Announcement

Collapse
No announcement yet.

Bcachefs Lands More Fixes Ahead Of Linux 6.7 Stable Debut

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • PuckPoltergeist
    replied
    Originally posted by guzz46 View Post
    I know its very early days, but I tried it recently, I also tried crash testing it, I had a web browser open, a file manager open, a text file open, and was playing a video in mpv when I pressed the restart button, on reboot I had to manually fsck my home partition, I rebooted again and logged into the desktop, where I discovered the video I had been watching was now at 0 bytes, I also did the same test on xfs, jfs, and f2fs, (on jfs I had no video playing, just an open text file) on jfs the text file was at 0 bytes, and on xfs and f2fs everything was fine, they rebooted without issue and I didn't loose any data, I did it a total of five times and never lost any data.
    You're saying you loosed a file that was used read-only during a crash/reset? Is this reproducable?

    Leave a comment:


  • guzz46
    replied
    I know its very early days, but I tried it recently, I also tried crash testing it, I had a web browser open, a file manager open, a text file open, and was playing a video in mpv when I pressed the restart button, on reboot I had to manually fsck my home partition, I rebooted again and logged into the desktop, where I discovered the video I had been watching was now at 0 bytes, I also did the same test on xfs, jfs, and f2fs, (on jfs I had no video playing, just an open text file) on jfs the text file was at 0 bytes, and on xfs and f2fs everything was fine, they rebooted without issue and I didn't loose any data, I did it a total of five times and never lost any data.

    Leave a comment:


  • spicfoo
    replied
    Originally posted by cynic View Post

    i'm not a hater (how does people hate software?)
    Oh please, you can't go by a single day in Phoronix where someone isn't hating on some software or the other constantly. Some ZFS fans bitching about Btrfs on every single Btrfs post is a near guarantee. Same goes for any post about Wayland, Systemd, GNOME or anything at all commonly used. Let's not pretend as if hating a software is remotely uncommon here.

    Leave a comment:


  • pWe00Iri3e7Z9lHOX2Qx
    replied
    Super pumped about bcachefs finally landing in 6.7. I'm going to play around with the tiered storage. I think I can create the topology I want, but I guess I'll find out in a few weeks when 6.7 hits Tumbleweed. I have 2x 4TB NVMe disks that I want in RAID0, and 4x 14TB drives that I want in RAID10, all as one big pool. That's definitely possible. But I want to make sure that the RAID10 pool always (eventually) has a copy of everything that's on the NVMe drives. It looks like the "durability" option will do what I want. Something like...

    Code:
    --label=ssd.ssd1 --durability=0​ /dev/nvme1n1 \
    --label=ssd.ssd2 --durability=0​ /dev/nvme2n1 \
    --label=hdd.hdd1 /dev/sdc \
    --label=hdd.hdd2 /dev/sdd \
    --label=hdd.hdd3 /dev/sde \
    --label=hdd.hdd4 /dev/sdf \
    --replicas=2 \
    --foreground_target=ssd \
    --promote_target=ssd \
    --background_target=hdd​

    Leave a comment:


  • Weasel
    replied
    When you see bug fixes do you:
    1. Rejoice for more bug fixes on your bug-ridden software?
    2. Feel terrible for having used it for so long with so many bugs?
    Optimism/pessimism personality test!

    Leave a comment:


  • Quackdoc
    replied
    Originally posted by fong38 View Post

    Note that it's only reflinks at the moment. It doesn't support FIDEDUPERANGE which is necessary for tools like duperemove to work. Same with ZFS too (though there's a PR implementing it).
    I am aware, however there are numerous dedupe tools that can use reflink for it. which is good enough in the interm​

    Originally posted by pkese View Post

    If you need deduplication, you'll be better off with btrfs, at least for a while.
    Although XFS and bcachefs have some form of deduplication, it is on file level, not on block level.

    Unless of course, you're planning to run RAID5/6 on your 256gb nvme. Then you should stay away from btrfs.
    btrfs has failed on me too many times to bother with it, btrfs for me will never sit on a single one of my drives again.

    Leave a comment:


  • pkese
    replied
    Originally posted by Quackdoc View Post
    I plan on flashing my nvme to bcachefs when I can, I mainly use it for games and for loading large bitrate videos so no major issues will occur, it's only 256gb so having dedupe will be very nice
    If you need deduplication, you'll be better off with btrfs, at least for a while.
    Although XFS and bcachefs have some form of deduplication, it is on file level, not on block level.

    Unless of course, you're planning to run RAID5/6 on your 256gb nvme. Then you should stay away from btrfs.

    Leave a comment:


  • cynic
    replied
    Originally posted by Steffo View Post

    Well, this can also mean: Use it now and it won't eat your data. Near present is also future. 😉
    or it means that it will not eat them in the future because he has already eaten them in the past​!

    Leave a comment:


  • Steffo
    replied
    Originally posted by User29 View Post
    btw it doesn't say "it doesn't eat your data". it actually says: "it won't eat your data". future tense.
    Well, this can also mean: Use it now and it won't eat your data. Near present is also future. 😉

    Leave a comment:


  • steve007
    replied
    Originally posted by cynic View Post
    i'm not a hater (how does people hate software?)
    Years of using Microsoft products?

    Leave a comment:

Working...
X