V16: Linux OS Upgrade from Debian 9 to Debian 10 is now Available

[MH_INDUS]

I upgraded our test instance to v18 and noticed some great improvements but the new webclient will cause a lot of problems with our users if it is implemented in its current state.

• Use this device for calls is too hidden, it should be at most 2 clicks away
• Queue log in /out is harder to access
• Can`t edit status in the menu where you can set it
• The icons on the left side are not enough for our users to fin the correct menu
• Call history is too small to easily read, v16 was better in this area

Positive:

• The settings menu for the admin is much cleaner
• Audit log is a great addition
• Codec settings are improved

As stated several times, SSO via SAML or LDAP as well as improvements to the phonebook are much needed.

 

[pmterp]

Is there a way to see if the install was from the 3CX ISO or if 3CX was added? Maybe a specific package we can look for?
Back at v15.5 we were installing Debian 8 and then adding 3CX. All of those have since been updated to v16 and Debian 9. At v16 we started using the 3CX ISO for new installs.
When we start the process moving everyone to v18 of course we'll make backups first but I don't think we have any record of which servers started as stock Debian as compared to using the 3CX ISO. It would be nice if we could identify them just so we know where issues may pop up.

 

[Dan JEM]

Is there a way to see if the install was from the 3CX ISO or if 3CX was added? Maybe a specific package we can look for?
Back at v15.5 we were installing Debian 8 and then adding 3CX. All of those have since been updated to v16 and Debian 9. At v16 we started using the 3CX ISO for new installs.
When we start the process moving everyone to v18 of course we'll make backups first but I don't think we have any record of which servers started as stock Debian as compared to using the 3CX ISO. It would be nice if we could identify them just so we know where issues may pop up.

So the original method was a update.sh script that was downloaded and ran. I found it on both test linux systems I have and one upgraded to Debian 10 just fine and the other failed. It didn't upgrade the OS and stripped off 3CX. I had done a VM snapshot so easy rollback. For bare metal systems that'll leave needing to take a backup and pull the backup file off the system.

Check for the update.sh to see if you left it on one of the user accounts home directories or on the file system somewhere. Also there should be an UPDATE.log folder in the /root folder if you followed 3CX's original instructions.

This was the original article which is depreciated: https://www.3cx.com/blog/docs/upgrading-debian/

At some point they also added in the ability to update it via the management console: https://www.3cx.com/blog/docs/upgrade-debian-9/

 

[ChrisKoll]

Version 18 Alpha (Includes OS upgrade from Debian 9 to 10), May break your system
I know it is in Alpha, however I needed to know what would happen if a client did the upgrade, I installed new Debian 9 from 3cx ISO, then did the upgrade from the management console, the system is now unusable. I recommend not using this if you are not willing to reinstall if needed.

 

[Harjeet]

Hi Guys
i am new to 3CX, we setup a new 3CX instance using AWS and it was working fine v16, then from web dashboard we clicked upgrade to v18, and now looks like system is broke. we can't login to web dashboard it says "internal server error" reboot of instance didnt help.

on instance manager in our 3CX portal says 11 out of 12 services failed to start and on more details says services doesn't exist.

not sure what is best way forward to fix it. it was clean install.

 

[weimerd]

Version 18 Alpha (Includes OS upgrade from Debian 9 to 10), May break your system
I know it is in Alpha, however I needed to know what would happen if a client did the upgrade, I installed new Debian 9 from 3cx ISO, then did the upgrade from the management console, the system is now unusable. I recommend not using this if you are not willing to reinstall if needed.

You may try to access to your installation using port 5001. So just add :5001 to the acces domain (example: yourInstallName.my3cx.de:5001)

I realized that in my test azure install I had similar behaviour where the installation won't respond to 443 port anymore. Did not investigate further here as I had a focus on function and design. But maybe will check later on.

 

[StefanW]

Hi Guys
i am new to 3CX, we setup a new 3CX instance using AWS and it was working fine v16, then from web dashboard we clicked upgrade to v18, and now looks like system is broke. we can't login to web dashboard it says "internal server error" reboot of instance didnt help.

on instance manager in our 3CX portal says 11 out of 12 services failed to start and on more details says services doesn't exist.

not sure what is best way forward to fix it. it was clean install.

You should get an email once the upgrade is completed or failed, did you got such an email?
In case you rebooted the instance during the process I cannot tell you in which state the system is now and would advise to take the pre-upgrade back up and go again with a new instance (pref. already on Deb10)

 

[Harjeet]

You should get an email once the upgrade is completed or failed, did you got such an email?
In case you rebooted the instance during the process I cannot tell you in which state the system is now and would advise to take the pre-upgrade back up and go again with a new instance (pref. already on Deb10)

Hi Stefan
the upgrade was started yesterday late night so i left it for the night, no emails received. when i checked in the morning i can see the admin login page but can't login as it says error. so looks like its completely broke.

i believe deleting AWS instance and then just create new one following pbxexpress wizard is what i need to do.

 

[ChrisC_3CX]

@GlobalSiteGuard

Version 18 Alpha (Includes OS upgrade from Debian 9 to 10), May break your system
I know it is in Alpha, however I needed to know what would happen if a client did the upgrade, I installed new Debian 9 from 3cx ISO, then did the upgrade from the management console, the system is now unusable. I recommend not using this if you are not willing to reinstall if needed.

Version 18 is indeed in Alpha and not recommended for instances running in a production environment, however, the upgrade process has been thoroughly tested and should not cause any issues on systems compliant with our hardware requirements and that have been deployed as per our guides using the 3CX ISO.

If the instance you are referring to is still running or you can replicate the issue with a new instance, please PM me so that we can look into this.

 

[Harjeet]

@GlobalSiteGuard

Version 18 is indeed in Alpha and not recommended for instances running in a production environment, however, the upgrade process has been thoroughly tested and should not cause any issues on systems compliant with our hardware requirements and that have been deployed as per our guides using the 3CX ISO.

If the instance you are referring to is still running or you can replicate the issue with a new instance, please PM me so that we can look into this.

Hi Chris
instance was created by 3CX install wizard on our AWS, so it should be supported. but anyway i am creting new one now and will try again.

 

[weimerd]

Hi Chris
instance was created by 3CX install wizard on our AWS, so it should be supported. but anyway i am creting new one now and will try again.

try using the iso instead of pbxexpress install.

 

[Harjeet]

try using the iso instead of pbxexpress install.

on AWS instance not sure if I can use ISO ?, or shall we do script based install on debian os

 

[ChrisC_3CX]

@Harjeet

Not sure if you can upload a custom ISO to AWS per se, but maybe you could deploy 3CX via the AWS Marketplace instead if you prefer.

Here's the guide: https://www.3cx.com/docs/pbx-amazon-aws-marketplace/

 

[njmsa]

Is it possible to upgrade to V18 Alpha on a Raspberry Pi4? It's not giving me the update option so far?

 

[ChrisC_3CX]

Is it possible to upgrade to V18 Alpha on a Raspberry Pi4? It's not giving me the update option so far?

This is not possible yet, you will most probably be able to upgrade to V18 on a Pi 4 once it is officially released.

 

[Ruymar Castro]

Do we already have any scripts defined to update systems that don't use ISO 3CX?

I have tested the current script on systems installed on Debian 9, without using ISO 3CX, I have had success on some but had problems upgrading one.

 

[ChrisC_3CX]

@Ruymar Castro

Do we already have any scripts defined to update systems that don't use ISO 3CX?

Not really, the upgrade is meant for 3CX Instances that have been deployed as per our guides and are therefore supported by 3CX. An instance not deployed using the 3CX ISO does not fall under this category so we cannot guarantee a smooth upgrade process or that it will successfully upgrade at all for that matter.

I have tested the current script on systems installed on Debian 9, without using ISO 3CX, I have had success on some but had problems upgrading one.

The above of course is not to say that the upgrade will definitely fail but we recommend avoiding it if possible by redeploying the instance before upgrading, in a 3CX Supported manner, in this case, using the 3CX ISO.


Note: Instances deployed via other 3CX supported methods such as via PBX Express or Hosting Provider's Marketplaces should of course be able to upgrade without issues as well.

 

[njmsa]

This is not possible yet, you will most probably be able to upgrade to V18 on a Pi 4 once it is officially released.

Aha, thank you for clearing that up! I'll test it on a Proxmox VM for now then.

 

[hroberts65616]

I told the system to perform this upgrade on my test unit. and now I can not login. I waited an hour, and then rebooted the server. When I am at the login screen and attempt to login I get the error message Server Error.

Any suggestions on where to start to look to troubleshoot?

 

[Roy Schotpoort]

I told the system to perform this upgrade on my test unit. and now I can not login. I waited an hour, and then rebooted the server. When I am at the login screen and attempt to login I get the error message Server Error.

Any suggestions on where to start to look to troubleshoot?

Try rebooting the machine or ssh to machine and see what service is down.