Mongrel users - Apr 2007 - Mongrels dying on FreeBSD 5.5-STABLE......why why why?

Yo Zed and everyone else, I''m having a major problem I''m
hoping someone can
help with.

I''ve been running mongrel clusters for a few months with no problems on
a
couple of my boxes. They both run Debian...

I recently moved one of my older Rails apps on a FreeBSD 5.5 box to mongrel
as well. Everything runs quickly and wonderfully - when it''s running!
My
problem is that over time each one of my 5 mongrel processes die.
There''s
nothing in the mongrel.log telling me why they die. They just die.

My question then is two-fold

1: Why are my mongrel processes dying? Where''s the log information?
2: Can these guys be automatically restarted somehow???

I''m on mongrel 1.0.1 and mongrel_cluster 1.0.1.1

HELP! My clients are going apeshit and want to cut my throat :p


--------------------
seth at subimage interactive
-----
http://www.subimage.com
http://sublog.subimage.com
-----
http://www.getcashboard.com
http://dev.subimage.com/projects/substruct
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://rubyforge.org/pipermail/mongrel-users/attachments/20070403/f8f32bdb/attachment.html

On Tue, 3 Apr 2007 11:05:50 -0700
"subimage interactive" <subimage at gmail.com> wrote:
> Yo Zed and everyone else, I''m having a major problem I''m
hoping someone can
> help with.
> 
> I''ve been running mongrel clusters for a few months with no
problems on a
> couple of my boxes. They both run Debian...
> 
> I recently moved one of my older Rails apps on a FreeBSD 5.5 box to mongrel
> as well. Everything runs quickly and wonderfully - when it''s
running! My
> problem is that over time each one of my 5 mongrel processes die.
There''s
> nothing in the mongrel.log telling me why they die. They just die.
You''ll have to be more specific than that.  "Die" could mean
they crash
and the process goes away, or that the process stops responding to
requests.  For the first kind you''ll need to turn on core dumps and do
a gdb inspect after.  With the second one you''ve gotta list out the
various gems you''re using and when you find a dead mongrel then attach
to it with strace or gdb to find out where it''s blocked.
> My question then is two-fold
> 
> 1: Why are my mongrel processes dying? Where''s the log
information?
> 2: Can these guys be automatically restarted somehow???
Yes, monit works great, other folks like runit.  There''s a bunch of
tutorials for both.

-- 
Zed A. Shaw, MUDCRAP-CE Master Black Belt Sifu
http://www.zedshaw.com/
http://www.awprofessional.com/title/0321483502 -- The Mongrel Book
http://mongrel.rubyforge.org/

> > I recently moved one of my older Rails apps on a FreeBSD 5.5 box to
> mongrel
> > as well. Everything runs quickly and wonderfully - when it''s
running! My
> > problem is that over time each one of my 5 mongrel processes die.
> There''s
> > nothing in the mongrel.log telling me why they die. They just die.
>
> You''ll have to be more specific than that.  "Die" could
mean they crash
> and the process goes away, or that the process stops responding to
> requests.  For the first kind you''ll need to turn on core dumps
and do
> a gdb inspect after.  With the second one you''ve gotta list out
the
> various gems you''re using and when you find a dead mongrel then
attach
> to it with strace or gdb to find out where it''s blocked.

Die meaning, dead...process is crashed and gone. How do I turn on core
dumps? This a mongrel switch?

> My question then is two-fold
> >
> > 1: Why are my mongrel processes dying? Where''s the log
information?
> > 2: Can these guys be automatically restarted somehow???
>
> Yes, monit works great, other folks like runit.  There''s a bunch
of
> tutorials for both.
>
>
Got a good starting point (besides google) for running either with a
mongrel_cluster?

Thanks for the quick response.

--------------------
seth at subimage interactive
-----
http://www.subimage.com
http://sublog.subimage.com
-----
http://www.getcashboard.com
http://dev.subimage.com/projects/substruct
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://rubyforge.org/pipermail/mongrel-users/attachments/20070403/e427a80e/attachment.html

You should able to do a search for monit and mongrel on this list or
google and get more than enough.

http://www.tildeslash.com/monit/


On 4/3/07, subimage interactive <subimage at gmail.com>
wrote:
>
>
> > > I recently moved one of my older Rails apps on a FreeBSD 5.5 box
to
> mongrel
> > > as well. Everything runs quickly and wonderfully - when
it''s running! My
> > > problem is that over time each one of my 5 mongrel processes die.
> There''s
> > > nothing in the mongrel.log telling me why they die. They just
die.
> >
> > You''ll have to be more specific than that.  "Die"
could mean they crash
> > and the process goes away, or that the process stops responding to
> > requests.  For the first kind you''ll need to turn on core
dumps and do
> > a gdb inspect after.  With the second one you''ve gotta list
out the
> > various gems you''re using and when you find a dead mongrel
then attach
> > to it with strace or gdb to find out where it''s blocked.
>
>
> Die meaning, dead...process is crashed and gone. How do I turn on core
> dumps? This a mongrel switch?
>
>
> > > My question then is two-fold
> > >
> > > 1: Why are my mongrel processes dying? Where''s the log
information?
> > > 2: Can these guys be automatically restarted somehow???
> >
> > Yes, monit works great, other folks like runit.  There''s a
bunch of
> > tutorials for both.
> >
> >
>
> Got a good starting point (besides google) for running either with a
> mongrel_cluster?
>
> Thanks for the quick response.
>
>
> --------------------
> seth at subimage interactive
> -----
> http://www.subimage.com
> http://sublog.subimage.com
> -----
> http://www.getcashboard.com
> http://dev.subimage.com/projects/substruct
> _______________________________________________
> Mongrel-users mailing list
> Mongrel-users at rubyforge.org
> http://rubyforge.org/mailman/listinfo/mongrel-users
>