Init stops at 30-config: executing about docker-tvheadend HOT 13 CLOSED

What's your host OS?

from docker-tvheadend.

Oh sorry, its Ubuntu 16.04.3.
But i seems there was a very long delay...after 1 hour the 30-config process was done but after that the log is full with messages like this:

2018-01-30 15:42:35.801 [ ALERT] CRASH: Fault address 0x3e8000003d2 (N/A)

By the way, way fast support :)

from docker-tvheadend.

Are you using official docker install or the one in ubuntu?

from docker-tvheadend.

Docker ppa.

from docker-tvheadend.

So it's this one, right?
https://docs.docker.com/install/linux/docker-ce/ubuntu/

from docker-tvheadend.

yes, exactly.

from docker-tvheadend.

Then I don't know what is happening.
You could try the 4.2 release tag and see if the same thing happen there.

from docker-tvheadend.

Already tried. I will create a fresh vm and try it again.
Maybe some problems with my docker host.

I will report.

from docker-tvheadend.

I'm also affected by this.

Right after spinning up the container looking around INSIDE the running container:

root@foo:/$ ps aux
[...]
  182 root       0:00 bash /var/run/s6/etc/cont-init.d/30-config
  185 root       0:00 chown -R abc:abc /config /picons
root@foo:/$ ls -la /picons/ | grep root | wc -l
5458
root@foo:/$ ls -la /picons/ | grep root | wc -l
4056
root@foo:/$ ls -la /picons/ | grep root | wc -l
4056
[...]
root@foo:/$ ls -la /picons/ | grep root | wc -l
3913
root@foo:/$ ls -la /picons/ | grep root | wc -l
2633
root@foo:/$ ls -la /picons/ | grep root | wc -l
2029
root@foo:/$ ls -la /picons/ | grep root | wc -l
1643
root@foo:/$ ls -la /picons/ | grep root | wc -l
557
root@foo:/$ ls -la /picons/ | grep root | wc -l
2
root@foo:/$ ps aux
PID   USER     TIME   COMMAND
    1 root       0:00 s6-svscan -t0 /var/run/s6/services
   30 root       0:00 s6-supervise s6-fdholderd
  186 root       0:00 bash
  316 root       0:00 s6-supervise tvheadend
  441 abc        0:00 /usr/bin/tvheadend -C -c /config
  454 root       0:00 ps aux

This took 5 minutes on my system, so it seems you needed to just wait. But until now I did not debug further why this takes so long.

from docker-tvheadend.

Great detective work waja, you are 1 millimeter away of a workaround, let me bridge the "gap"

( I use Armbian on Orange Pi One)

The picons folder contains over 5700 files, taking the 6 minutes to chown every time the container is started, as you found out

So I added : -v my_picons_folder:/picons
and filled that folder with my only 150 ish icons

You can pick the icons manually or automated with with https://github.com/reuab/picons-source / but that is out of scope of this issue.

I cant see a startup delay anymore.

So what needs to be solved by the lsioarmhf/tvheadend guys

• dont chown the picon files
  or
• dont put 5700 picon files in the image, certainly not in 1 folder
  or
• just explain this workaround on your Repository pages

Thanks guys !

from docker-tvheadend.

@Roger53 I'll check here how long it takes on arm.
This is not the arm repo though. This is x86

from docker-tvheadend.

Hi tobbenb,

Right, I indeed did not notice that this issue is about a X86 repo, as my search machine found it for me and this was exactly my problem, and helped me to workaround the problem as described.

I am also using Docker-ce (18.03.0-ce) on Ubuntu 16 (Welcome to ARMBIAN 5.38 stable Ubuntu 16.04.4 LTS 4.14.18-sunxi) so yes a lot of similarities with above.

I think modern arm and X86 are not so different as many think anyway, except e.g. performance and the amount of available docker images. ;-(

For that I applaud the linuxserver guys !

Should I move to / open same issue and include my workaround in the armhf section ?

from docker-tvheadend.

We have integrated changes to help alleviate this. Please pull from latest or release-4.2 if you are running into this.

from docker-tvheadend.