Sorry I am getting back to it just now, I have been very busy.

I added the code snippet at the end of my deploy.rb file (as it is
loaded by the Capfile, right?) However, I cannot notice any change. It
still tells me the "command finished" and still, there is no mongrel
running on the server. I introduced a bug on purpose to make sure the
deploy.rb is read, and indeed it did give me an error message. I am
not sure how I can debug this problem.
My deploy.rb now looks like:
http://pastie.org/409283

I tried the following:

I logged into my server using the user capistrano uses
I ran the command manually
it only worked if I modified it to
  mongrel_rails start -C <configfile>           (I removed the
"cluster" and the --clean)
I also found out that it only works if I do not use sudo, because if I
do the path does not contain the directory of my gems. With "sudo su"
it has the appropriate path - where do I have to add that info for
sudo to have it? Anyway, I bypassed that problem by modifying my
deploy.rb to use the absolute path to access mongrel_rails. This
manually works even with sudo. However, Capistrano still won't do
anything that it has not done before.

Next step I tried was to install the gem mongrel_cluster, which indeed
let me now use the "cluster" and the "--clean". If I type
mongrel_rails cluster::start -C <configfile> --clean
It tells me:
starting port 8000
starting port 8001

even though my mongrel_cluster.yml says nothing about two ports. It
looks like this:
http://pastie.org/409309

Also, when I now do a "ps aux | grep mongrel" it does not find
anything but the grep itself. If I am very fast, I get to see a
1000     28034 37.7  0.9  24044 20656 ?        R    14:47   0:01 /usr/
bin/ruby1.8 /var/lib/gems/1.8/bin/mongrel_rails start -d -e production
-p 8000 -P /var/run/mongrel.8000.pid -l /var/log/mongrel/mongrel.
8000.log
1000     28038 38.0  0.9  24032 20632 ?        R    14:48   0:01 /usr/
bin/ruby1.8 /var/lib/gems/1.8/bin/mongrel_rails start -d -e production
-p 8001 -P /var/run/mongrel.8001.pid -l /var/log/mongrel/mongrel.
8001.log

But those processes vanish shortly after. If I check the log file
directory, there's not a single file in it.

I really don't now what else I could try. I understand that a cluster
is always more than one mongrel. But why it does not work and even
more: the start without cluster works manually, but not if I put it in
the deploy.rb, that I don't understand. Any ideas?

--~--~---------~--~----~------------~-------~--~----~
To unsubscribe from this group, send email to 
[email protected]
For more options, visit this group at http://groups.google.com/group/capistrano
-~----------~----~----~----~------~----~------~--~---

Reply via email to