Conversation

Notices

1. 受不了包 (shibao@misskey.bubbletea.dev)'s status on Sunday, 19-Feb-2023 14:24:02 JST 受不了包

   xfs or brtfs be like asking do you want to eat the redshat bugs or live in the redshat pod

   • LS (lain@lain.com)'s status on Sunday, 19-Feb-2023 14:23:57 JST LS

     in reply to

     • Moon
     • feinzer
     @shibao @Moon @feinzer btrfs had a bug in their raid code that destroyed data, years after the code functionality was marked stable. At least by now zfs is a thing everywhere so there's no need to even look at btrfs.
   • feinzer (feinzer@ak.airis.work)'s status on Sunday, 19-Feb-2023 14:23:58 JST feinzer

     in reply to
     @shibao i mean i do run a pretty simple setup now so it might shit itself on specific hardware or something
   • 受不了包 (shibao@misskey.bubbletea.dev)'s status on Sunday, 19-Feb-2023 14:23:58 JST 受不了包

     in reply to

     • Moon
     • feinzer

     @feinzer@ak.airis.work idk @Moon@shitposter.club if you can give more context behind your catastropic data losses but for lall pretty sure it was just an unclean shutdown

   • 受不了包 (shibao@misskey.bubbletea.dev)'s status on Sunday, 19-Feb-2023 14:23:59 JST 受不了包

     in reply to

     • feinzer

     @feinzer@ak.airis.work this conversation only brought up again cuz moon mentioned like 5-6 catastropic data loss from brtfs and (afaik) lall's database getting corrupted while also on top of brtfs, but yeah maybe this is all just anecdotal and very skewed but idk

   • 受不了包 (shibao@misskey.bubbletea.dev)'s status on Sunday, 19-Feb-2023 14:24:00 JST 受不了包

     in reply to

     • feinzer

     @feinzer@ak.airis.work literally every time i hear about someone with data corruption on a redundant filesystem it's always brtfs

   • feinzer (feinzer@ak.airis.work)'s status on Sunday, 19-Feb-2023 14:24:00 JST feinzer

     in reply to
     @shibao HOW
     
     The last time I had any data corruption issue was like a decade ago and I was messing with it completely on purpose
     
     Or with android which I was also messing with it and disabling fsync lol
   • feinzer (feinzer@ak.airis.work)'s status on Sunday, 19-Feb-2023 14:24:01 JST feinzer

     in reply to
     @shibao why
     
     Installed fedora with btrfs in my work laptop and it's going great
   • LS (lain@lain.com)'s status on Sunday, 19-Feb-2023 14:27:55 JST LS

     in reply to

     • Moon
     • feinzer
     @shibao @Moon @feinzer apparently user tools didn't even tell you about it for years, https://www.phoronix.com/news/Btrfs-Warning-RAID5-RAID6
     
     Btrfs probably runs fine for 99% of users so I don't think there's much of a need so switch away from it if you're already using it, but I can't find a good reason to use it on a new system.
   • 受不了包 (shibao@misskey.bubbletea.dev)'s status on Sunday, 19-Feb-2023 14:27:56 JST 受不了包

     in reply to

     • Moon
     • LS
     • feinzer

     @lain@lain.com @Moon@shitposter.club @feinzer@ak.airis.work omg that's horrible

   • 受不了包 (shibao@misskey.bubbletea.dev)'s status on Sunday, 19-Feb-2023 14:30:08 JST 受不了包

     in reply to

     • Moon
     • LS
     • feinzer

     @lain@lain.com @Moon@shitposter.club @feinzer@ak.airis.work to me it feels like apache vs nginx, like it's basically the same for most cases except the extreme slim chance where you get a vulnerability or something apache just has 10x more for no reason or whatever, except worse with filesystems cuz permanent

     LS likes this.
   • feinzer (feinzer@ak.airis.work)'s status on Sunday, 19-Feb-2023 14:33:03 JST feinzer

     in reply to

     • Moon
     • LS
     @lain @shibao @Moon but like, is the bug still not fixed that they need to strongly discourage the users from doing it?
     
     I know you might not be able to fix already formatted drives but can't you push an updated revision for future users?
   • LS (lain@lain.com)'s status on Sunday, 19-Feb-2023 14:33:03 JST LS

     in reply to

     • Moon
     • feinzer
     @feinzer @shibao @Moon as far as I can tell this hasn't been fixed because it would involve a new on-disk format and nobody wants to do that.
   • Moon (moon@shitposter.club)'s status on Sunday, 19-Feb-2023 14:34:16 JST Moon

     in reply to

     • LS
     • feinzer
     @shibao @lain @feinzer tons of people switched to nginx because when it came out it was way faster but i think now it's comparable again
   • LS (lain@lain.com)'s status on Sunday, 19-Feb-2023 14:34:16 JST LS

     in reply to

     • Moon
     • feinzer
     @Moon @feinzer @shibao I think pretty much because apache rewrote their code to work more like nginx
   • 受不了包 (shibao@misskey.bubbletea.dev)'s status on Sunday, 19-Feb-2023 14:35:01 JST 受不了包

     in reply to

     • Moon
     • LS
     • feinzer

     @Moon@shitposter.club @lain@lain.com @feinzer@ak.airis.work i just remember seeing a comparison of reported cves for apache vs nginx over the last x years but i can't find it at all so it might be hearsay
     i also remember something about the codebase size being an order of magnitude different but idk

     LS likes this.