1 Return-Path: <tomi.ollila@iki.fi>
\r
2 X-Original-To: notmuch@notmuchmail.org
\r
3 Delivered-To: notmuch@notmuchmail.org
\r
4 Received: from localhost (localhost [127.0.0.1])
\r
5 by olra.theworths.org (Postfix) with ESMTP id 8EBEB431FAF
\r
6 for <notmuch@notmuchmail.org>; Wed, 8 Feb 2012 01:59:38 -0800 (PST)
\r
7 X-Virus-Scanned: Debian amavisd-new at olra.theworths.org
\r
11 X-Spam-Status: No, score=0 tagged_above=-999 required=5 tests=[none]
\r
13 Received: from olra.theworths.org ([127.0.0.1])
\r
14 by localhost (olra.theworths.org [127.0.0.1]) (amavisd-new, port 10024)
\r
15 with ESMTP id kEOkO54WAUez for <notmuch@notmuchmail.org>;
\r
16 Wed, 8 Feb 2012 01:59:37 -0800 (PST)
\r
17 Received: from guru.guru-group.fi (guru-group.fi [87.108.86.66])
\r
18 by olra.theworths.org (Postfix) with ESMTP id B8D41431FAE
\r
19 for <notmuch@notmuchmail.org>; Wed, 8 Feb 2012 01:59:36 -0800 (PST)
\r
20 Received: by guru.guru-group.fi (Postfix, from userid 501)
\r
21 id ADC3668055; Wed, 8 Feb 2012 11:59:36 +0200 (EET)
\r
22 From: Tomi Ollila <tomi.ollila@iki.fi>
\r
23 To: David Edmondson <dme@dme.org>, notmuch@notmuchmail.org
\r
24 Subject: Re: [PATCH v3] emacs: Call "notmuch tag" once when applying tag
\r
25 changes to a thread.
\r
26 In-Reply-To: <1328690169-6991-1-git-send-email-dme@dme.org>
\r
27 References: <1328632303-31877-1-git-send-email-jani@nikula.org>
\r
28 <1328690169-6991-1-git-send-email-dme@dme.org>
\r
29 User-Agent: Notmuch/0.11.1+164~g6619341 (http://notmuchmail.org) Emacs/23.3.1
\r
30 (x86_64-unknown-linux-gnu)
\r
31 X-Face: HhBM'cA~<r"^Xv\KRN0P{vn'Y"Kd;zg_y3S[4)KSN~s?O\"QPoL
\r
32 $[Xv_BD:i/F$WiEWax}R(MPS`^UaptOGD`*/=@\1lKoVa9tnrg0TW?"r7aRtgk[F
\r
33 !)g;OY^,BjTbr)Np:%c_o'jj,Z
\r
34 Date: Wed, 08 Feb 2012 11:59:36 +0200
\r
35 Message-ID: <m2liody7av.fsf@guru.guru-group.fi>
\r
37 Content-Type: text/plain; charset=us-ascii
\r
38 X-BeenThere: notmuch@notmuchmail.org
\r
39 X-Mailman-Version: 2.1.13
\r
41 List-Id: "Use and development of the notmuch mail system."
\r
42 <notmuch.notmuchmail.org>
\r
43 List-Unsubscribe: <http://notmuchmail.org/mailman/options/notmuch>,
\r
44 <mailto:notmuch-request@notmuchmail.org?subject=unsubscribe>
\r
45 List-Archive: <http://notmuchmail.org/pipermail/notmuch>
\r
46 List-Post: <mailto:notmuch@notmuchmail.org>
\r
47 List-Help: <mailto:notmuch-request@notmuchmail.org?subject=help>
\r
48 List-Subscribe: <http://notmuchmail.org/mailman/listinfo/notmuch>,
\r
49 <mailto:notmuch-request@notmuchmail.org?subject=subscribe>
\r
50 X-List-Received-Date: Wed, 08 Feb 2012 09:59:38 -0000
\r
52 On Wed, 8 Feb 2012 08:36:09 +0000, David Edmondson <dme@dme.org> wrote:
\r
53 > Optimize thread tagging by combining all the tagging operations to a
\r
54 > single "notmuch tag" call.
\r
56 > For threads in the order of tens or a hundred inbox tagged messages,
\r
57 > this gives a noticeable speedup. On two different machines, archiving
\r
58 > a thread of about 50 inbox tagged messages goes down from 10+ seconds
\r
59 > to about 0.5 seconds.
\r
61 > The bottleneck is not within emacs; the same behaviour can be observed
\r
62 > in the CLI. This approach has the added benefit of being more
\r
63 > reliable: any of the individual tagging operations might face a locked
\r
64 > database, leading to partial results.
\r
66 > This introduces a limitation to the number of messages that can be
\r
67 > archived at the same time (through ARG_MAX limiting the command
\r
68 > line). While at least on Linux this seems more like a theoretical
\r
69 > limitation than a real one, it could be avoided by archiving at most a
\r
70 > few hundred messages at a time.
\r
72 I did a simple test program:
\r
74 --8<----8<----8<----8<----8<----8<----8<----8<----8<----8<----8<--
\r
80 die "Usage: $0 <arglen> <# of args>\n" unless @ARGV == 2;
\r
82 my $arg = 'x' x $ARGV[0];
\r
83 my @args = ( $arg ) x $ARGV[1];
\r
85 print "One arg: '$arg'\n";
\r
86 print 'Number of args: ', scalar @args, "\n";
\r
88 exec '/bin/true', @args;
\r
89 --8<----8<----8<----8<----8<----8<----8<----8<----8<----8<----8<--
\r
91 This program executes /bin/true with given number of args all args
\r
92 of some length: i.e.
\r
94 ./test_cmdlimit 100 10000
\r
96 makes 10000 100 character arguments (total of million characters)
\r
97 and executes /bin/true with those ten thousand 100-char args.
\r
99 In one machine where getconf ARG_MAX returns 131072
\r
100 (Debian Lenny ia32)
\r
102 ./test_cmdlimit 200 10000 succeeds
\r
104 ./test_cmdlimit 20 100000 gives
\r
105 Can't exec "/bin/true": Argument list too long at ./test_cmdlimit.pl line 14.
\r
107 Hmm, actually there:
\r
108 ./test_cmdlimit 19 100000 succeeds and
\r
109 ./test_cmdlimit.pl 209 10000 fails
\r
111 More with args which lengths are 1 and 2 chars:
\r
113 ./test_cmdlimit.pl 1 1046163 succeeds
\r
114 ./test_cmdlimit.pl 1 1046164 fails
\r
116 ./test_cmdlimit.pl 2 697442 succeeds
\r
117 ./test_cmdlimit.pl 2 697443 fails
\r
119 1046163 was close to 1048576 (1024 * 1024) --
\r
120 697442 * 2 is 1394884...
\r
122 >From these I can make an educated guess that when message id:s
\r
123 are typically between 30 to 70 characters something like
\r
124 20 000 messages are safe to be tagged at once in this
\r
125 test system (./test_cmdlimit.pl 50 40000 succeeds).
\r
128 > Based on code from Jani Nikula <jani@nikula.org>.
\r