!MthpOIxqJhTgrMNxDS:nixos.org

NixOS ACME / LetsEncrypt

92 Members
Another day, another cert renewal42 Servers

Load older messages


SenderMessageTime
3 Jun 2024
@stephank:stephank.nlStéphanI just have no idea how to test it08:49:04
@stephank:stephank.nlStéphanI reused the fixperms service, because I was worried about bind mounts. I'm not sure if bind mounts are preserved from ExecStartPre to ExecStart, or if they are recreated correctly when the underlying directory changed.08:50:13
@stephank:stephank.nlStéphan Now that I think about it, maybe a simple -e or -d check won't work because the $newHash directory will always be created via BindPaths? 08:51:01
@stephank:stephank.nlStéphanLooks like it's always created: https://www.freedesktop.org/software/systemd/man/latest/systemd.exec.html#BindPaths=08:51:59
@stephank:stephank.nlStéphanOh wait, the fixperms / migration service doesn't use BindPaths. So what I cooked up there may work.08:55:41
@arianvp:matrix.orgArianWe have quite an extensive NixOS test which we could change. But doing NixOS tests for "transitions" is always a bit tricky 09:22:26
@arianvp:matrix.orgArian Stéphan: dont this based on stateVersion wont work 09:24:03
@arianvp:matrix.orgArian * Stéphan: doing this based on stateVersion wont work I think 09:24:23
@arianvp:matrix.orgArianah no nvm forget what I said09:26:04
@arianvp:matrix.orgArianHmm how do we handle people who rollback boot into a 23.11 configuration ? :/ 09:29:01
@arianvp:matrix.orgArianmaybe we should symlink or copy the directory instead of moving? idk. Maybe it's not worth it supporting rollbacks. Just thinking out loud of another failure mode here09:29:39
@stephank:stephank.nlStéphanI like that idea, but no idea if a symlink works09:31:44
@arianvp:matrix.orgArian

urgh I feel bad that this shipped. The whole reason we created the ACME Team was to catch these kind of things :(

I’m kind of confused why automation didnt tag us for review. Given we’re set up as meta.maintainers for that module. maybe that automation doesn’t exist for NixOS? only for nixpkgs?

09:34:37
@stephank:stephank.nlStéphan I think BindPaths should follow a symlink, because I found this, which says not following a symlink is something you otherwise need to specify explicitely: https://github.com/systemd/systemd/issues/32366 09:35:27
@stephank:stephank.nlStéphan
In reply to @arianvp:matrix.org

urgh I feel bad that this shipped. The whole reason we created the ACME Team was to catch these kind of things :(

I’m kind of confused why automation didnt tag us for review. Given we’re set up as meta.maintainers for that module. maybe that automation doesn’t exist for NixOS? only for nixpkgs?

See: https://github.com/NixOS/nixpkgs/pull/270221#issuecomment-2144652323
09:36:12
@arianvp:matrix.orgArianah awesome09:36:19
@arianvp:matrix.orgArianOh well no hard feelings. we’re all human. just makes my blood pressure go up to know I’m breaking people’s TLS setup :’) 09:37:15
@arianvp:matrix.orgArianAdded a WIP PR to add the team to code-owners: https://github.com/NixOS/nixpkgs/pull/316854 09:45:09
@arianvp:matrix.orgArianIf there are any volunteers to join the team just yell ;) 09:45:39
@sandro:supersandro.deSandro 🐧You could symlink the old hash to the new one if the new directory doesn't exist and the contents are similar enough to be compatible09:52:47
@sandro:supersandro.deSandro 🐧
In reply to @arianvp:matrix.org
maybe we should symlink or copy the directory instead of moving? idk. Maybe it's not worth it supporting rollbacks. Just thinking out loud of another failure mode here
Copy means you have old, potentially ran out certs
09:52:47
@sandro:supersandro.deSandro 🐧
In reply to @arianvp:matrix.org
(and it's important. E.g. German government has been issueing LEtsEncrypt certificates for a lot of XMPP servers through MITM'ing through middleboxes at Hetzner datacenters and got caught redhanded multiple times last year)

I know of the one case that went on Hackernews.

DNS challenge works against that, does it?

09:52:47
@sandro:supersandro.deSandro 🐧I have a PR for a test improvement open for 4 months to prevent sich issues in the future and no one really cared, so I just gave up https://github.com/NixOS/nixpkgs/pull/28699909:52:47
@arianvp:matrix.orgArianYeh no blame on you at all. 09:53:22
@sandro:supersandro.deSandro 🐧Going back to null is also not that great because then we rely on the lego defaults which could change in the future09:56:08
@sandro:supersandro.deSandro 🐧If you have a change I could test, throw it over the fence10:00:00
@arianvp:matrix.orgArianyeh I think the only solution is to do some state mangling. Or just put in the release notes that the hash changed and call it a day 10:00:10
@sandro:supersandro.deSandro 🐧I really thought we already had that in the release notes...10:00:36
@arianvp:matrix.orgArianWe used to have bugs where we would recreate the same account multiple times: https://github.com/NixOS/nixpkgs/pull/106857 and the account creation rate limiting is very aggressive (5 per day?) But I think we dont run into that issue anymore10:00:39
@arianvp:matrix.orgArian So the rate-limit issue is probably less of a problem; unless you have A lot of domains 10:01:25

Show newer messages


Back to Room ListRoom Version: 6