From: Ralph Sennhauser <ralph.sennhauser@gmail.com>
To: Richard Weinberger <richard@nod.at>
Cc: Geert Uytterhoeven <geert@linux-m68k.org>,
Florian Fainelli <f.fainelli@gmail.com>,
Artem Bityutskiy <dedekind1@gmail.com>,
LEDE Development List <lede-dev@lists.infradead.org>,
Adrian Hunter <adrian.hunter@intel.com>,
Christoph Hellwig <hch@infradead.org>,
"linux-mtd@lists.infradead.org" <linux-mtd@lists.infradead.org>,
OpenWrt Development List <openwrt-devel@lists.openwrt.org>
Subject: Re: [OpenWrt-Devel] [RFC] Pre-seeded files/directories for UBIFS
Date: Sun, 21 May 2017 14:23:36 +0200 [thread overview]
Message-ID: <20170521142336.6bdc8782@gmail.com> (raw)
In-Reply-To: <b193086e-322b-16a6-0306-5d9c47068872@nod.at>
Hi Richard
On Sun, 21 May 2017 10:40:05 +0200
Richard Weinberger <richard@nod.at> wrote:
> Geert,
>
> Am 21.05.2017 um 10:37 schrieb Geert Uytterhoeven:
> > On Sat, May 20, 2017 at 9:36 PM, Ralph Sennhauser
> > <ralph.sennhauser@gmail.com> wrote:
> >> There is also the size consideration. Unless a seeded ubifs can get
> >> close to squashfs in terms of compression there would still be a
> >> use-case for squashfs with an ubifs overlay. My current root as
> >> ubifs instead of squashfs is 76.8% bigger.
> >
> > As seeded files are stored and kept unmodified, they could be
> > stored in compressed form?
>
> This is what UBIFS already does.
Right, or it would be some 400-500% difference.
Another advantage of the squashfs with overlay came to mind. OpenWrt
doesn't do a factory reset per se but boots into a "failsafe" mode which
just doesn't mount the overlay. This allows to fix the overlay (some
screwed up config value which prevents connection) or to reset the
password for example without having to reinstall packages and
reconfigure the device from ground up. Wiping the overlay is basically
the last resort.
With the expectation distributions to use the same approach for as many
devices possible I see the seeded ubifs running out of users fast. Some
commercial vendor backporting it to a 3.10 kernel maybe?
If the seeded ubifs could be generalized to snapshot support ala btrfs
that would change things a lot as it would enable uses far beyond just
factory reset. No idea how feasible that is but might be worth
considering instead.
Ralph
>
> Thanks,
> //richard
next prev parent reply other threads:[~2017-05-21 12:23 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-05-20 16:12 [RFC] Pre-seeded files/directories for UBIFS Richard Weinberger
2017-05-20 18:39 ` Florian Fainelli
2017-05-20 18:48 ` Richard Weinberger
2017-05-20 19:36 ` [OpenWrt-Devel] " Ralph Sennhauser
2017-05-20 19:57 ` Richard Weinberger
2017-05-20 21:04 ` Ralph Sennhauser
2017-05-21 8:37 ` Geert Uytterhoeven
2017-05-21 8:40 ` Richard Weinberger
2017-05-21 12:23 ` Ralph Sennhauser [this message]
2017-05-21 15:38 ` Richard Weinberger
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20170521142336.6bdc8782@gmail.com \
--to=ralph.sennhauser@gmail.com \
--cc=adrian.hunter@intel.com \
--cc=dedekind1@gmail.com \
--cc=f.fainelli@gmail.com \
--cc=geert@linux-m68k.org \
--cc=hch@infradead.org \
--cc=lede-dev@lists.infradead.org \
--cc=linux-mtd@lists.infradead.org \
--cc=openwrt-devel@lists.openwrt.org \
--cc=richard@nod.at \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox