ALERT Buh-bye CentOS.

[Eliad]

After re-doing the install over SSH the web interface is working. What is the best way to migrate from Incredible PBX 2020 to Incredible PBX 2021?

that is the problem. You have to run incrediblepbx install script through SSH. During this install the local network and the computer IP you installed from is whitelisted. I think the easiest way is to start from scratch instead of trying to fix the firewall settings. @ward said many, many times, do not mess up with the firewall.

 

[Eliad]

After re-doing the install over SSH the web interface is working. What is the best way to migrate from Incredible PBX 2020 to Incredible PBX 2021?

In the Incredible 2020 you have incrediblebackup16, do not use that, it will hose your system during the backup. You have to download incrediblebackup2021 and use that one instead. Make a snapshot of the incredible2020 system before you do anything from now on. run fwconsole ma upgradeall then ./sig-fix 2 times, then fwconsole reload.
Then run incrediblebackup2021 and it will create a file in /backup . copy this file (starts with CN) to the new incredible2021 (make sure is all updated before that). Then run incrediblerestore2021, it will complain is a different platform ( i think is referring to Centos versus Debian). Then rerun the fwconsole ma upgradeall then ./sig-fix 2 times, then fwconsole reload. I you should be ok. reboot the new 2021 several times to make sure is stable.
If UCP does not load see the other thread on that

 

[smccloud]

Fixed the UCP issue. Still showing all phones registered, which is weird since I haven't moved them to it yet.

 

[herbertshades]

Installed this 2021. This feels more polished that the Centos based. All was smooth sailing until I went to configure MoH. I can not configure Type:Custom Application. Error: SQLSTATE[22007]: Invalid datetime format: 1366 incorrect integer value: "for column `asterisk`.`music`.`random` at row 1 File:/var/www/html/admin/modules/music/Music.class.php:136.

Any solutions?

 

[n9yty]

Testing this out on an AzureVM instance as I have easy access to that at the moment without setting up billing anywhere else.

Here are some things I saw during install... And at the end it was a failed attempt.

There was a warning that the version of NodeJS being used is no longer supported... Is/was that fixed during the install with an update, or what? Seems running currently maintained versions would provide more security:

Node.js 11.x is no longer actively supported!

I don't know the reasons behind the choices or any incompatibilities later versions may have introduced, so I'll leave this as "just mentioning it" and not as a problem.

I also saw this, but couldn't find the related error:

. . . home/asterisk/.node/lib/node_modules/npm-cache/.jshintignore home/asterisk/.node/lib/node_modules/npm-cache/README.md home/asterisk/.node/lib/node_modules/npm-cache/.jshintrc home/asterisk/.node/lib/node_modules/npm-cache/index.js home/asterisk/.node/lib/node_modules/npm-cache/CONTRIBUTING.md home/asterisk/.node/lib/node_modules/npm-cache/util/ home/asterisk/.node/lib/node_modules/npm-cache/util/logger.js home/asterisk/.node/lib/node_modules/npm-cache/util/parseUtils.js tar: Exiting with failure status due to previous errors

and


Setting base permissions...chmod: changing permissions of '/var/www/html/index.php': Operation not permitted chown: changing ownership of '/var/www/html/index.php': Operation not permitted

and a bit further on:

Running Asterisk post from Clearlytrunking module No valid token found for Clearly Trunking Running Asterisk post from Core module Running Asterisk post from Dahdiconfig module Running Asterisk post from Smsclient module Running Asterisk post from Ucp module Starting UCP Node Server... [>---------------------------] < 1 sec Started UCP Node Server. PID is 8697 Reload Started [FATAL] Running an unsupported version of Asterisk. Supported Asterisk versions: 13, 14, 15, 16, 17. Detected Asterisk version: 18.1.1 Taking too long? Customize the chown command, See http://wiki.freepbx.org/display/FOP/FreePBX+Chown+Conf Setting Permissions... Setting base permissions...chmod: changing permissions of '/var/www/html/index.php': Operation not permitted chown: changing ownership of '/var/www/html/index.php': Operation not permitted Done

That showed up a bit later too...

Starting UCP Node Server... [>---------------------------] < 1 sec Started UCP Node Server. PID is 11215 Reload Started [FATAL] Running an unsupported version of Asterisk. Supported Asterisk versions: 13, 14, 15, 16, 17. Detected Asterisk version: 18.1.1

Random line:
./IncrediblePBX2021.sh: line 540: debconf-set-selection: command not found

And then when it finished it cleared the terminal so I couldn't look for more. But after it rebooted, I can no longer get into the system. I rebooted it from the Azure console, still cannot get in. Guess I have to wipe and start over. Not really familiar with Azure, looks like I have to destroy the VM and reload, so working on it. Also, I was logged in as the user created during the install, and used `sudo bash` before installing things, but this time I am going to change things so I can log in as root in case it matters.

 

[wardmundy]

Looks like Asterisk 18.1.1 threw another monkey wrench into FreePBX. Will back off to 18.1. Try a new install in a few minutes. Other issues are harmless and either required or intended at this juncture.

 

[wardmundy]

@n9yty: Verified a successful new install on Vultr moments ago.

 

[smccloud]

I have noticed that after doing a restore from IncrediblePBX 2020, it no longer says 2021 in the UI. Is there an easy fix for this?

 

[wardmundy]

I have noticed that after doing a restore from IncrediblePBX 2020, it no longer says 2021 in the UI. Is there an easy fix for this?

Edit /etc/pbx/.version

 

[kenn10]

Looks like Asterisk 18.1.1 threw another monkey wrench into FreePBX. Will back off to 18.1. Try a new install in a few minutes. Other issues are harmless and either required or intended at this juncture.

Remember, in FreePBX Settings > Advanced Settings in the View attachment 3187

When I loaded 18.1.1, I didn't have to add this so I don't know what changed. It's possible the Framework and Core got modified and that was left out again. There were so many different parts of FreePBX that check for version including the reload module.

@wardmundy I wonder if you can write the version 16.0.1 to the DB before trying fwconsole commands.

 

[jerrm]

What version of the Framework module is this? The current install tarball and module should support Asterisk v18, Does Ward have the "before ransomware" version lock in place for Debian?

 

[wardmundy]

No. It turns out this 18.1.1 issue was a red herring. There are various install issues along the way to the final product. But the final module upgrade of all modules at the end of the script solves the version issues so we have restored the previous script.

 

[wardmundy]

Note how well CloudAtCost keeps time, not to mention allocating a 5GB swap file.

 

[n9yty]

I did eventually get a working install, learning Azure and this at the same time, competing firewalls, etc.. I did find out how to attach a web serial console to the VM so I was able to get in and sort things out after a new install. I am having some other issues that I will make a separate post on, but this one is done. I don't know if it was because I was not actually root that caused the problem or not, but since I couldn't get logged in before I just wiped it all out we will never know.

 

[wardmundy]

Hint: RTFM

And, yes, installing without root credentials would be a major problem.

 

[wardmundy]

CloudAtCost also revealed another missing piece. Will push out time synchronization shortly.

apt-get install ntp -y
systemctl enable ntp

 

[kenn10]

CloudAtCost also revealed another missing piece. Will push out time synchronization shortly.

I didn't have a problem on my servers but neither Free Range Cloud nor Hotlineservers had NTP installed either. Your update installed easily on them.

 

[n9yty]

Thanks, @wardmundy for your tireless help and dedication. I launched it with sudo, but not actually "being" root can cause strange issues. At any rate, it is resolved now, I just had to tweak the default Azure/Debian config to get to a proper starting point.

 

[Starburst]

Sad.
Well looks like back to 3CX.
At least won't have to worry about getting a DID Provider configured or a phone.
It's already built-in.

FreePBX could do the same thing to IncrediblePBX with Non-GPL code as well.

Over the years, as I started with CP/M as well.
RedHat did this back around 2002 (I think) when they went Commercial, and announced the GPL version would now be Fedora.
Debian/Ubuntu didn't even exist back then.

Only disto that hasn't come and gone is FreeBSD.

 

[Eliad]

Thanks, @wardmundy for your tireless help and dedication. I launched it with sudo, but not actually "being" root can cause strange issues. At any rate, it is resolved now, I just had to tweak the default Azure/Debian config to get to a proper starting point.

It happened to me on AWS, just running the install script under sudo did not work well. I had to be root with su before running incrediblepbx script.