Reply to thread

Message: <blockquote data-quote="paalkr" data-source="post: 711303" data-attributes="member: 78371">Hi Stepko!I have tried to analyse the log files  according to your suggestions. Unfortunately I didn't really understand what you would like me to do. But from what I have seen there are often a gap in the sequence numbers just after the IPTVfilter passes data to the next filter.[CODE]01-02-2011 22:01:28.063 [bb8]RTP: Sending data to next filter: Len: 7896, firstTS: 0, lastTS: 590726893, firstSeqNr: 50509, lastSeqNr: 50510, continous: 001-02-2011 22:01:28.063 [bb8]RTP: Sending data to next filter: Len: 7896, firstTS: 0, lastTS: 590726893, continous: 001-02-2011 22:01:28.063 [bb8]RTP: Got unContinous packet!01-02-2011 22:01:28.063 [bb8]RTP: Exiting Fillbuffer01-02-2011 22:01:28.065 [bb8]Enter FillBuffer. Protokoll is 201-02-2011 22:01:28.066 [bb8]RTP: Enter CheckDataCanBePosted: 1024001-02-2011 22:01:28.066 [bb8]RTP: Received bytes: 132801-02-2011 22:01:28.066 [bb8]RTP: Enter addPacket. SequenceNr: 5108901-02-2011 22:01:28.067 [bb8]RTP: Enter CheckDataCanBePosted: 1024001-02-2011 22:01:28.067 [bb8]RTP: Received bytes: 132801-02-2011 22:01:28.067 [bb8]RTP: Enter addPacket. SequenceNr: 5109501-02-2011 22:01:28.069 [bb8]RTP: Enter CheckDataCanBePosted: 1024001-02-2011 22:01:28.072 [bb8]RTP: Add dummy packet at beginning. LastSeqNr: 50510, new firstSequenceNr: 5051501-02-2011 22:01:28.073 [bb8]RTP: Add dummy packet at beginning. LastSeqNr: 50510, new firstSequenceNr: 5051401-02-2011 22:01:28.073 [bb8]RTP: Add dummy packet at beginning. LastSeqNr: 50510, new firstSequenceNr: 5051301-02-2011 22:01:28.073 [bb8]RTP: Add dummy packet at beginning. LastSeqNr: 50510, new firstSequenceNr: 5051201-02-2011 22:01:28.074 [bb8]RTP: Got unContinous packet. LastSequenceNr: 50510, currentSequenceNr: 5051101-02-2011 22:01:28.074 [bb8]RTP: Sending data to next filter: Len: 9212, firstTS: 0, lastTS: 590727289, firstSeqNr: 50511, lastSeqNr: 50517, continous: 001-02-2011 22:01:28.074 [bb8]RTP: Sending data to next filter: Len: 9212, firstTS: 0, lastTS: 590727289, continous: 001-02-2011 22:01:28.074 [bb8]RTP: Got unContinous packet![/CODE]The first couple of seconds in each logs seems to be OK though, and that indicated packet loss due to the windsock buffer running out of space while the next filters are handling the forwarded packet, if I understood you correct. So the CPU load of the IPTVfilter it self is not the issue, but the time spent for a packet to travel the "whole graph" takes to long for the windsock buffer to hold all new packets that arrives in the mean time. Sorry, I guess this could be rephrased in a better way but my English is far from perfect.BTW. I'm planning to building my new AMD Athlon II X4 640 (3.0 GHz) based TV server with a SSD drive this weekend, so maybe the increase in CPU power will lead to a faster processing of packets (by the other filters), and then maybe the windsock buffer don't run out of space. I will report back.So, what happens next? Are there anything more I can contribute with on a short term?Do you think a larger windsock buffer could help? Is that hard to implement for a testing purpose? <img src="data:image/gif;base64,R0lGODlhAQABAIAAAAAAAP///yH5BAEAAAAALAAAAAABAAEAAAIBRAA7" class="smilie smilie--sprite smilie--sprite5" alt=":confused:" title="Confused    :confused:" loading="lazy" data-shortname=":confused:" /> I guess making the IPTV filter multi threaded for handling received packets while other filters are in action, is a bigger task. ;|Regards,PK</blockquote>

Verification