pkgsrc-Bugs archive
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]
Re: pkg/58264: Enable FAILOVER_FETCH by default
The following reply was made to PR pkg/58264; it has been noted by GNATS.
From: =?ISO-8859-1?Q?J=F6rg?= Sonnenberger <joerg%bec.de@localhost>
To: pkg-manager%netbsd.org@localhost, gnats-admin%netbsd.org@localhost, pkgsrc-bugs%netbsd.org@localhost,
gnats-bugs%netbsd.org@localhost
Cc:
Subject: Re: pkg/58264: Enable FAILOVER_FETCH by default
Date: Fri, 17 May 2024 02:31:40 +0200
This is a multi-part message in MIME format.
--nextPart26415902.1r3eYUQgxm
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain; charset="utf-8"
On Friday, May 17, 2024 1:20:00=E2=80=AFAM GMT+2 rwhitlock22%gmail.com@localhost wrot=
e:
> >How-To-Repeat:
> Read pkgsrc/mk/fetch/fetch.mk and discover that FAILOVER_FETCH needs to b=
e defined for checksums to be checked.
That's not true. Distfiles are always checked (unless the check is explicit=
ly
disabled) in the "checksum" phase. With FAILOVER_FETCH, the check is
earlier and the newly fetched file is thrown away if it doesn't match. So
unless a mirror later in the list does have the correct file, it will try o=
ver
and over to fetch the url. Given that we have distfiles that are 100s of MB,
that can waste a lot of network IO.
I'm not saying that the current default is the best, but your description
is certainly not right and there is a purpose behind the current setting.
Joerg
--nextPart26415902.1r3eYUQgxm
Content-Transfer-Encoding: quoted-printable
Content-Type: text/html; charset="utf-8"
<html>
<head>
<meta http-equiv=3D"content-type" content=3D"text/html; charset=3DUTF-8">
</head>
<body><p style=3D"margin-top:0;margin-bottom:0;margin-left:0;margin-right:0=
;">On Friday, May 17, 2024 1:20:00=E2=80=AFAM GMT+2 rwhitlock22%gmail.com@localhost w=
rote:</p>
<p style=3D"margin-top:0;margin-bottom:0;margin-left:0;margin-right:0;">>=
; >How-To-Repeat:</p>
<p style=3D"margin-top:0;margin-bottom:0;margin-left:0;margin-right:0;">>=
; Read pkgsrc/mk/fetch/fetch.mk and discover that FAILOVER_FETCH needs to b=
e defined for checksums to be checked.</p>
<br /><p style=3D"margin-top:0;margin-bottom:0;margin-left:0;margin-right:0=
;">That's not true. Distfiles are always checked (unless the check is expli=
citly</p>
<p style=3D"margin-top:0;margin-bottom:0;margin-left:0;margin-right:0;">dis=
abled) in the "checksum" phase. With FAILOVER_FETCH, the check is=
</p>
<p style=3D"margin-top:0;margin-bottom:0;margin-left:0;margin-right:0;">ear=
lier and the newly fetched file is thrown away if it doesn't match. So</p>
<p style=3D"margin-top:0;margin-bottom:0;margin-left:0;margin-right:0;">unl=
ess a mirror later in the list does have the correct file, it will try over=
</p>
<p style=3D"margin-top:0;margin-bottom:0;margin-left:0;margin-right:0;">and=
over to fetch the url. Given that we have distfiles that are 100s of MB,</=
p>
<p style=3D"margin-top:0;margin-bottom:0;margin-left:0;margin-right:0;">tha=
t can waste a lot of network IO.</p>
<br /><p style=3D"margin-top:0;margin-bottom:0;margin-left:0;margin-right:0=
;">I'm not saying that the current default is the best, but your descriptio=
n</p>
<p style=3D"margin-top:0;margin-bottom:0;margin-left:0;margin-right:0;">is =
certainly not right and there is a purpose behind the current setting.</p>
<br /><p style=3D"margin-top:0;margin-bottom:0;margin-left:0;margin-right:0=
;">Joerg</p>
<br /></body>
</html>
--nextPart26415902.1r3eYUQgxm--
Home |
Main Index |
Thread Index |
Old Index