From mboxrd@z Thu Jan 1 00:00:00 1970 From: ludo@gnu.org (Ludovic =?UTF-8?Q?Court=C3=A8s?=) Subject: bug#33586: "updating substitutes" spam Date: Tue, 04 Dec 2018 15:20:49 +0100 Message-ID: <871s6x1igu.fsf@gnu.org> References: <87k1ksgjnm.fsf@gmail.com> <87pnuhhj3l.fsf@elephly.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]:47994) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1gUBaO-0005CZ-Oh for bug-guix@gnu.org; Tue, 04 Dec 2018 09:22:06 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1gUBaN-0007P5-43 for bug-guix@gnu.org; Tue, 04 Dec 2018 09:22:04 -0500 Received: from debbugs.gnu.org ([208.118.235.43]:55721) by eggs.gnu.org with esmtps (TLS1.0:RSA_AES_128_CBC_SHA1:16) (Exim 4.71) (envelope-from ) id 1gUBaM-0007OY-Fz for bug-guix@gnu.org; Tue, 04 Dec 2018 09:22:03 -0500 Received: from Debian-debbugs by debbugs.gnu.org with local (Exim 4.84_2) (envelope-from ) id 1gUBaM-0007eL-BF for bug-guix@gnu.org; Tue, 04 Dec 2018 09:22:02 -0500 Sender: "Debbugs-submit" Resent-Message-ID: In-Reply-To: <87pnuhhj3l.fsf@elephly.net> (Ricardo Wurmus's message of "Tue, 04 Dec 2018 08:00:30 +0100") List-Id: Bug reports for GNU Guix List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: bug-guix-bounces+gcggb-bug-guix=m.gmane.org@gnu.org Sender: "bug-Guix" To: Ricardo Wurmus Cc: 33586@debbugs.gnu.org Ricardo Wurmus skribis: > Taylan Kammer writes: > >> A minor issue, but I wanted to create a bug-report for this to make >> sure it's on the radar. >> >> During updates, guix may print this message hundreds of times: >> >> substitute: updating substitutes from 'http://...' 100% >> >> Repetitions of this message should probably be coalesced into a single >> message. > > Each of these messages is for a different substitute query. We are > currently not showing the query, so it looks like the same work is done > multiple times. > > The substitute queries are executed at different times. Coalescing the > messages would mean that the messages for previous queries are delayed > until the last query is performed. This would not be good. > > I don=E2=80=99t know if we can reorder the queries so that they all happe= n at > once. In the absence of grafts, the system is designed to perform all the substitute queries at once (with HTTP pipelining and all.) So you see only one message. However, the current implementation of grafts relies on substitutes (when enabled) to determine whether a given package should be grafted. Due to the way this is implemented, the resulted queries are not coalesced, leading to these messages and, more importantly, poor network usage (no pipelining, no connection reuse, etc.) This is really a shortcoming of the current implementation of grafts, much as discussed in . Thanks, Ludo=E2=80=99.