Mongrel users - Feb 2007 - Monit / mongel_cluster 0.2.2 / mongrel 1.0.1

Hi,

I''ve a few problems with my rails app at the minute, causing mongrels
in a cluster to die, while I debug I''ve setup monit to keep the site
running.

Problem is, whenever monit starts one of my mongrels via mongrel_rails
cluster::start --only 8000 --clean -C /url/to/yml/file I get the
following in my log/mongrel.log:

** Mongrel available at 127.0.0.1:8000
** Writing PID file to log/mongrel.8000.pid
Define INLINEDIR or HOME in your environment and try again

I also had to make changes to mongrel_cluster 0.2.2''s init.rb to give
it the full path to mongrel_rails, otherwise it failed to find it.

If I run the same command line myself, the pup comes up fine.

Anyone got a working example with the new mongrel_cluster functionality?

monitrc:

check process mongrel-8000 with pidfile /path/to/mongrel.8000.pid
        start program = "/usr/local/bin/ruby
/usr/local/bin/mongrel_rails cluster::start --only 8000 --clean -C
/www/path/to/config/mongrel_cluster.yml"

Thanks in advance,
Paul

Hi,

I''ve fixed the below (previous post) which was simply a case of
reading the monit manual (it nukes most of the environment settings
when it runs the script).

Now I''m in a position to debug my problem, which is (I think)
image_science crashing mongrel during processing of an uploaded file
with a message of: terminate called after throwing an instance of
''int''.

I''m finding it difficult to reproduce the crash reliably as uploading
the same file works sometimes and not others.  I''ve tried running
mongrel with -B, and using killall -USR1 but nothing out of the
ordinary is being reported.

I''ve also tried strace-ing the mongrel process and this snippet is the
closest I can get to something useful:

open("/tmp/rpupload6375.0", O_RDONLY)   = 12 <0.000113>
futex(0xb6179e2c, FUTEX_WAKE, 2147483647) = 0 <0.000088>
futex(0xb60a41a4, FUTEX_WAKE, 2147483647) = 0 <0.000081>
write(2, "terminate called after throwing an instance of \''",
48) = 48
<0.000137>
write(2, "int", 3)                      = 3 <0.000092>
write(2, "\''\n", 2)                     = 2 <0.000091>
rt_sigprocmask(SIG_UNBLOCK, [ABRT], NULL, 8) = 0 <0.000146>
tgkill(6375, 6375, SIGABRT)             = 0 <0.000083>
--- SIGABRT (Aborted) @ 0 (0) ---

Is it perhaps some sort of file locking issue?   Any ideas on how I
could go about better tracking down the problem?

Thanks,
- Paul

----------------
> I''ve a few problems with my rails app at the minute, causing
mongrels in a cluster to die, while I debug I''ve setup monit to keep
the site running.
> Problem is, whenever monit starts one of my mongrels via mongrel_rails
cluster::start --only 8000 --clean -C /url/to/yml/file I get the following in my
log/mongrel.log:
> ** Mongrel available at 127.0.0.1:8000
> ** Writing PID file to log/mongrel.8000.pid
> Define INLINEDIR or HOME in your environment and try again
> I also had to make changes to mongrel_cluster 0.2.2''s init.rb to
give it the full path to mongrel_rails, otherwise it failed to find it.
[snip]