Heiko Wundram
2014-Jun-30 09:55 UTC
[asterisk-users] Asterisk 11.10.2 update breaks snom TLS transport
Hey all, I've recently updated an Asterisk installation to 11.10.2 (from the Gentoo packaged build), and the update has (seemingly) broken TLS transport for snom phones (I could reproduce on snom 3x0 and 7x0 with 8.4.3x and 8.7.x.y firmwares). What happens is that the registration to the PBX is successful, and inbound calls work normally (i.e., when the phone is being signalled that a call is coming in), but dispatching a call from the snom phone to the PBX makes the TLS transport "hang" (i.e., the PBX doesn't see any more TLS packets from the phone, and OPTIONS is no longer replied to). The phone then kills the TLS channel after a while (probably times it out) and Asterisk marks the phone as unreachable due to QUALIFY, and after some time (roundabout two minutes) the phone reconnects the TLS channel and is reachable again. I couldn't reproduce this behaviour with other phones (i.e., Gigaset devices which I also use with TLS transport). Is this a known regression, and generally: how can I help diagnose this problem further for opening a bug-report, besides the description given above? Has anyone else encountered this and found another workaround besides switching to TCP/UDP-transport for the phones? Thanks for any hints in advance! -- --- Heiko Wundram.
Heiko Wundram
2014-Jun-30 19:47 UTC
[asterisk-users] Asterisk 11.10.2 update breaks snom TLS transport
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Am 30.06.2014 11:55, schrieb Heiko Wundram:> Is this a known regression, and generally: how can I help diagnose > this problem further for opening a bug-report, besides the > description given above? Has anyone else encountered this and found > another workaround besides switching to TCP/UDP-transport for the > phones?To answer my own question: applying the (latest) patch from https://issues.asterisk.org/jira/browse/ASTERISK-18345 fixes the issues with snom phones that I was seeing, but for me only starting with 11.10.2, earlier releases like those posting in that bug didn't have the problem I described. As I haven't looked through the additional implications of that patch applying to current Asterisk releases, is anybody willing to comment on whether that patch is safe to apply in production environments? - -- - --- Heiko. -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.22 (MingW32) iQEcBAEBAgAGBQJTsb7lAAoJEDMqpHf921/Sb1oH/j5Xbu6n2OC1WHy1ecVaT65S Q2AKraCMnAj8IOA0yu+3X41PN5mJnLvuGBFN3BZDFpnrgi3t09MCV3oaZbJ0xrb3 Mfy7EEcoQZMaJDEZK7VY1hR8HZOU22Jg+be0+XIParvUTTprNnyKDgQaum0QFR9x 2ZiKIJyD2g4YdqIu8p/vLK8Jxb22b3aG+Pzu3qfHxVBbVWKB1+yk16YAWOU9mGCW 7leIpi7tN4j3eEU9NoVg/K6cmrP/Lh0iaPesmXD1i710+czcfg4hz/mNOOlYH/5U iXnoKBGKgltZvsEEAfpOhiDX85moguGjN3zadK0gkSbZDaAv5gQQk1etHTd6VGc=HhP6 -----END PGP SIGNATURE-----
Apparently Analagous Threads
- Asterisk 1.8.15-cert7, 1.8.28.2, 11.6-cert4, 11.10.2, 12.3.2 Now Available (Security/Regression Release)
- Asterisk 1.8.15-cert7, 1.8.28.2, 11.6-cert4, 11.10.2, 12.3.2 Now Available (Security/Regression Release)
- Calls not hanging up
- Asterisk not honoring astetcdir
- Extra REGISTER messages sent by Asterisk when subscribe for MWI is defined in zoiper