I seem to have 2 'lazy' extensions

[toolfolks]

I have a Grandstream GXP1628 and HT108 ATA.

I have our mobile software running on IOS & Android and also GSWave on IOS & Android.

All can call each other put on hold etc etc.
I have a DialPlan that checks for the caller ID and if it matches transfers the call direct to ext 1053.

However after a while the 2 hardware extensions 'go to sleep'. When I call on an inbound number there's a silence for a while then the call goes to voice mail. Same if I call from a sip phone.

To wake them up I have to dial another extension and all is ok again

 

[DigitalDaz]

What are your registration timers set at, that sounds like classic NAT tunnel timeout

 

[toolfolks]

Cheers,

as usual as a newbie I go to check the settings and there is a few different options.

Which one is the registration timers

 

[ad5ou]

Registration expiration is way too high (60 minutes) change setting to 2
Also enable “options keep-alives ”

 

[toolfolks]

Cheers.
I’ve changed them.
test in the morning

 

[Adrian Fretwell]

Registration expiration is way too high (60 minutes) change setting to 2

You should never need to set registration expiration to such a low value as long as your keep alive pings are working correctly.

If NAT holes are getting closed, it is better to properly investigate the cause rather than make your system work around it.

Short registration expiry times puts extra load on your system. If you have a relatively small number of extensions it's not a problem but it certainly will not help you scale well.

 

[toolfolks]

What are the ‘ optimum ‘ settings.
It’s a minefield as there are 6 settings on the 6128 so how does one test the various combinations.

Any documentation out there.

 

[DigitalDaz]

If you are also not that savvy with troubleshooting, setting a low reg timer is the easy solution.

Some nat keep alive mechanisms are not that great too you want at least options of info. I usually use options.

Also, low reg timers are necessary for many of us doing failover anyway so we are used to it.

 

[Lorena Watkins]

Hi guys, I know this is an old post, but I am dealing with the exact problem that @toolfolks described on the first message: "However after a while the 2 hardware extensions 'go to sleep'. When I call on an inbound number there's a silence for a while then the call goes to voice mail. Same if I call from a sip phone."

In our case it happened just after a server reboot. We find out that when rebooting the phone it will come back alive. So my question is: it is a normal procedure that all phones should be rebooted after a server reboot? Since we have a many phones in different locations it makes it difficult to do it so, is there a way to bring alive this phones without having to manually reboot them?

Thanks in advance

 

[Adrian Fretwell]

Phones should not need to be rebooted if you reboot your server. FreeSWITCH maintains it list of registrations in its SQLite database, it will pick the registration state up again after a restart, unless you have moved SQLite on to an in memory filing system in which case a restart of FreeSWITCH will be fine but a server reboot will lose registration information. You can even work around this by copying the SQLlite DB after FreeSWITCH is shut down but few people bother about doing it.

 

[Lorena Watkins]

Thanks a lot for our answer @Adrian Fretwell I will try this.

I guess maybe it was originated due to a difference between the time on my freeswitch vs the time on the server?

Since I run a fs_cli -x 'fsctl sync_clock' and it seems to be fine now.

 

[pnwitguy]

Thanks a lot for our answer @Adrian Fretwell I will try this.

I guess maybe it was originated due to a difference between the time on my freeswitch vs the time on the server?

Since I run a fs_cli -x 'fsctl sync_clock' and it seems to be fine now.

What other things did you try? I tried time sync command and my extension is still quite and eventually goes to voicemail.