From mboxrd@z Thu Jan 1 00:00:00 1970 From: ludo@gnu.org (Ludovic =?utf-8?Q?Court=C3=A8s?=) Subject: Re: Performance issues with /gnu/store in virtual machines Date: Tue, 22 Aug 2017 16:35:35 +0200 Message-ID: <87r2w33c20.fsf@gnu.org> References: <20170817210547.3fafde57@cbaines.net> <87shgj7vz4.fsf@gnu.org> <20170822122034.6879df0e@cbaines.net> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable Return-path: Received: from eggs.gnu.org ([2001:4830:134:3::10]:43829) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1dkAHL-0001Wc-40 for help-guix@gnu.org; Tue, 22 Aug 2017 10:35:40 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1dkAHJ-0004v2-9u for help-guix@gnu.org; Tue, 22 Aug 2017 10:35:39 -0400 In-Reply-To: <20170822122034.6879df0e@cbaines.net> (Christopher Baines's message of "Tue, 22 Aug 2017 12:20:34 +0100") List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: help-guix-bounces+gcggh-help-guix=m.gmane.org@gnu.org Sender: "Help-Guix" To: Christopher Baines Cc: help-guix@gnu.org Christopher Baines skribis: > On Tue, 22 Aug 2017 12:11:43 +0200 > ludo@gnu.org (Ludovic Court=C3=A8s) wrote: > >> Hi Chris, >>=20 >> Christopher Baines skribis: >>=20 >> > 1.13 0.000019 0 3853 openat=20=20 >>=20 >> [...] >>=20 >> > 13.56 0.193014 50 3853 openat=20=20 >>=20 >> So openat(2) is 4 orders of magnitude slower in the VM, which is a >> problem. >>=20 >> I guess a large part of that is due to the fact that VM created with >> =E2=80=98guix system vm=E2=80=99 use unionfs as their root file system, = which is a >> FUSE thing (see the =E2=80=9Cvolatile=E2=80=9D options in (gnu system vm= ), (gnu system >> linux-initrd), and (gnu build linux-boot).) I=E2=80=99m not sure how we= could >> avoid it. > > I'm unsure if the unionfs will be involved though, as find is searching > in the /gnu/store directory, which is shared with the host: > > TAG_gnu_store on /gnu/store type 9p (ro,relatime,sync,dirsync,trans=3Dv= irtio) > > This might be misunderstanding how mounts work, but this could be > either layered on top, or below the unionfs, and I'm guessing its on > top, which in my mind would mean that the unionfs isn't involved? Oh you=E2=80=99re probably right. It might be that 9p over virtio is this = slow. We should look at what others are doing, I guess it=E2=80=99s a fairly comm= on use case. Ludo=E2=80=99.