Timeout on sync after v18 upgrade

I am running a self-hosted server and after upgrade, syncs from remote machines timeout. Status shows I can still see the remote machine, and syncs on the local machine still work.

I can curl the remote server and it does respond. I just get the error on sync. Are there some logs I can turn on to start debugging if this is machine or not?

Update: I think it may have something to do with DNS resolution. I am running tailscale on both machines. If I use the dns name that resolves to a tailscale ip it doesnt work, but if I change to the local ip address the sync does work. Seems like atuin is using the wrong dns server. Not super surprising consider the mess linux dns resolution is.

I also tried a sync from a remote mac that uses tailscale as well, and that does sync. So while this probably related to my machine, it will come up for other people who use VPNs or Tailscale on linux,.

1 Like

We did introduce timeouts in v18, so this is probably why

Firstly, if you’re able to try the new sync, that would be best. It’s much faster :rocket:

Otherwise, you can see the timeout config here: Config | Atuin Docs

Thanks for the info. Actually the first thing I did was try the new sync. I love it! But it still has the same problem.

1 Like

glad you like it!

hm that is weird. if you run ATUIN_LOG=debug atuin sync, it’ll probably have some traces about DNS resolution

Very strange. It does resolve to the correct IP, and if I use the tailscale ip in the config it also times out.

Maybe the Atuin webserver in configured to not listen to that IP?