On Thu, Feb 11, 2021 at 11:02:51AM +, Rusty Bird wrote:
> Peter Todd:
> > On Fri, Feb 05, 2021 at 06:59:05PM +0100, donoban wrote:
> > > On 2/5/21 5:24 PM, Jinoh Kang wrote:
> > > > - Switch to reflinks. BTRFS has been around 10+ years, so I assume it's
> > > > stable enough?
> > >
> > >
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA512
Peter Todd:
> On Fri, Feb 05, 2021 at 06:59:05PM +0100, donoban wrote:
> > On 2/5/21 5:24 PM, Jinoh Kang wrote:
> > > - Switch to reflinks. BTRFS has been around 10+ years, so I assume it's
> > > stable enough?
> >
> > I've been using BTRFS for a
On 2/11/21 6:06 AM, Peter Todd wrote:
> I'd like to see btrfs added for its checksums as well. I've personally had
> `btrfs scrub` catch hardware problems before more data got silently corrupted.
It isn't default but is official supported [1]. The only problem is that
it could be a little tricky
On Fri, Feb 05, 2021 at 06:59:05PM +0100, donoban wrote:
> On 2/5/21 5:24 PM, Jinoh Kang wrote:
> > Possible solutions would be:
> >
> > - Patch dm-thin. Now to implement this properly, dm-thin must first
> > slice the range, have it ref'd somewhere and then issue discards (so
> > that it
On 2/5/21 5:24 PM, Jinoh Kang wrote:
> Possible solutions would be:
>
> - Patch dm-thin. Now to implement this properly, dm-thin must first
> slice the range, have it ref'd somewhere and then issue discards (so
> that it can handle power disruptions).
> - Switch to reflinks. BTRFS has been
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA256
TL;DR: I'm growingly getting tired of #5426 [1].
Every time I shut down a domain, I have to endure _seconds_ of freezing
in every other VMs of the same LVM Thin pool just because I have discard
passdown enabled. A quickly wearing SSD is the last
