I mixed up a few things and didn't explain the whole story... I wrote my own STP (without any digium code), and I'm still using Asterisk 1.8.12 + my own patches for ISUP/SIP conversion, but I'm planning to completely replace asterisk with a KISS solution that only does basic ISUP<->SIP conversion and TDM<->alaw RTP voice conversion, and let a pure SIP switch do all the complex stuff (but that's still probably 6 months away). The STP is working very well. Actually that route is a single remote STP route, which actually goes completely down (and then we have the problems with CICs described). Its not a situation where we have two links and one out of two is always up. The old 1.0 only looked for ALARM to detect a link down, it should instead only care about presence of FISU / MSUs keeping the link alive, but the dahdi MTP2 mode hides us from that heartbeat. So I wrote my own kernel DAHDI patch that keeps FISU repeating on the TX side, and only reports different FISU/MSU on the RX side, but detects a silent RX link and converts that case into a LSSU_SIOS to the user layer. It also does the reverse, if the userland is silent or closes the channel, it sends LSSU_SIOS to the remote side. This also helps the fact that my asterisks mostly don't use TDM SS7 links anymore, but instead SS7 TDM like protocol on UDP, keeps the code 99.9% the same (only socket setup/tear down different). The STP does the actual TDM links (in most cases). On Wed, Apr 20, 2016 at 12:49 PM, Goke Aruna <goksie at gmail.com> wrote: > Thanks Kolayan, > All seems okay with few warning mesages on unequipped E1. > However, I will update the link as the test goes on. > Regards > > On Wed, Apr 20, 2016 at 2:50 PM, Michal Ryb?rik <michal at rybarik.sk> wrote: > >> Hi, >> >> I'm running libss7 with SS7-27 patches on Asterisk 11 for few years, and >> it works perfectly. Millions of calls connected without any issue, restart, >> unstability, etc. During interoperability testing we had some problems with >> deadlocks after BLOcking CICs with older SS7-27 patches, but I debugged it >> and Kaloyan fixed it then, fix is included in official sources now. All >> this should be published as libss7 2.0. Kaloyan did a lot of work on this >> and I really don't see _any_ issues now. >> >> -- >> Michal Rybarik >> >> >> On 04/19/2016 07:27 PM, Marcelo Pacheco wrote: >> >>> I have my own extensive unpublished patch for Asterisk 1.8 that fixes >>> lots of SS7/ISUP issues for me. >>> I tried Asterisk 11 with unpatched sources and found lots of issues >>> again. I couldn't even get a stable environment with just all trunks >>> aligned. >>> So I'm not surprised you're having problems. >>> No, my patch isn't available, not it will for free. Most problems have >>> been reported to this list before, and haven't been addressed in the 3 >>> years passed since I reported them. >>> So I don't advise anybody to use either ss7 solutions, you will have >>> serious problems. The libss7 development process is unable to deliver a >>> serious ss7 solution that will work properly. >>> >>> >> >> -- >> _____________________________________________________________________ >> -- Bandwidth and Colocation Provided by http://www.api-digital.com -- >> >> asterisk-ss7 mailing list >> To UNSUBSCRIBE or update options visit: >> http://lists.digium.com/mailman/listinfo/asterisk-ss7 >> > > > -- > _____________________________________________________________________ > -- Bandwidth and Colocation Provided by http://www.api-digital.com -- > > asterisk-ss7 mailing list > To UNSUBSCRIBE or update options visit: > http://lists.digium.com/mailman/listinfo/asterisk-ss7 > -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.digium.com/pipermail/asterisk-ss7/attachments/20160420/74de9045/attachment-0001.html>
|