GNU bug report logs - #32600
‘guix system vm-image’ spends a very long time “registering closures”

Previous Next

Full log

View this message in rfc822 format

From: help-debbugs <at> gnu.org (GNU bug Tracking System)
To: ludo <at> gnu.org (Ludovic Courtès)
Cc: tracker <at> debbugs.gnu.org
Subject: bug#32600: closed (‘guix system
 vm-image’ spends a very long time “registering closures”)
Date: Sun, 23 Sep 2018 21:39:02 +0000

[Message part 1 (text/plain, inline)]

Your message dated Sun, 23 Sep 2018 23:38:03 +0200
with message-id <87mus7rj2c.fsf <at> gnu.org>
and subject line Re: bug#32600: ‘guix system vm-image’ spends a very long time “registering closures”
has caused the debbugs.gnu.org bug report #32600,
regarding ‘guix system vm-image’ spends a very long time “registering closures”
to be marked as done.

(If you believe you have received this mail in error, please contact
help-debbugs <at> gnu.org.)


-- 
32600: http://debbugs.gnu.org/cgi/bugreport.cgi?bug=32600
GNU Bug Tracking System
Contact help-debbugs <at> gnu.org with problems

[Message part 2 (message/rfc822, inline)]

From: ludo <at> gnu.org (Ludovic Courtès)
To: bug-guix <at> gnu.org
Cc: Leo Famulari <leo <at> famulari.name>
Subject: ‘guix system vm-image’
 spends a very long time “registering closures”
Date: Fri, 31 Aug 2018 12:07:19 +0200

Hi,

A couple of days ago, Leo reported that ‘guix system vm-image’ spends a
very long time in the “registering closures” phase:

--8<---------------cut here---------------start------------->8---
<lfam> I wonder about the slowness of 'registering closures' with `guix system
       vm-image` and similiar commands. I don't remember this taking so long
       in the past  [22:52]

[...]

<lfam> Created a bare-bones VM takes ~90 minutes for me recently
<lfam> s/Created/Creating
<civodul> with vm-image?
<lfam> Yes

[...]

<lfam> With #:reset-timestamps #f, I cancelled the task after 10 minutes of
       'registering closures'  [23:33]

[...]

<civodul> lfam: i added some pk's and it seems that it's deduplucation that's
	  killing us  [00:41]
<civodul> because it involves retraversing each store item and reading each
	  file  [00:42]
<lfam> It does it for each file?
<lfam> Or, you could just point me to the relevant code so I can learn it :)
<civodul> yeah, it has to read them to be able to determine if there are
	  duplicates
<civodul> (guix store database) calls the deduplication code  [00:43]
<civodul> which is in (guix store deduplication)
<civodul> (reepca's code from last year's GSoC)
<civodul> maybe c45477d2a1a651485feede20fe0f3d15aec48b39 or something close
	  inadvertently turned on deduplication  [00:44]
<civodul> so in (gnu system vm), we could pass #:deduplicate? #f to
	  'root-partition-initializer'
--8<---------------cut here---------------end--------------->8---

Ludo’.

[Message part 3 (message/rfc822, inline)]

From: ludo <at> gnu.org (Ludovic Courtès)
To: 32600-done <at> debbugs.gnu.org
Cc: Leo Famulari <leo <at> famulari.name>
Subject: Re: bug#32600: ‘guix system vm-image’ spends a very long time “registering closures”
Date: Sun, 23 Sep 2018 23:38:03 +0200

Hello,

ludo <at> gnu.org (Ludovic Courtès) skribis:

> A couple of days ago, Leo reported that ‘guix system vm-image’ spends a
> very long time in the “registering closures” phase:

> <lfam> Created a bare-bones VM takes ~90 minutes for me recently

As of commit fce225471128c4a5246342b3aa1b7e53a8066211, ‘guix system
vm-image bare-bones.tmpl’ runs in ~2m42s on my SSD-based laptop.  The
key commits here are fce225471128c4a5246342b3aa1b7e53a8066211 and
bb3b6ccb05550fbfbeb459c68819a752327d6e1e.

Regarding deduplication (commit
b5460d95e970608bee7a794590be1b051b57d3a3), I found that it doesn’t
change the image size of bare-bones.tmpl (1.2G), but enabling it makes
things slower, ~6m.

Lemme know how it goes for you!

Thanks,
Ludo’.

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.