Hello all,
First of all, merry Christmas and thank you for your awesome work on OPNsense.
Since 2.7.10 (2.7.9 was fine), I am observing a quite strange issue. My setup includes a Captive Portal running on my WireGuard interface and as soon as I enable it, my internet connection is getting slower (after roughly an hour it goes down to 600 Mbit/s from 1 Gbit/s (download)). However, I don't see anything suspicious in the logs.
The authentication of the Captive Portal is "Local + Timebased One Time Password" and I have already recreated the Captive Portal from scratch but the issue keeps showing up.
Could this be related to: https://forum.opnsense.org/index.php?topic=50177.0 ?
Please let me know if you need any further logs or information.
Thank you and best,
a happy OPNsense user
Hi,
It could be related to .10 but not the ticket. Unsure how to debug this, but you can use
# opnsense-revert -r 25.7.9 opnsense
and a reboot to confirm that is the actual issue or not.
Cheers,
Franco
25.7.9 is working flawlessly. So it seems to be related to .10 . Please let me know if I can somehow help you by reproducing this issue and thank you for your support.
After updating to 25.7.10 the issue is back again - so it is definitely related to .10 .
Not just you! I have a similar setup (albeit LDAP) and the same issue.
It eventually takes Opnsense down to the point where latency is in seconds and the internet is unusable.
That are good and bad news at the same time. Bad news: it seems as if there are more user affected, good news: it should be easier to reproduce and seems to be directly traceable to .10 .
Two more data points:
1. If I configure an allowed IP address and ping it regularly, I can observe heavy packet loss.
2. If I change the authentication to "local database", the issue seems to be not present.
Please let me know if you need any further information to reproduce the issue.
How to progress here? Should I open a bug ticket on github.com and if so, is it more likely to be related to the "core" or "src" project?
Seems like the issue is gone with the latest update. Thank you for your support!
I think you have an entirely different problem... the code didn't change from the one introduced in 25.7.10.
Cheers,
Franco