Ancestors

Written by h3artbl33d :openbsd: :ve: on 2025-01-11 at 20:36

Our S3 backend was shortly offline - kernel panic. Same issue as before, doesn't seem to be fully resolved.

panic: tcp_do_segment: sent too much

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=282605

=> More informations about this toot | More toots from h3artbl33d@exquisite.social

Toot

Written by h3artbl33d :openbsd: :ve: on 2025-01-11 at 20:45

Because our entire stack runs with encrypted storage, manual intervention is needed. Currently, this is a mild inconvenience.

Aside from this bug, #HardenedBSD is rock solid and a true pleasure to work with.

This bug is in upstream and could, in theory, be MacGuyver'ed around by not using disk encryption and have the machine auromatically come online again.

However, not going to do that. FDE || GTFO :flan_hacker:

=> More informations about this toot | More toots from h3artbl33d@exquisite.social

Descendants

Written by Shawn Webb on 2025-01-11 at 20:47

@h3artbl33d There's progress on this bug: we've nailed it down to one (or more) of three commits coming from #FreeBSD:

If you want to follow along, I'm helping diagnose this issue with someone in the #HardenedBSD Signal Group chat: https://signal.group/#CjQKIAjHzdcnLy4paolCY6yX9CUoAgR7Q-CZYlZxItkQorMoEhBPrt7SR9zSIDdtHStVVopY

=> More informations about this toot | More toots from lattera@bsd.network

Written by h3artbl33d :openbsd: :ve: on 2025-01-11 at 20:52

@lattera

You are truly fantastic :flan_heart:

Two machines 'suffer' from this bug. One runs 15-current, the other one 14-stable. Both run HardenedBSD with the january build.

I'll join the Signal group and help out where I can.

=> More informations about this toot | More toots from h3artbl33d@exquisite.social