v20 On Prem Upgrade - NIC Keeps changing

[micro13]

I have spent countless hours trying to Upgrade from version 18 to version 20. This is an on premise vm. instance, the upgrade appears to go fine and the process completes. I can make calls from my soft phone mobile app. but all the other on Prem phones cannot provision.

The IP address of the server has changed from .14 to .150 .

• I've changed the necessary firewall ports and redirects.
• I have performed the split DNS properly.
• Thank God for snapshots so I can revert back to version 18 and try again - after each attempt. Literally tried several over 6 different reinstall attempts with varying steps
• I even went into each phone. logged in and changed the IP manually. The system then seemed then seemed to work fine. All the phones were listed although it was a tedious process. But the moment I reset the passwords for all users so that they can log in to the new system, all the phones lost their provisioning to v20 and I couldn't communicate with them.

I have performed SEVERAL upgrades ondifferent platforms - some cloud based some On Prem with no problems but this install has cost me way too many hours. Anyway, can anyone offer any suggestions?

This most definitely appears to be a phone provisioning issue between the upgrades. i know v20 provisions a bit differently. THese are all Yealink phones, which I never have problems with.

HELP! Thanks

 

[VoIPTools]

I assume you have a static internal IP address and not using DHCP? Or perhaps a reservation in DHCP (not recommended)? It sounds like you have a handle on the infrastructure, but thought I would ask nevertheless.

 

[micro13]

I did have a reservation - but removed it. I also went into the interface and assigned a static ip there - the same one as v18 (.14). I then went into each phone and ...

1. Deleted the old Phone, Added the new phone back in V20 w/ same MAC
2. Logged in directly to each (using OLD phone passwords from v18), - Manually changed the provisioning link
3. removed IP from blacklist - IT KEPT blacklisting them because v20 assigned new passwords

System is up and running BUT...

When I sent the RESET for a single phone (to test) to get user the new credentials, it lost the provisioning again! I am sure the password is not making it's way over to the phone (Yealink). Same problem as before

What am i missing???

 

[Alphabetic]

Have you checked your DHCP server to make sure the scope isnt giving out the IP of the SIP server? (Option 66)

 

[NicholasP_3CX]

When you are creating a VM you should always have a static MAC address AND a static IP Address.

Between Version 18 and Version 20, did you change your HTTPS or HTTP port the phones are provisioning to?

You didnt mention if it is a Windows PBX or Linux.

Because the Windows edition requires uninstalling and resintalling the PBX, it will ask you for the HTTP/S ports of the system. If you just clicked on next in the wizard it would revert to 80 and 443 instead of 5000 and 5001, and here starts your provisioning issues. The port numbers are different.

 

[micro13]

Have you checked your DHCP server to make sure the scope isnt giving out the IP of the SIP server? (Option 66)

ASsigned ip is outside the scope of DHCP and Option 66 is enabled properly

 

[micro13]

When you are creating a VM you should always have a static MAC address AND a static IP Address.

Between Version 18 and Version 20, did you change your HTTPS or HTTP port the phones are provisioning to?

You didnt mention if it is a Windows PBX or Linux.

Because the Windows edition requires uninstalling and resintalling the PBX, it will ask you for the HTTP/S ports of the system. If you just clicked on next in the wizard it would revert to 80 and 443 instead of 5000 and 5001, and here starts your provisioning issues. The port numbers are different.

This is a Linux VM. I UPGRADED from v18 to v20 in place. Initially it wreaked havoc when it assigned a new IP and all phones lost their provisioning. I manually changed the IP assignment to new IP in 3cx but it was still crazy.
Fortuantely I could roll back to the original version with the snapshot capabilities of VMs. I attempted this upgrade over 6 or 7 times with different variations - so many hours!!

After that, I went ahead and manually changed the IP in linux using sudo nano command (interface). This gave me the most consistent results. I manually logged into EACH IP PHONE and changed the provisioning link to the new one and the phone would provision properly.

Problem is when I roll out new Welcome email (RESET) with instructions for users - the phone loses it s provisoning AGAIN.

 

[Alphabetic]

What are the commands youre running and have you confirmed its taking the new IP by pinging the FQDN?

 

[micro13]

What are the commands youre running and have you confirmed its taking the new IP by pinging the FQDN?

Tested Split DNS - FQDN successfully resolves to local IP, I can ping the 3cx internal server. To reset user pw. i click on the reset button in the web console while selecting checkboxes for spefic users that i wish to reset passwords and send email to