From: Henning Schild <henning.schild@siemens.com>
To: Jan Kiszka <jan.kiszka@siemens.com>
Cc: Baurzhan Ismagulov <ibr@radix50.net>, <isar-users@googlegroups.com>
Subject: Re: [PATCHv7 00/29] base-apt-rework
Date: Wed, 25 Mar 2020 12:25:14 +0100 [thread overview]
Message-ID: <20200325122514.12542375@md1za8fc.ad001.siemens.net> (raw)
In-Reply-To: <7d4eb625-6d64-43b5-a72a-92e938e21998@siemens.com>
On Wed, 25 Mar 2020 12:13:29 +0100
Jan Kiszka <jan.kiszka@siemens.com> wrote:
> On 25.03.20 11:35, Henning Schild wrote:
> > On Tue, 24 Mar 2020 11:49:42 +0100
> > Baurzhan Ismagulov <ibr@radix50.net> wrote:
> >
> >> On Mon, Mar 23, 2020 at 04:25:07PM +0100, Henning Schild wrote:
> >>>>> As pointed out in the other thread, we should try hard to
> >>>>> understand why only this series triggers the issue before we
> >>>>> declare it an upstream problem.
> >>>>
> >>>> As I said, I'd like to check what the problem is.
> >>>
> >>> I think i am responsible to do that. A first quick shot at a CI
> >>> bot did not have the wanted effect ...
> >>>
> >>> Please let me know if you plan to look into it as well, we should
> >>> not duplicate the work. I currently assume the ball is on my side
> >>> again ...
> >>
> >> That would be great, please continue. The CI is up again and is
> >> building your branch. From my builds and the nightly ones some
> >> passed, some failed, have to check what is going on.
> >
> > The problem shows when debootstrap works on an imported cache, a
> > think i repaired/enabled in v7.
> > My guess it that it is less strict about the exact versions to
> > install and might be picking a candidate it should not ... false
> > sharing with other archs of the same suite or something like that.
> >
> > I see two ways to proceed:
> > 1. not import the cache for debootstrap
> > 2. have a close look at the logs and identify the false sharing
> >
> > 1 would mean to always download the minbase again, offline would
> > stil work because of base-apt
> >
> > 2 would mean not downloading twice but having to dig deeper. Here my
> > idea would be to also cache the list of file-names of a debootrap
> > run on the debootstrap export step. The next import would only take
> > files from that list.
> > But that could become complicated since the list might differ i.e.
> > gnupg and https support ... i assume
> >
> > What do you guys think? On the one hand i prefer a deep
> > understanding, on the other hand i would like to eventually get
> > this merged and tested more widely. Not everybody does multiconfig
> > ... and dealing with that i keep thinking ... "i told you so" ...
> > remembering my strong words against the complexity we will have to
> > deal with ... and our users.
>
> I'm also a big fan of understanding the problem before papering over.
> Option 1 can remain on the table, but it should not be picked until
> we know what actually happens.
>
> Can you reproduce the issue locally? Freeze a state that leads to it?
I guess i can, just not feasible in the home-office where i am mainly
conneted via VPN ... Maybe on another machine or on a machine in the
office. I will find a way and dig a bit. Just afraid that this might be
a mix of a weird upstream situation and a bug ... and my repro case
going away with new upstream ... but i guess i should switch to
https://snapshot.debian.org/ which means i have time ;)
Henning
> Jan
>
next prev parent reply other threads:[~2020-03-25 11:25 UTC|newest]
Thread overview: 41+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-03-21 8:31 Henning Schild
2020-03-21 8:31 ` [PATCHv7 01/29] repository: new class to deal with repos Henning Schild
2020-03-21 8:31 ` [PATCHv7 02/29] dpkg-base: add download caching of apt:// downloads Henning Schild
2020-03-21 8:31 ` [PATCHv7 03/29] meta: dpkg-base: convert "exit" into "return 0" Henning Schild
2020-03-21 8:31 ` [PATCHv7 04/29] base-apt: change the sources.list to also offer deb-src Henning Schild
2020-03-21 8:31 ` [PATCHv7 05/29] base-apt: add deb-src packages as well Henning Schild
2020-03-21 8:31 ` [PATCHv7 06/29] base-apt: do not skip gpg check when it is signed Henning Schild
2020-03-21 8:31 ` [PATCHv7 07/29] ci: conf: add "hello" to the sample config and every build Henning Schild
2020-03-21 8:31 ` [PATCHv7 08/29] meta: split all apt-get invocations into download and execution Henning Schild
2020-03-21 8:31 ` [PATCHv7 09/29] meta: create DL_DIR support for all apt-get downloaded .debs Henning Schild
2020-03-21 8:31 ` [PATCHv7 10/29] meta: import DL_DIR debs before apt-get download steps Henning Schild
2020-03-21 8:31 ` [PATCHv7 11/29] meta: include DL_DIR deb import/export into bootstrap Henning Schild
2020-03-21 8:31 ` [PATCHv7 12/29] base-apt: populate from DEBDIR as well Henning Schild
2020-03-21 8:31 ` [PATCHv7 13/29] base-apt: drop the "apt_cache" feature Henning Schild
2020-03-21 8:31 ` [PATCHv7 14/29] base-apt: do not copy debs directly out of rootfss anymore Henning Schild
2020-03-21 8:31 ` [PATCHv7 15/29] base-apt: rework base-apt population Henning Schild
2020-03-21 8:31 ` [PATCHv7 16/29] base-apt: move class "base-apt-helper" into only user Henning Schild
2020-03-21 8:31 ` [PATCHv7 17/29] CI: include "isar-disable-apt-cache" into all CI images Henning Schild
2020-03-21 8:31 ` [PATCHv7 18/29] CI: include "cowsay" into default build to test dpkg-gbp Henning Schild
2020-03-21 8:31 ` [PATCHv7 19/29] CI: set BB_NO_NETWORK for cached rebuild Henning Schild
2020-03-21 8:31 ` [PATCHv7 20/29] meta: repository: implement repo_contains_package and use it in base_apt Henning Schild
2020-03-21 8:31 ` [PATCHv7 21/29] meta: repository: simplify the matching of packages Henning Schild
2020-03-21 8:31 ` [PATCHv7 22/29] meta: deb-dl-dir: do not cache debs from isar-apt Henning Schild
2020-03-21 8:31 ` [PATCHv7 23/29] base-apt: pull base-apt population to the front of the build chain Henning Schild
2020-03-21 8:31 ` [PATCHv7 24/29] meta: base-apt: remove isar-apt check from population Henning Schild
2020-03-21 8:31 ` [PATCHv7 25/29] meta: deb-dl-dir: only export newly downloaded files Henning Schild
2020-03-21 8:31 ` [PATCHv7 26/29] CI: test a custom kernel build in the base-apt offline rebuild Henning Schild
2020-03-21 8:31 ` [PATCHv7 27/29] meta/deb-dl-dir: make subshell verbose and fail on error Henning Schild
2020-03-21 8:31 ` [PATCHv7 28/29] meta: deb-dl-dir: fix "command not found" error when excluding pkgs Henning Schild
2020-03-21 8:31 ` [PATCHv7 29/29] RECIPE-API-CHANGELOG: add a few bits on the base-apt caching changes Henning Schild
2020-03-21 8:41 ` [PATCHv7 00/29] base-apt-rework Henning Schild
2020-03-23 10:56 ` Baurzhan Ismagulov
2020-03-23 11:01 ` Jan Kiszka
2020-03-23 12:22 ` Baurzhan Ismagulov
2020-03-23 15:25 ` Henning Schild
2020-03-24 10:49 ` Baurzhan Ismagulov
2020-03-25 10:35 ` Henning Schild
2020-03-25 11:13 ` Jan Kiszka
2020-03-25 11:25 ` Henning Schild [this message]
2020-03-25 11:14 ` Baurzhan Ismagulov
2020-03-26 12:53 ` Henning Schild
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20200325122514.12542375@md1za8fc.ad001.siemens.net \
--to=henning.schild@siemens.com \
--cc=ibr@radix50.net \
--cc=isar-users@googlegroups.com \
--cc=jan.kiszka@siemens.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox