Solved ISP Changed now 3CX Mobile App wont Connect

Status
Not open for further replies.

TaylorStevens23

SOHO User
Joined
Jul 27, 2020
Messages
51
Reaction score
9
This morning we experienced an Outage with our ISP. We switched over 3CX to our backup ISP provider and contacted our SIP provider and updated our new IP. The dashboard information is showing the new IP in green and set to Static in the management console. The issue we are currently having is the 3CX mobile app is not connecting and calls will route internally but cannot route externally. If someone from outside the organization calls they show as connected in the receptionist view but the phone is not ringing on the callers side and they are not hearing any of our messaging from the IVR. The PBX shows that external calls into our main line are hitting our IVR but there is no answer on the external callers side. If you try to place an outbound call you get a busy signal.

We have a pro license and I see online it can take up to 6 hours to propagate. We made this change around 2-3 hours ago and still are having the same issues. Is there anything I am missing or is there a way to speed up this process?

  • 3CX Version, Professional Annual 16.0.8.16
  • Server OS, Debian 10
  • Is the 3CX Server Hosted and where? On Prem
  • Provisioning Method: Local
  • Trunk Provider or Gateway Make/Mode: 123net
  • Are custom Phone Templates being used: YES
 
This morning we experienced an Outage with our ISP. We switched over 3CX to our backup ISP provider and contacted our SIP provider and updated our new IP. The dashboard information is showing the new IP in green and set to Static in the management console. The issue we are currently having is the 3CX mobile app is not connecting and calls will route internally but cannot route externally. If someone from outside the organization calls they show as connected in the receptionist view but the phone is not ringing on the callers side and they are not hearing any of our messaging from the IVR. The PBX shows that external calls into our main line are hitting our IVR but there is no answer on the external callers side. If you try to place an outbound call you get a busy signal.

We have a pro license and I see online it can take up to 6 hours to propagate. We made this change around 2-3 hours ago and still are having the same issues. Is there anything I am missing or is there a way to speed up this process?

  • 3CX Version, Professional Annual 16.0.8.16
  • Server OS, Debian 10
  • Is the 3CX Server Hosted and where? On Prem
  • Provisioning Method: Local
  • Trunk Provider or Gateway Make/Mode: 123net
  • Are custom Phone Templates being used: YES
UPDATE: We cycled the services and we can now get calls to connect if calling externally to direct numbers. I can now connect external calls to the PBX. I am experiencing an issue with outbound calls going directly to busy and hanging up.
 
Last edited:
Make sure your firewall has port forwarding for your backup ISP if you use multi-WAN.
 
Make sure your firewall has port forwarding for your backup ISP if you use multi-WAN.
I worked with my network guy to get the network fixed and I can now call outbound and accept calls inbound but my remote users are not getting audio on their desktop app when trying to accept calls. Does this desktop app, like the phone need the DNS to propagate? I can pass audio through if I use the app on the local network.
 
Does this desktop app, like the phone need the DNS to propagate
On each phone (as in, ext -> Phone Provisioning -> choose '3CX App') there is a choice of "Network interface for registration and provisioning" for either the hostname or IP. TTL for a 3CX hostname is 6 hours unless one has an Enterprise license in which case it is 5 minutes.

Edit: the hostname there is the server IP so is relevant if your server is in your office and therefore its IP changed.
 
  • Like
Reactions: Evolute IT
On each phone (as in, ext -> Phone Provisioning -> choose '3CX App') there is a choice of "Network interface for registration and provisioning" for either the hostname or IP. TTL for a 3CX hostname is 6 hours unless one has an Enterprise license in which case it is 5 minutes.

Edit: the hostname there is the server IP so is relevant if your server is in your office and therefore its IP changed.
We are getting extremely close or passing the 6 hour mark at this time. Can it take longer than that or is there a potential issue?
 
We are getting extremely close or passing the 6 hour mark at this time. Can it take longer than that or is there a potential issue?
It shouldn't take more than 6h.
 
We are getting extremely close or passing the 6 hour mark at this time. Can it take longer than that or is there a potential issue?
You can check..."ping hostname" should use a cached value, but "nslookup hostname 8.8.8.8" or other DNS server should query direct, though keep in mind all DNS servers should cache it so you could try a few, or look up the NS records for the 3CX domain and query them directly.

TTL depends on when it was used at the client side. If it was cached 5:59 ago then it should pick up the change in 1 minute. If it was cached 0:01 ago then it will take 5:59 to cache it again.
 
  • Like
Reactions: Evolute IT
You can check..."ping hostname" should use a cached value, but "nslookup hostname 8.8.8.8" or other DNS server should query direct, though keep in mind all DNS servers should cache it so you could try a few, or look up the NS records for the 3CX domain and query them directly.

TTL depends on when it was used at the client side. If it was cached 5:59 ago then it should pick up the change in 1 minute. If it was cached 0:01 ago then it will take 5:59 to cache it again.
Can you give me some insight on why it takes so long
 
Nearly 24 hours and still has not updated
 
  • Like
Reactions: AWS2P
Go to "dnschecker.org", enter your FQDN and check if the IP result is the same IP as your backup ISP.
I have done this and it shows the old IP still
 
I have done this and it shows the old IP still
OK, go to the 3CX, into Settings > Network, set the IP as dynamic. Reboot the server or restart the services, wait a few minutes and check the IP in your customer portal as well.

Are you using an Std or Pro license?
 
OK, go to the 3CX, into Settings > Network, set the IP as dynamic. Reboot the server or restart the services, wait a few minutes and check the IP in your customer portal as well.

Are you using an Std or Pro license?
I have seen this suggestion before but I am not sure what doing this does. Can you explain why changing the IP to dynamic and rebooting will help anything?

We have a pro license. I am in the customer portal and I do not see a spot with an IP listed. I do see our FQDN listed under our subscription but that is still resolving to our old IP address.
 
I have seen this suggestion before but I am not sure what doing this does. Can you explain why changing the IP to dynamic and rebooting will help anything?

We have a pro license. I am in the customer portal and I do not see a spot with an IP listed. I do see our FQDN listed under our subscription but that is still resolving to our old IP address.
Doing so kinda "forces" a FQDN update. You can set back to static afterwards.

Pro licenses have a 6 hour TTL, this is a limit 3CX set. Enterprise licenses get 5 min TTL. If you use your own FQDN you can control the time.
 
  • Like
Reactions: Evolute IT
I have seen this suggestion before but I am not sure what doing this does. Can you explain why changing the IP to dynamic and rebooting will help anything?

We have a pro license. I am in the customer portal and I do not see a spot with an IP listed. I do see our FQDN listed under our subscription but that is still resolving to our old IP address.
When setting the IP to dynamic, it forces the system to recheck all of its network public IPs and this forces the 3CX DNS to update again.

The IP would be in Instance Details I believe. I still use the old portal so we see it in there.

I would recommend upgrading to Enterprise next renewal so that you enjoy 5min TTL instead. We never sell anything other than that for that unique reason (which allows us to quickly move an instance if needed, with minimal downtime.)
 
I have done this and it shows the old IP still
If no one's DNS has the updated IP then your server hasn't told 3CX to update the IP yet. See the post I referenced, above.

re: why TTL takes so long... TTL is Time To Live. That number tells the world how long to cache a DNS answer. If a DNS server has an answer in its cache it doesn't look it up again. It sounds like that wasn't your issue though.
 
  • Like
Reactions: Evolute IT
When setting the IP to dynamic, it forces the system to recheck all of its network public IPs and this forces the 3CX DNS to update again.

The IP would be in Instance Details I believe. I still use the old portal so we see it in there.

I would recommend upgrading to Enterprise next renewal so that you enjoy 5min TTL instead. We never sell anything other than that for that unique reason (which allows us to quickly move an instance if needed, with minimal downtime.)
I believe we identified the issue, there was an oversight from our network setup that fixed most of our issues once resolved. This was also causing the new IP not to propagate, after we fixed it we briefly moved the the server to dynamic and back to static. We did not reboot the server or services due to users on the phones. Does this step need to be done?
 
  • Like
Reactions: Evolute IT
Status
Not open for further replies.

Forum statistics

Threads
112,147
Messages
590,959
Members
165,167
Latest member
Finatra.us