All of lore.kernel.org
 help / color / mirror / Atom feed
From: Joe MacDonald <Joe_MacDonald@mentor.com>
To: <mark.asselstine@windriver.com>, <rehsack@gmail.com>,
	<openembedded-devel@lists.openembedded.org>
Subject: Re: [meta-networking][PATCH] dnsmasq: get systemd only working again
Date: Tue, 19 Apr 2016 10:21:01 -0400	[thread overview]
Message-ID: <20160419142059.GB5970@mentor.com> (raw)
In-Reply-To: <20160419074100.GA9350@ad.chargestorm.se>

[-- Attachment #1: Type: text/plain, Size: 4256 bytes --]

[Re: [meta-networking][PATCH] dnsmasq: get systemd only working again] On 16.04.19 (Tue 09:41) Anders Darander wrote:

> * Joe MacDonald <joe_macdonald@mentor.com> [160419 05:29]:
> 
> > From: Mark Asselstine <mark.asselstine@windriver.com>
> 
> > The changes made in commit 2497cf2960537152427c99629b2af412787eb6c2
> > [dnsmasq: steal resolvconf support from Ubuntu] broke systemd only
> > dnsmasq runtime. No sysvinit scripts are included in systemd only
> > builds (and should not be) and the dnsmasq executable has not moved to
> > /usr/sbin.
> 
> > Reverting to the previous version of the systemd service file. If
> > folks want the local dnsmasq instance to be queried before going to
> > an external DNS they should add 'nameserver 127.0.0.1' to
> > /etc/resolv.conf. Or submit a change which will work with systemd.
> 
> > Signed-off-by: Mark Asselstine <mark.asselstine@windriver.com>
> > Signed-off-by: Joe MacDonald <joe_macdonald@mentor.com>
> > ---
> 
> > Mark,
> 
> > Thanks for the patch.  In the interests of not causing any more hassle
> > than I already have, I'm proposing this as an alternative to what you sent
> > earlier today.  This hopefully preserves the resolvconf/start-stop
> > functionality while not relying on any sysvinit-ness.  What do you think
> > about this as a middle-ground?
> 
> > Anders,
> 
> > I trust that nothing Mark and I are discussing here changes the issue you
> > noted about dnsmasq just falling over after a few minutes.  I propose we
> > dig into that separately, since it looks like a different problem probably
> > related to the uprev.  If it turns out that once we merge something to
> > support systemd-only builds that resolves the 'dnsmasq vandalizes itself'
> > issue, even better.
> 
> Well, I've tried this patch, but dnsmasq isn't running. Was it run-time
> tested?

Yes, it worked on my setup last night, but the testing I did was a bit
of a fake due to trying it out on a live system rather than waiting for
the filesystem generation to happen and copying it over to the target.
Thus the error you noted below.

> Isn't MAINPID the PID of the main process started by systemd?

In the case of Type=forking, $MAINPID is the value out of PIDFile unless
PIDFile is unset, then systemd needs to guess at the value (unless
you've also set GuessMainPID='no', in which case I have no idea what
happens, but the default is 'yes', so it shouldn't matter in this case).
At least that's my reading of the docs at
https://www.freedesktop.org/software/systemd/man/systemd.service.html,
I'm no expert when it comes to systemd service files, though, so please
correct me if I'm wrong.

> And -x should specify a path to the pid-file?

Yes, it absolutely should.

-J.

> Logs below.
> 
> Cheers,
> Anders
> 
> [FAILED] Failed to start DNS forwarder and DHCP server.
> See 'systemctl status dnsmasq.service' for details.
> 
> Running `systemctl status dnsmasq.service` gives:
> 
> [[0;1;31m●[[0m dnsmasq.service - DNS forwarder and DHCP server    Loaded: loaded (/lib/systemd/system/dnsmasq.service; enabled; vendor preset: enabled)
>    Active: [[0;1;31mfailed[[0m (Result: exit-code) since Fri 2016-04-08 11:07:27 CEST; 8min ago
>   Process: 384 ExecStart=/usr/bin/dnsmasq -x $MAINPID [[0;1;31m(code=exited, status=1/FAILURE)[[0m
>   Process: 381 ExecStartPre=/usr/bin/dnsmasq --test (code=exited, status=0/SUCCESS)
> 
> Apr 08 11:07:27 ccu systemd[1]: Starting DNS forwarder and DHCP server...
> Apr 08 11:07:27 ccu dnsmasq[381]: dnsmasq: syntax check OK.
> Apr 08 11:07:27 ccu dnsmasq[384]: dnsmasq: bad command line options: try --help
> Apr 08 11:07:27 ccu dnsmasq[384]: [[0;1;31mbad command line options: try --help[[0m
> Apr 08 11:07:27 ccu dnsmasq[384]: [[0;1;31mFAILED to start up[[0m 
> Apr 08 11:07:27 ccu systemd[1]: [[0;1;39mdnsmasq.service: Control process exited, code=exited status=1[[0m
> Apr 08 11:07:27 ccu systemd[1]: [[0;1;31mFailed to start DNS forwarder and DHCP server.[[0m
> Apr 08 11:07:27 ccu systemd[1]: [[0;1;39mdnsmasq.service: Unit entered failed state.[[0m
> Apr 08 11:07:27 ccu systemd[1]: [[0;1;39mdnsmasq.service: Failed with result 'exit-code'.[[0m
> 
-- 
-Joe MacDonald.
:wq

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 484 bytes --]

  reply	other threads:[~2016-04-19 14:21 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-04-19  3:30 [meta-networking][PATCH] dnsmasq: get systemd only working again Joe MacDonald
2016-04-19  7:41 ` Anders Darander
2016-04-19 14:21   ` Joe MacDonald [this message]
2016-04-19 15:37     ` [meta-networking][PATCH v3] " Joe MacDonald
2016-04-20  8:24       ` Anders Darander
2016-04-20  8:29         ` Anders Darander
2016-04-20 18:56           ` Joe MacDonald
2016-04-20 19:09             ` Joe MacDonald
2016-04-21 13:39               ` Anders Darander
2016-04-19 13:54 ` [meta-networking][PATCH] " Mark Asselstine
  -- strict thread matches above, loose matches on Subject: below --
2016-04-18 20:34 Mark Asselstine
2016-04-19  4:39 ` Anders Darander

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=20160419142059.GB5970@mentor.com \
    --to=joe_macdonald@mentor.com \
    --cc=mark.asselstine@windriver.com \
    --cc=openembedded-devel@lists.openembedded.org \
    --cc=rehsack@gmail.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.