VoIP Mailing List Archives

[jolexpert at gmail.com]

hello,

Did you try to upgrade to latest release 1.10.6 (march 2021) and check if this "blackout" is solved or still present ?.


regards

[remi.marand at odigo.com]

Hello,

It seems we have the same problem, it doesn't happen often, but the Register part of internal profile does crash.

I observe that the number of connections established on the WSS port (7443) sometimes exceeds the number of active registrations. (netstat -tpa)..
It could be really important (2 or 3 times the number of Registrations), it could of course be a software problem on the WebRTC (Front) client side.

Is there a way on the Freeswitch side in the configuration to disconnect TLS connections that are no longer registered? On timeout for example.

Best regards.

Rémi Marand.
remi.marand@odigo.com

De : FreeSWITCH-users <freeswitch-users-bounces@lists.freeswitch.org> De la part de Kakiman Expert
Envoyé : vendredi 30 avril 2021 15:34
À : freeswitch-users@lists.freeswitch.org
Objet : [Freeswitch-users] RE - Internal Interface suddenly freezes


hello


Did you find any solution to this issue ?



I am in production with 1.10.5 version and since a week, I have the same issue.

The internal profile is out of service and registration is impossible.



and we don't have any answer to "sofia status profile internal" , the prompt in fs_cli keeps waiting.



the only way to solve is to restart freeswitch



thanks

regards

[s.kainz at wnt.at]

Another one
At this point we have a script that restarts freeswitch every time this happens.
We don’t really have a solution for this so far, and are out of things to test …

May I ask if you have an external interface in your setup? Does that interface freeze too?
We have a external interface, and it looks like it doesn’t happen to that one.
We also have a third interface called “alternate_routing�, which doesn’t handle registrations or presence. Like the external interface it doesn’t freeze.

Do you use Mariadb/ODBC for odbc-core?

Im hoping the GitHub-issue gets some attention soon.
Maybe it helps if some of you also post a backtrace when this happens.

To your problem, can you confirm that this isn’t some sort of SYN-Flood-Attack?
Maybe check the IP’s that are connected. I think its pretty easy to find out if those are real-users or hackers.

Regards,

[remi.marand at odigo.com]

Hello,

Thank you for the various information about this incident on the userlist and github.

We are using Postgresql on Freeswitch server itself.
The version of Freeswitch is 1.10.5-release-17-25569c1631~64bit On Debian 10.

I cannot reproduce the problem on our development environment with thousands of registrations and thousands of calls, for this reason I suspect some unexpected behavior of the web client, and I note that a few TCP connections: on wss:7443 remain after the Register they carry is completed.
I am looking for a solution on the Freeswitch side to close them on inactivity timeout.

If we are successful in causing the problem, we will release the core as you recommend.
As for the transition to version 1.10.6, this is a track that we will try.

Best Regards.

remi.marand@odigo.com

De : FreeSWITCH-users <freeswitch-users-bounces@lists.freeswitch.org> De la part de Stefan Kainz
Envoyé : lundi 24 mai 2021 17:42
À : FreeSWITCH Users Help
Objet : Re: [Freeswitch-users] RE - Internal Interface suddenly freezes



Another one


At this point we have a script that restarts freeswitch every time this happens.

We don’t really have a solution for this so far, and are out of things to test …



May I ask if you have an external interface in your setup? Does that interface freeze too?

We have a external interface, and it looks like it doesn’t happen to that one.

We also have a third interface called “alternate_routing”, which doesn’t handle registrations or presence. Like the external interface it doesn’t freeze.



Do you use Mariadb/ODBC for odbc-core?



Im hoping the GitHub-issue gets some attention soon.

Maybe it helps if some of you also post a backtrace when this happens.



To your problem, can you confirm that this isn’t some sort of SYN-Flood-Attack?

Maybe check the IP’s that are connected. I think its pretty easy to find out if those are real-users or hackers.



Regards,

[s.kainz at wnt.at]

Thank you.

We will try 1.10.6 with mod_mariadb this week!

Thanks!

Von: FreeSWITCH-users <freeswitch-users-bounces@lists.freeswitch.org> Im Auftrag von Rémi MARAND
Gesendet: Montag, 24. Mai 2021 19:24
An: FreeSWITCH Users Help <freeswitch-users@lists.freeswitch.org>
Betreff: Re: [Freeswitch-users] RE - Internal Interface suddenly freezes



Hello,

Thank you for the various information about this incident on the userlist and github.

We are using Postgresql on Freeswitch server itself.
The version of Freeswitch is 1.10.5-release-17-25569c1631~64bit On Debian 10.

I cannot reproduce the problem on our development environment with thousands of registrations and thousands of calls, for this reason I suspect some unexpected behavior of the web client, and I note that a few TCP connections: on wss:7443 remain after the Register they carry is completed.
I am looking for a solution on the Freeswitch side to close them on inactivity timeout.

If we are successful in causing the problem, we will release the core as you recommend.
As for the transition to version 1.10.6, this is a track that we will try.

Best Regards.

remi.marand@odigo.com (remi.marand@odigo.com)

De : FreeSWITCH-users <freeswitch-users-bounces@lists.freeswitch.org (freeswitch-users-bounces@lists.freeswitch.org)> De la part de Stefan Kainz
Envoyé : lundi 24 mai 2021 17:42
À : FreeSWITCH Users Help
Objet : Re: [Freeswitch-users] RE - Internal Interface suddenly freezes



Another one


At this point we have a script that restarts freeswitch every time this happens.

We don’t really have a solution for this so far, and are out of things to test …



May I ask if you have an external interface in your setup? Does that interface freeze too?

We have a external interface, and it looks like it doesn’t happen to that one.

We also have a third interface called “alternate_routing”, which doesn’t handle registrations or presence. Like the external interface it doesn’t freeze.



Do you use Mariadb/ODBC for odbc-core?



Im hoping the GitHub-issue gets some attention soon.

Maybe it helps if some of you also post a backtrace when this happens.



To your problem, can you confirm that this isn’t some sort of SYN-Flood-Attack?

Maybe check the IP’s that are connected. I think its pretty easy to find out if those are real-users or hackers.



Regards,