Upgrade to v18 and Debian 10 (Buster)

  • Thread starter Thread starter B.lueken
  • Start date Start date
Hi there. I have an Azure Hosted v16 to update. I am not seeing the V18 update option when checking for updates through the web management.

Can you let me know if this is realted to the issues in this thread?

The virtual server is running:

Linux (debian 9.13)

on

Standard D2as v4 (2 vcpus, 8 GiB memory)

with version 16.0.9

Thanks
 
Hi there. I have an Azure Hosted v16 to update. I am not seeing the V18 update option when checking for updates through the web management.

Can you let me know if this is realted to the issues in this thread?

The virtual server is running:

Linux (debian 9.13)

on

Standard D2as v4 (2 vcpus, 8 GiB memory)

with version 16.0.9

Thanks
Yes, it's related to this:
https://www.3cx.com/community/threa...-debian-10-upgrade-failing.86161/#post-405968

As mentioned by StefanW, we should have everything sorted out by the end of today, so give it another try tomorrow.
 
  • Like
Reactions: DSXVOICE
hope this explains why mine won't upgrade to 10... Fingers crossed.
 
  • Like
Reactions: NickD_3CX
Just for information on this thread. I have carried out a sucessful 16 > 18 update this morning and went just fine.

Thank you for your efforts to get that back so quickly.
 
mine failed still. Log attached. Should I start a new thread?

Bash:
Dec 01 12:08:58 deb3cx systemd[1]: Started LSB: Autogenerate and use a swap file.
[33m====== [13:20:21] Preparation: Checking locales[39m
[33m====== [13:20:21] Preparation: Common list 3cxpbx.list found but it seems to be OK[39m
[31m[13:20] Preparation: Found an uncommon source list in /etc/apt/sources.list.d/: 3cxpbx.list.save - 0[39m
> Execute apt command simulation on "install dphys-swapfile"
Reading package lists...
Building dependency tree...
Reading state information...
 
@Jonners
mine failed still. Log attached. Should I start a new thread?
SSH to the machine and move the following file temporarily to a different location outside of the sources lists:
/etc/apt/sources.list.d/3cxpbx.list.save

Run
apt update

Give it another go and if it still fails go ahead and start a new thread here so that we can take a closer look at it.
 
@Jonners

SSH to the machine and move the following file temporarily to a different location outside of the sources lists:


Run


Give it another go and if it still fails go ahead and start a new thread here so that we can take a closer look at it.
Thank you... It failed again. :-( I've got to pick up my youngest from school. Will start a thread when back. Thank you.
 
  • Like
Reactions: ChrisC_3CX
I'm trying to update version 3CX ver 16.0.8.9 to the new version 18 as I have the update option, my install is using Google Cloud with Debian 9. I get the 3cx event log error:

Your 3CX Debian OS failed to update. ID 10012

I've run an apt update via SSH in GCP hoping that might help. Any pointers?
 
I'm trying to update version 3CX ver 16.0.8.9 to the new version 18 as I have the update option, my install is using Google Cloud with Debian 9. I get the 3cx event log error:

Your 3CX Debian OS failed to update. ID 10012

I've run an apt update via SSH in GCP hoping that might help. Any pointers?
P.S. i have checked for the /etc/apt/sources.list.d/3cxpbx.list.save which is not there, just 3cxpbx.list
 
I'm trying to update version 3CX ver 16.0.8.9 to the new version 18 as I have the update option, my install is using Google Cloud with Debian 9. I get the 3cx event log error:

Your 3CX Debian OS failed to update. ID 10012

I've run an apt update via SSH in GCP hoping that might help. Any pointers?
Don't do apt update as that will put you in unsupported terrority. If you try to do the upgrade via the console and it doesn't work, you should receive an email with a log file detailing why. But as always, the recommendation is to take a backup and copy it off the instance and then shutdown the instance and deploy a new one via the customer portal to remain in a supported configuration.
 
Run the following 3 cmds in a terminal
sudo mv /etc/apt/sources.list.d/3cxpbx.list.save /home/administration/Desktop
sudo apt-get --purge remove appstream
wget -O- http://downloads-global.3cx.com/downloads/3cxpbx/public.key | sudo apt-key add -
Take another snapshot and name it - Changes pre reboot
Reboot the "Machine" (Guest)
**Check the Software and Updates to ensure all the boxes are ticked. I found mine were not so the upgrade failed to action and had to redo the previous step***
in a terminal
sudo apt-get update
sudo apt-get upgrade
Take another Snapshot and name it - Upgraded pre Dist upgrade (I also included the cmds I'd used in the description, just in case)
sudo apt-get dist-upgrade
Take another Snapshot and name it - Dist upgraded pre reboot
sudo reboot
In a terminal:
sudo apt-get --purge autoremove cleans out the rubbish
hostnamectl Tells me if upgraded
So now I am running Debian 10 A few extra steps for safety, namely the creation of a lot of Snapshots, but they only take 10 seconds each and can be deleted afterwards.
For some reason, 3CX didn't seem to be installed, but simply in a terminal
sudo apt-get install 3cxpbx
and it was installed and had all my settings
Please don't provide recommendations like this without a disclaimer. First of all, standard Debian deployments won't have the folder /home/administration/Desktop. You have that because you created a user and installed a GUI that wouldn't be there in a normal 3CX installation from ISO. Second, removing the appstream package is not something I've seen 3CX recommend before so I doubt it solves all use cases. I understand that these steps worked for you in your environment, but as your environment as a whole is unsupported from a 3CX standpoint, it's not something you should recommend to others without disclosing that.
 
Just to confirm as the instance manager doesn't specifically offer a shutdown option only a restart.
Also the beta portal offers a reinstall option, can i use that option, and are there any pre-requisites to doing so, other than taking a backup which I have done? Pls find the fail log attached FYI.

It looks like it fails due to the TimeSeries
3CX PhoneSystem 01 CallFlow Server...
Dec 6 22:26:11 gc-custom-europe-west2-2020-03-04-17-26-10-1um8m2oppulr9 collectd[865]: write_gcm: Unsuccessful HTTP request 400: {#012 "error": {#012 "code": 400,#012 "message": "Field timeSeries[2].points[0].interval.start_time had an invalid value of \"2021-12-06T14:25:09.982834-08:00\": The start time must be before the end time (2021-12-06T14:25:09.982834-08:00) for the non-gauge metric 'agent.googleapis.com/agent/api_request_count'.",#012 "status": "INVALID_ARGUMENT"#012 }#012}

I have seen in the Cloud Monitoring API about a 30% fail rate associated with TimeSeries. Clearly there is some date time stamping issue going on where the end time is before the start time somewhere. Presumably a language timezone issue being set incorrectly somewhere?!
 

Attachments

Last edited:
Thank you for the heads up @cobaltit . @Jonners is having issues with the upgrade because he modified the installation. Yes then an automated upgrade becomes difficult. Leave the 3CX install stock and the upgrade goes smoothly. So everybody who has modified the 3CX debian instance and is having issues with upgrading - Take a back up, deploy a new instance to a supported cloud provider or use Hosted by 3CX and then do a restore and dont touch the machine. You can hide that you modified the instance but its just wasting your own and others time because it will show up in the log files.
 
  • Like
Reactions: pmterp
@ochapple as I can see your upgrade failed because of " [31m[22:32] Preparation: Found an uncommon source list in /etc/apt/sources.list.d/: datadog.list - "

Remove this source ist, then do an apt update and try again. If PBX is a vanilla installation and there are not any other modifications, then it will succeed.
If any further modifications are detected, it will rollback again and you will receive a new log file.
 
Looking at the packages listed in that log file I'd recommend the backup and then fresh install via PBXpress. This doesn't look like a vanilla 3CX deployment, or at the very least perhaps an old manual install that would benefit from a refresh
 
  • Like
Reactions: N_G and NikosT_3CX
About 18 months ago I've installed 3CX on a dedicated server using the provided image on the website. Except for automated updates no further modifications were made. I recently tried to upgrade to Debian 10 through the updater but unfortunately the process didn't went through.

Attached you find the generated log file. Any ideas how to fix the problem?
 

Attachments

About 18 months ago I've installed 3CX on a dedicated server using the provided image on the website. Except for automated updates no further modifications were made. I recently tried to upgrade to Debian 10 through the updater but unfortunately the process didn't went through.

Attached you find the generated log file. Any ideas how to fix the problem?
I suspect the problem is probably this:
Code:
Preparation: There seems to be an apt file lock in place. Don't touch anything.

Best thing you can do, reboot the whole server, not just the 3CX Services, then try again.
This does happen sometimes on Debian with apt, but has nothing to do with 3CX.
 
Thanks guys for your help. Indeed it was the Datadog installation that was the problem. Having uninstalled that service, and manually having to uninstall its datadog.list repository file, I was able to run the V18 upgrade no problem. As you say the 3CX upgrade indeed needs a vanilla install to work without issues. I had to edit the permissions of the datadog.list file using sudo chmod from SSH, but once i did that I was able to remove it.
 
Where are the update log files located on the 3cx ISO? The update from debian 9 to 10 seems to go fine, but it seems some 3CX services never start or something. PBX nor webserver are online, and I'm not a skilled enough linux admin to dig much further. The install is unmodified from the 3CX ISO we loaded about a year and a half ago, besides the regular updates.

I took a clone of the botched upgrade so I can find the logs, but reverted to my v16 snapshot. Here is the top about 30 minutes post-OS-upgrade:

fBSstbl.png


For what it's worth, last weekend I updated a different site's instance and it went great! It was a more recent install, but the one-click worked a treat.
 

Forum statistics

Threads
111,975
Messages
590,084
Members
164,901
Latest member
Silent_Guru