Pritunl Self Hosted in AWS

We have been hosting multiple Pritunl servers for the past 2 years. Two in AWS and two in our Data Centres.

Since 05:00 GMT our Pritunl instances hosted in AWS showing error messages and we are no longer able to connect to the profiles that they host. We can connect to different profiles hosted by Pritunl instances in our Data Centres.

The EC2 instances have been rebooted, rolled back to a previous instance of Pritunl, checked connectivity to Mongo Atlas DB.

The errors we see from the AWS instances are as follows

AttributeError: ‘NoneType’ object has no attribute ‘vxlan_addr’
UnboundLocalError: local variable ‘network’ referenced before assignment

”systemctl status pritunl.service” shows as started.

Any assistance would be greatly appreciated.

Thanks.

probably issue with last release v1.32.4512.98 if your server run on ubuntu as described here Issue with new release - #3 by zach

We are using Alma Linux in all deployments for Pritunl.

The same for me on Oracle 9 with stable:

  • pritunl-1.32.4512.98-1.el9.oraclelinux.x86_64

and unstable:

  • pritunl-openvpn-2.6.17-1.el9.oraclelinux.x86_64.

I needed to roll it back to the previous release that was working:

  • pritunl-1.32.4400.99-1.el9.oraclelinux.x86_64.

The logs:

[vpn][2026-02-04 10:43:01,333][INFO] Signing certificate...
[vpn][2026-02-04 10:43:01,636][ERROR] Server error occurred while running
Traceback (most recent call last):
  File "/usr/lib/pritunl/usr/lib/python3.9/site-packages/pritunl/server/instance.py", line 1453, in reserve_route_advertisement
'vxlan_addr': self.vxlan.vxlan_addr,
AttributeError: 'NoneType' object has no attribute 'vxlan_addr'

This is being fixed in the v1.32.4528.97 release. This will be on the unstable repositories later tonight.