systemd post-install failure

systemd post-install failure

[Jack Mitchell]

[-- Attachment #1: Type: text/plain, Size: 650 bytes --]

Booting master today I had a systemd fault. Enabling debugging I 
captured the (attached) output. The gist of which is:

.....

Spawned /lib/systemd/system-generators/systemd-system-update-generator as 81

.....

Received SIGCHLD from PID 81 (n/a).
Got SIGCHLD for process 90 (systemd-sysctl)
Child 90 died (code=exi

So, looking into systemd-system-update-generator, it seems this is what 
takes card of the post install scripts, which I know we're having 
problems with.

Is there anything I can do to help?

Cheers,

-- 

   Jack Mitchell (jack@embed.me.uk)
   Embedded Systems Engineer
   http://www.embed.me.uk

--


[-- Attachment #2: screenlog.0 --]
[-- Type: application/octet-stream, Size: 26311 bytes --]

Re: systemd post-install failure

[Burton, Ross]

On 22 April 2013 15:13, Jack Mitchell <ml@communistcode.co.uk> wrote:
> Booting master today I had a systemd fault. Enabling debugging I captured
> the (attached) output. The gist of which is:
>
> So, looking into systemd-system-update-generator, it seems this is what
> takes card of the post install scripts, which I know we're having problems
> with.

Not quite - the system update generator would be used if we were using
systemd's native systemd update protocol, which we're not.  This
generator shouldn't do anything unless you've got a symlink
/system-update (see
http://freedesktop.org/wiki/Software/systemd/SystemUpdates for the
details).

Ross

Re: systemd post-install failure

[Jack Mitchell]

On 22/04/13 15:40, Burton, Ross wrote:
> On 22 April 2013 15:13, Jack Mitchell <ml@communistcode.co.uk> wrote:
>> Booting master today I had a systemd fault. Enabling debugging I captured
>> the (attached) output. The gist of which is:
>>
>> So, looking into systemd-system-update-generator, it seems this is what
>> takes card of the post install scripts, which I know we're having problems
>> with.
> Not quite - the system update generator would be used if we were using
> systemd's native systemd update protocol, which we're not.  This
> generator shouldn't do anything unless you've got a symlink
> /system-update (see
> http://freedesktop.org/wiki/Software/systemd/SystemUpdates for the
> details).
>
> Ross

Ah, ok then, I don't have a /system-update. So square one, any ideas why 
this would have suddenly flagged up? I was building and booting fine on 
Wed/Thurs. That was the only useful bit of the debug log I could 
determine...

-- 

   Jack Mitchell (jack@embed.me.uk)
   Embedded Systems Engineer
   http://www.embed.me.uk

--

Re: systemd post-install failure

[Burton, Ross]

On 22 April 2013 15:44, Jack Mitchell <ml@communistcode.co.uk> wrote:
> Ah, ok then, I don't have a /system-update. So square one, any ideas why
> this would have suddenly flagged up? I was building and booting fine on
> Wed/Thurs. That was the only useful bit of the debug log I could
> determine...

That log isn't massively useful, can you explain exactly what isn't working?

Ross

Re: systemd post-install failure

[Jack Mitchell]

On 22/04/13 15:56, Burton, Ross wrote:
> On 22 April 2013 15:44, Jack Mitchell <ml@communistcode.co.uk> wrote:
>> Ah, ok then, I don't have a /system-update. So square one, any ideas why
>> this would have suddenly flagged up? I was building and booting fine on
>> Wed/Thurs. That was the only useful bit of the debug log I could
>> determine...
> That log isn't massively useful, can you explain exactly what isn't working?
>
> Ross

Tell me about it, I boot, systemd crashes, I get a capture of the serial 
log up to where it crashed.

I've managed to get a tiny little bit more output that may shed a little 
bit more information.

_*scrap that*_

My power supply was crapping out on me, so when the going got a bit 
tough the cpu required some more amps and bam, dead system.

Sorry for the noise. Should have checked that earlier.


-- 

   Jack Mitchell (jack@embed.me.uk)
   Embedded Systems Engineer
   http://www.embed.me.uk

--