms
|
2010-08-22 10:12:57 -04:00 |
|
Richard Alam
|
bbd6d1904b
|
- cleanup and change speex transcoder to increment timestamp by 20ms.
|
2010-08-22 10:08:16 -04:00 |
|
Richard Alam
|
3fa480baaf
|
- use refactored RtpPacket
|
2010-08-20 13:27:44 -04:00 |
|
Richard Alam
|
32019cd622
|
- modify to use new gradle task
|
2010-08-20 11:03:32 -04:00 |
|
Richard Alam
|
e0844956b0
|
- add testng.xml for unit testing
|
2010-08-20 11:02:57 -04:00 |
|
Richard Alam
|
3edc00369e
|
- refactor RtpPacket and add unit tests
|
2010-08-20 11:01:20 -04:00 |
|
Richard Alam
|
d820c400c4
|
- add testng and easymock dependencies
|
2010-08-20 11:00:09 -04:00 |
|
Richard Alam
|
d08a1bdfdf
|
- set timestamps to increment by 20ms
|
2010-08-17 16:44:59 -04:00 |
|