<div dir="ltr"><div><div><div><div><div><div><div>Hello All,<br><br></div>We are facing a peculiar problem in which the frequent connect/disconnect events on the PRI line results in FreeSwitch not able to handle any further incoming calls.<br>
</div>The workaround to the above problem is to <b>restart mod_freetdm</b> or the complete freeswitch.<br><br><br></div><div>I'm providing some of the logs with my analysis:<br></div><div><br>30111 2014-01-31 10:32:53.522513 [NOTICE] mod_freetdm.c:1889 Alarm raised on channel 13:15<br>
<b>30112 2014-01-31 10:32:53.522513 [DEBUG] ftmod_wanpipe.c:1630 [s3c7][3:7] read wanpipe event 17<br>30113 2014-01-31 10:32:53.522513 [NOTICE] mod_freetdm.c:1889 Alarm raised on channel 3:22<br>30114 2014-01-31 10:32:53.522513 [INFO] ftmod_sangoma_isdn_stack_rcv.c:663 [SNGISDN PHY] D-chan 3 : T1/E1/BRI Disconnected<br>
30115 2014-01-31 10:32:53.522513 [DEBUG] mod_freetdm.c:2416 got clear channel sig [ALARM_TRAP]</b><br>30116 2014-01-31 10:32:53.522513 [DEBUG] ftmod_wanpipe.c:1385 [s3c7][3:7] Ignoring wanpipe link disconnected event<br>30117 2014-01-31 10:32:53.522513 [INFO] ftmod_sangoma_isdn_stack_rcv.c:663 [SNGISDN PHY] D-chan 3 : T1/E1/BRI Disconnected<br>
30118 2014-01-31 10:32:53.522513 [DEBUG] mod_freetdm.c:2416 got clear channel sig [ALARM_TRAP]<br>30119 2014-01-31 10:32:53.522513 [DEBUG] ftmod_wanpipe.c:1630 [s3c8][3:8] read wanpipe event 17<br><br><br>30313 2014-01-31 10:32:53.682516 [INFO] ftmod_sangoma_isdn_stack_rcv.c:1004 sng_isdn->s16: Invalid Q.921/Q.931 frame - ignoring len:1<br>
30314 2014-01-31 10:32:53.722506 [DEBUG] mod_freetdm.c:866 Dropping frame! (write note ready) in channel FreeTDM/3:8/43850327 device 3:8!<br>30315 2014-01-31 10:32:53.722506 [DEBUG] mod_freetdm.c:866 Dropping frame! (write note ready) in channel FreeTDM/3:4/43850118 device 3:4!<br>
30316 2014-01-31 10:32:53.722506 [DEBUG] ftmod_wanpipe.c:945 [s14c11][14:11] Rx Queue length exceeded 80% hreshold (9/10)<br>30317 2014-01-31 10:32:53.722506 [DEBUG] ftmod_wanpipe.c:945 [s23c11][23:11] Rx Queue length exceeded 80% hreshold (9/10)<br>
30318 2014-01-31 10:32:53.722506 [DEBUG] mod_freetdm.c:866 Dropping frame! (write note ready) in channel FreeTDM/3:5/43851158 device 3:5!<br>30319 2014-01-31 10:32:53.722506 [DEBUG] ftmod_wanpipe.c:945 [s24c2][24:2] Rx Queue length exceeded 80% hreshold (9/10)<br>
30320 2014-01-31 10:32:53.722506 [DEBUG] mod_freetdm.c:866 Dropping frame! (write note ready) in channel FreeTDM/3:6/43850118 device 3:6!<br>30321 2014-01-31 10:32:53.722506 [DEBUG] ftmod_wanpipe.c:945 [s6c17][6:18] Rx Queue length exceeded 80% hreshold (9/10)<br>
30322 2014-01-31 10:32:53.742514 [DEBUG] mod_freetdm.c:866 Dropping frame! (write note ready) in channel FreeTDM/3:3/43850523 device 3:3!<br><b>30323 2014-01-31 10:32:53.762508 [WARNING] mod_freetdm.c:763 Too many timeouts while waiting I/O in channel FreeTDM/3:6/43850118 device 3:6!<br>
30324 2014-01-31 10:32:53.762508 [ERR] mod_freetdm.c:808 clearing IO in channel FreeTDM/3:6/43850118 device 3:6!</b><br>30325 2014-01-31 10:32:53.762508 [WARNING] mod_freetdm.c:763 Too many timeouts while waiting I/O in channel FreeTDM/3:1/43852488 device 3:1!<br>
30326 2014-01-31 10:32:53.762508 [ERR] mod_freetdm.c:808 clearing IO in channel FreeTDM/3:1/43852488 device 3:1!<br><br><br></div>Because of the timeout in the read channel, the TFLAG_IO gets cleared and leads to a recurrent logs w.r.t the TFLAG_IO flag (as shown in line 30413)<br>
<br>30403 2014-01-31 10:32:53.762508 [WARNING] mod_freetdm.c:763 Too many timeouts while waiting I/O in channel FreeTDM/3:8/43850327 device 3:8!<br>30404 2014-01-31 10:32:53.762508 [ERR] mod_freetdm.c:808 clearing IO in channel FreeTDM/3:8/43850327 device 3:8!<br>
30405 2014-01-31 10:32:53.762508 [DEBUG] switch_ivr_bridge.c:505 FreeTDM/3:8/43850327 ending bridge by request from read function<br>30406 2014-01-31 10:32:53.762508 [WARNING] mod_freetdm.c:763 Too many timeouts while waiting I/O in channel FreeTDM/3:4/43850118 device 3:4!<br>
30407 2014-01-31 10:32:53.762508 [ERR] mod_freetdm.c:808 clearing IO in channel FreeTDM/3:4/43850118 device 3:4!<br>30408 2014-01-31 10:32:53.762508 [DEBUG] switch_ivr_bridge.c:505 FreeTDM/3:4/43850118 ending bridge by request from read function<br>
30409 2014-01-31 10:32:53.762508 [WARNING] mod_freetdm.c:763 Too many timeouts while waiting I/O in channel FreeTDM/3:10/43115865 device 3:10!<br>30410 2014-01-31 10:32:53.762508 [ERR] mod_freetdm.c:808 clearing IO in channel FreeTDM/3:10/43115865 device 3:10!<br>
30411 2014-01-31 10:32:53.762508 [DEBUG] switch_ivr_play_say.c:1678 done playing file /srv/sounds/usr/159d/0491/-a6f/0-48/e4-8/1df-/1823/c766/9 6e8.mp3<br>30412 2014-01-31 10:32:53.762508 [DEBUG] switch_ivr_play_say.c:1306 Codec Activated L16@8000hz 1 channels 20ms<br>
<b>30413 2014-01-31 10:32:53.762508 [DEBUG] mod_freetdm.c:747 TFLAG_IO is not set in channel FreeTDM/3:10/43115865 device 3:10!</b><br>30414 2014-01-31 10:32:53.762508 [ERR] mod_freetdm.c:808 clearing IO in channel FreeTDM/3:10/43115865 device 3:10!<br>
<br></div>Upon further looking at the code, it appears that the TFLAG_IO gets cleared because of the timeout (as a result of PRI connect/disconnect), but as this flag is not inturn set upon PRI line recovery, the channel_read_frame (in mod_freetdm.c) is not able to read any further frames.<br>
<br></div>Please provide your comments in this regard and any further inputs would be much appreciated.<br><br></div>Thanks<br></div>Deepika<br><div><div><div><div><div><br></div></div></div></div></div></div>