Bug 205281 - net/avahi-app: runs after mountd causing hostname resolution failures
Summary: net/avahi-app: runs after mountd causing hostname resolution failures
Status: Open
Alias: None
Product: Ports & Packages
Classification: Unclassified
Component: Individual Port(s) (show other bugs)
Version: Latest
Hardware: Any Any
: --- Affects Only Me
Assignee: freebsd-desktop (Team)
URL:
Keywords: needs-patch, needs-qa
Depends on:
Blocks:
 
Reported: 2015-12-13 05:04 UTC by Brendan Molloy
Modified: 2020-11-03 06:55 UTC (History)
6 users (show)

See Also:

Attachments
BEFORE: mountd patch (1.15 KB, patch)
2015-12-14 00:15 UTC, Brendan Molloy 		no flags Details | Diff
avahi-daemon failing to start, with 'BEFORE' line in place (5.56 KB, text/plain)
2018-08-17 13:11 UTC, mvharding 		no flags Details
avahi-daemon starting properly, with 'BEFORE' line removed (5.84 KB, text/plain)
2018-08-17 13:12 UTC, mvharding 		no flags Details
View All Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Brendan Molloy 2015-12-13 05:04:12 UTC 
I have added a line to my /usr/local/rc.d/avahi-daemon to ensure it runs before mountd:

# BEFORE:  mountd

If I don't do this, and I am using avahi for local domain resolution (eg nas.local), it fails to resolve and breaks due to mountd loading before avahi-daemon has.

There may be a better way to solve this, but this is how I fixed it. :)
Comment 1 Rodrigo N. Hernandez freebsd_triage 2015-12-13 06:32:30 UTC 
Hello Brendan. Many thanks for your feedback about the issue. Could you provide ansvn diff against the current ports rcd script? It would be very much appreciated.
Comment 2 Brendan Molloy 2015-12-14 00:15:01 UTC 
Created attachment 164212 [details]
BEFORE: mountd patch
Comment 3 Brendan Molloy 2015-12-22 14:02:17 UTC 
(In reply to Rodrigo N. Hernandez from comment #1)

Any update on this one?
Comment 4 Walter Schwarzenfeld freebsd_triage 2018-01-12 23:37:56 UTC 
Some lines are changed in the initscript.  But not BEFORE..
Is this still relevant?
Comment 5 commit-hook freebsd_committer freebsd_triage 2018-08-13 01:59:52 UTC 
A commit references this bug:

Author: zi
Date: Mon Aug 13 01:59:20 UTC 2018
New revision: 477047
URL: https://svnweb.freebsd.org/changeset/ports/477047

Log:
  - Ensure avahi-daemon starts before mountd

  PR:		205281
  Submitted by:	Brendan Molloy <brendan+freebsd@bbqsrc.net>

Changes:
  head/net/avahi-app/files/patch-initscript_freebsd_avahi-daemon.sh.in
Comment 6 mvharding 2018-08-17 03:52:38 UTC 
This patch keeps avahi-daemon from running at all.  According to https://www.freebsd.org/doc/en_US.ISO8859-1/articles/rc-scripting/article.html, 'BEFORE:' means 'mountd' will run before avahi-daemon, not after.  'rcorder' also now shows a loop, and avahi-daemon goes down in flames since it can't find the dbus socket.  More details tomorrow AM.
Comment 7 mvharding 2018-08-17 04:00:48 UTC 
Sorry, I misspoke about 'BEFORE:' but rcorder generates a loop so rcorder can't figure out what order to run things in. This may be happening on my system because I have gnome or other things installed.  I'll post the 'rcorder' output tomorrow when I have a chance to boot the machine up.
Comment 8 mvharding 2018-08-17 13:11:32 UTC 
Created attachment 196282 [details]
avahi-daemon failing to start, with 'BEFORE' line in place
Comment 9 mvharding 2018-08-17 13:12:17 UTC 
Created attachment 196283 [details]
avahi-daemon starting properly, with 'BEFORE' line removed
Comment 10 commit-hook freebsd_committer freebsd_triage 2018-08-17 13:26:51 UTC 
A commit references this bug:

Author: zi
Date: Fri Aug 17 13:26:17 UTC 2018
New revision: 477424
URL: https://svnweb.freebsd.org/changeset/ports/477424

Log:
  - Revert init script patch due to dependency issue with dbus/mountd
  - Bump PORTREVISION

  PR:		205281

Changes:
  head/net/avahi-app/Makefile
  head/net/avahi-app/files/patch-initscript_freebsd_avahi-daemon.sh.in
Comment 11 Ryan Steinmetz freebsd_committer freebsd_triage 2018-08-17 13:28:07 UTC 
Brendan,

Are you able to dive into this some more?  I think we would need to remove DAEMON from the dbus REQUIRE list, but am unable to dive into what else this might affect at the moment.