Search results

Its a while since I messed with this, I'm not using it in production but it definitely worked. I had also abandoned the gateway method of ping, I think it was too unreliable and instead was using crons every minute, one for each ms server: /usr/bin/sipp -sf...

This help? <gateway name="msft-sip"> <param name="username" value="dummy"/> <param name="password" value="password"/> <param name="proxy" value="sip.pstnhub.microsoft.com:5061"/> <param name="from-domain" value="$${sbc_contact_name}"/> <param name="from-user"...

I cannot remember exactly, I'll check later but look at the difference between the kamailio and freeswitch when it comes to the contact and from fields.

The containers will work very well without timing issues. One of the companies I work for have freeswitch running in a couple of docker instances, they are processing a couple of thousand concurrent calls between them.

It can be done but to what benefit? Unless someone is going to constantly maintain it then it will be a real pain. Following the docker principle of on task per container there will be many containers, then for each commit in Fusionpbx it may be a different container to upgrade. An lxc...

You have the cdrs too. Also, I have seen mysql fail so many times for people though I do like the look of galera myself, I do have a three node cluster running.

I think it already has bits in that are postgres only, why do you want sqlite?

Ah yes, I see a CANCEL previous to that

Go into fs_cli and do a: reload mod_sofia That should show you the error within that, it will be pretty verbose.

Same here, not a clue what a sticky agent is.

Look again at the freeswitch logs and maybe do a: sofia global siptrace on The Freeswitch appears to be hanging up the other call, I can see a BYE from it.

I actually use a heavily modified version of pyfreebilling v2. Its replicated across two datacenters. I've recently discovered, and I have been working on for my own use, that I can build similar functionality using just opensips. I'm currently working on that.

What do you guys do to prevent fraud? This is always top of my list. For example, every call that I process, we have thing like max connection fee, max cost per minute, max connection fee international , max cost per minute international etc, etc These things get checked real time as the call...

We need better logs to be able to help with this

I guess the diversion header would have to be set by the carrier, you may never get to see this, you would need to check the incoming sip in sngrep.

It would appear you have some NAT/Firewall issue there. Either that or the source IP is banned in fail2ban. As you can see, the INVITE is being received but there is no reply being sent out. When you added the IP to the ACL, did you reload ACLs? Check the freeswitch log, or watch from the...

You could start by checking the postgres logs. If postgres is not restarting, again, the logs should tell you the problem.

In short, no, that is how freeswitch distinguishes between tenants.

Rather than try and do this in time conditions I would probably add a table to the db with dates for the deacons and just call a lua script called whosthedeacon.lua or something that just queries the db.

It will work, all that it is looking for is the nic drivers, they are broadcom I think, that image has all the firmware preinstalled.