[lug] Debugging a startup problem

[Michael J. Hammel]

Also make sure the network devices are available.  Could be the driver
is not finding the phy's right away because the HW is starting to flake
out (see USB problem).  Check /sys/class/net to see if the devices are
showing up there at boot.  Try unloading and reloading the nic driver
and see if anything shows up in the logs or on the console.

On Sun, 2018-07-08 at 20:59 -0600, Davide Del Vento wrote:
> If the problem is reproducible, try also removing and reattaching the
> eth cable and see if that makes the 30 minutes wait become "as soon
> as the cable goes out-and-then-in". Watch the logs for clues while
> doing so...
> 
> On Sun, Jul 8, 2018 at 8:09 PM, Jed S. Baer <blug at jbaer.cotse.net>
> wrote:
> > On Sun, 8 Jul 2018 19:42:09 -0600
> > Lee Woodworth wrote:
> > 
> > > If you don't already have some kind of verbose logging enabled
> > > for init scripts during boot, I would turn that on.
> > > 
> > > Maybe this would help (search terms linux upstart verbose
> > logging):
> > >      https://serverfault.com/questions/601806/upstart-script-not-
> > logging-to-file
> > 
> > Hey, that looks like fun. Just add --debug to the end of the kernel
> > command line at boot. grep will be my friend.
> > 
> > > I had a machine hang during boot because it was mounting a
> > > non-existent nfs export. Ntp can also cause hangs (e.g. ntp-
> > client),
> > > dns lookup failures can also cause issues.
> > 
> > No NFS here, and ntp (I use chrony, actually) would come later,
> > after the
> > network starts.
> > 
> > > You probably know this already, but in my experience power
> > supplies
> > > and their fans are the root causes of many system failures. So
> > > if you start vacuuming, the power supply and case fans may also
> > > benefit from attention.
> > 
> > Been there, done that. Too many times. Wore out the t-shirt. I
> > don't
> > suspect it in this case. It'll be Wednesday at the soonest before I
> > try
> > to tackle this.
> > _______________________________________________
> > Web Page:  http://lug.boulder.co.us
> > Mailing List: http://lists.lug.boulder.co.us/mailman/listinfo/lug
> > Join us on IRC: irc.hackingsociety.org port=6667
> > channel=#hackingsociety
> > 
> 
> _______________________________________________
> Web Page:  http://lug.boulder.co.us
> Mailing List: http://lists.lug.boulder.co.us/mailman/listinfo/lug
> Join us on IRC: irc.hackingsociety.org port=6667
> channel=#hackingsociety
-- 
Michael J. Hammel <mjhammel at graphics-muse.org>