Re: Btrfs scrub going past 100%

[Paul <p@redscript.org>]

On 7/2/21 5:48 PM, Dan Ritter wrote:

> Paul wrote: 
> > I run regular btrfs scrubs through btrfsmaintenance on my disk array without
> > issues. This time, though, the scrub percentage went over 100%. I kept it
> > running so that I can file a bug if this is one.
> >
> > Which package do I file this against? I believe it's in the Linux kernel not
> > btrfs-progs, but I thought I would ask since I'm new to this.
> >
> > UUID:             xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx
> > Scrub started:    Thu Jul  1 00:03:28 2021
> > Status:           running
> > Duration:         37:13:49
> > Time left:        21358913:01:19
> > ETA:              Mon Feb 11 06:18:36 4458
> > Total to scrub:   29.08TiB
> > Bytes scrubbed:   29.24TiB  (100.56%)
> > Rate:             228.79MiB/s
> > Error summary:    no errors found
> >       
> Reported to the btrfs list last year, doesn't look like there's
> any followup. Check out your time left calculation, too.
>
> https://lore.kernel.org/linux-btrfs/a17a65d0-68d4-e220-0b2d-5c19cdc7be20@petaramesh.org/
>     
They mention "When the filesystem has grown during scrub...". I'm not sure if they mean they increased the size of the FS while it was scrubbing. Definitely not a good idea, IMO.
> It is probably significant that over the last year, I see a fair
> number of issues with crashes and unrecoverable data on the btrfs list.
>     
I've been using Btrfs for several years. It works quite well. I chose it over ZFS for its flexibility. The only issues I had are mostly my idiocy. Thankfully I was able to recover and didn't lose any data.
Just a data point ;)
> On the ZFS list, most of the discussion is about optimal
> configurations.
>
> -dsr-
>
>     
An update about the scrub: an hour or so later the percentage went back to 1%. I don't remember what happened to the other parameters, but I just cancelled it anyway. I didn't have any errors, thankfully, as usual.