- When performing an upgrade is always a good idea to backup your production system. It is also good to perform the update on a test system first in order to mitigate risks in production.
- It is also NOT recommended you use the administration web interface to upgrade your system from one version to another. The administration web interface for upgrading is best used for minor bug updates.
- You can upgrade from any version to any version as long as you follow the manual steps both before and after your upgrade.
- If you have multiple systems in a cluster (all managed by one Management server). Complete upgrade process on the master server before proceeding to the slave nodes. Once all the systems have been upgraded, log in to the the management UI, go to
System -> Servers, select all the servers and press
Send Profiles. Reboot any services when administration service instructs you to do so.
Step 1. Follow each of the pre-install instructions starting with the version you currently have installed. Then come back to this step and continue. So if you currently have version 4.0.4 installed, you want to start with the instructions for 4.0.4 and then proceed to the end of the pre-install instructions for each version.
Then in a root window on the system you wish to upgrade run
Step 3. Shutdown sipxecs and update your system.
It's good to let your system update all its packages because
- The binaries were probably built with the latest packages
- If ever you needed to update a package for security reasons, you'd want to minimize the changes to your system
- It's often a requirement for getting assistance on mailing list
- If you do have any problems, most likely others will see the problem as well and provide a quick solution to your issue.
Step 4. Follow each of the post-install instructions starting with the version you currently have installed. Then come back to this step and continue.
Step 5. Start your system back up.
Complete. if you are upgrading a clustered system, see the Notes section at the top of this document for upgrading each slave node.
Before upgrading from 3.6 or earlier
Consider upgrading to sipXecs 3.10 then following instructions from there.
Before upgrading from 3.10.x
The local domain bind zone is "emptied", if you are using the system as a DNS server, you should copy or backup your zone file before the update. It doesn't effect other zone files on the system, just the one sipx is attached to.
Before upgrading from 4.0.4
If you originally installed your system from the sipfoundry ISO in 4.0.4 and are having issues updating, here's some help:
First you have to determine what your redhat-release is:
If it doesn't say CentOS but says (sipXecs 4.x), then here's what you need to do, AFTER YOU GET A PROPER BACKUP..
Rename your repo file:
If you do not have a CentOS-Base.repo or other repos in there, it's OK, this will probably fix a lot, just hold on.
Edit the repo to make it JUST a CentOS repo again - remove these lines and save it:
Before upgrading from 4.2.x
Remove the old repo file:
Before upgrading from 4.4.0
Version 4.4.0 is available on CentOS 5 but not CentOS 6. Version 4.6.0 is available on CentOS 6 but not CentOS 5. Therefore you must make a backup your 4.4.0 system, then install a fresh 4.6.0 system, then restore your 4.4 backups by uploading them from the "Restore" web page. I would strongly suggest you test the this process in a lab to reduce the change you will encounter any problems when upgrading the real system as it will be time consuming to revert back. For example, a VMWare virtual machine would be fine or any virtual environment just to ensure restore goes smoothly. The test machine does not have to have same IP address as the original machine as was the case in prior versions.
Be sure to read "After Upgrading to 4.6.0" in this document for notes about restoring from backups.
After upgrading from 3.10.x
The DHCP 120 test will fail if you are using the on board DHCP server. To correct this you need to add the following to the /etc/dhcpd.conf and restart dhcpd (service dhcpd restart). This is not required if you aren't using the Counterpath Enterprise Clients.
After upgrading from 4.0.4
If you have voicemail messages from the old version, let's say <=4.0.X, you might hit the XX-9461
A quick fix for that is:
If you have customized versions of vxml files, you'll lose them starting on 4.2.0 as VXML are no longer used for Voicemail.
After upgrading from 4.2.x
No special instructions
After upgrading from 4.4.0
If you wish to install OpenACD run
If you wish to install support for audiocodes HD phones
After upgrading from 4.6.0
It's unlikely you upgraded directly to 4.6.0 because installation on CentOS required a full reinstall. You probably restored from backups.
When restoring from backups, not all settings may have migrated correctly. You will need to reset which services you want enabled on each machine. These were called "roles" in 4.4, and are now simply a list of services. Navigate to menu in administration user interface : System/Servers. There you will see links along the left hand side of the web page such as "Core Services" or "Telephony Services". Go through each link and enable the services you would like to have running. Hover your mouse over each service for more information about the service and what function it provides. Services will automatically start in about 10-30 seconds after you've applied your changed.
If the 'superadmin' password is not correctly decrypted during the restore, you can use 'service sipxconfig reset-admin' to reset it. You can then log in to the Admin GUI with no password and set a new password.