Screenshot of GW active tor and correct sdwdate WS with systemcheck error

You can see from the screenshot that Gateway tor is active and Workstation Sdwdate is active with correct time in agreement on both WS and Fedora host. TBB will not connect on WS. WS systemcheck states that control port cannot be reached.

When upgrade-nonroot is performed on WS, sdwdate turns off (x through circular lock icon). Otherwise, if sdwdate is (re)started, time is correct but TBB doesn’t connect.

Next troubleshooting step?

Do you have a connection to tor established on your GW? Or is that failing?

2 Likes

Tor is connecting when sdwdate is on-the-minute correct, otherwise not.

There is a way sdwdate and VM clocks are disrupted on KVM, VirtualBox, and Gnome Boxes. No one has provided a conclusive answer how this is done. htpdate (TAILS) is more reliable. I have to manually reset sdwdate for the tor connection to work for a few minutes if that in KVM. VirtualBox has to be monitored to make certain time is being synchronized correctly but works much more dependably than KVM Whonix on my system.

1 Like

@ReCreateQubesS76G6 That’s very interesting.

I had similar tor connection issues on my KVM based test deployment and found that a change in firewall rules (on the gateway VM) got my tor client connected.

I did see errors about sdwdate as well, so perhaps the fix didn’t fix tor as I had suspected but instead maybe fixed sdwdate which in turn fixed tor.

If you’d like to try it, I could give you a command to run on your gateway and see if it resolves the sdwdate issue and gets you connected reliably.

If it works then it could help to prove a pattern though I’m not sure it will be the final/best fix.

Yes I would like to try that command, but first I need to get my F38 out of emergency shell.

If you know how to do that, I can try your suggestion out.

Thanks : )