Heh guys. Here's the agent script: https://gist.github.com/marianogg9/9072387, from https://github.com/newrelic-platform/newrelic_rackspace_load_balancers_plugin
And this got my attention: unless $:.include?(File.dirname(__FILE__) + '/../lib/') $: << File.dirname(__FILE__) + '/../lib' end Isn't it an env path definition? keep in mind you need to be pwd in /path/newrelic_rs_dfw/ to run bundle exec bin/newrelic_rs.rb Thanks. On 18 February 2014 08:56, Mariano González < [email protected]> wrote: > Hi Steve and Clint. Thanks for your comments. > > @Steve > > You're totally right, it's not necessary to write sudo -u root as it's the > same as writing sudo. My fault there. > But, the command to run the agent is the following: > > $ cd /path/newrelic_rs_dfw/ > $ bundle exec ./bin/newrelic_rs > > As bundle exec is a ruby gem located in /path/newrelic_rs_dfw/. So, to > this script to function it'd need to locate in /path/newrelic_rs_dfw/ first > and then run the bundle exec. Which it doesn't do if I write down those two > previous steps in a script. Neither it does if I specify an env variable or > source stanza. That's why I think this is an env issue. > > @Clint > > I'm trying to run this agent as an Upstart job, that's why I'm writing an > /etc/init/newrelic.conf script. Sorry if I missunderstood your comments if > it's not what you meant. > > Finally, there's no trace of newrelic logs in /var/log/ besides /newrelic/ > folder where OTHER newrelic agents does write their logs in this vm. > > The only logs I have are the ones written in /messages.log > > Thanks again! > > > On 17 February 2014 12:40, Clint Byrum <[email protected]> wrote: > >> Excerpts from Mariano González's message of 2014-02-17 06:43:16 -0800: >> > Hi guys. >> > I'm asking for help to write a script to monitor a newrelic agent. >> > Here's what I have now: https://gist.github.com/marianogg9/9002995 >> > As you may see, the service won't start and will respawn 10 times. I >> think >> > this is an env issue, but I tried with export and env stanzas without >> > success. >> > Any comments will be really appreciated. >> > Thanks! >> > >> >> If you are running it as root, why not have it as a system job from >> /etc/init/something.conf ? Then you can drop sudo. >> >> Also you should have logs in ~/.cache/upstart if it is a user job, or >> /var/log/upstart if it is a system job. Those should give a clue as to >> why your command is failing. >> >> -- >> upstart-devel mailing list >> [email protected] >> Modify settings or unsubscribe at: >> https://lists.ubuntu.com/mailman/listinfo/upstart-devel >> > > > > -- > Mariano Gabriel González > -- Mariano Gabriel González
-- upstart-devel mailing list [email protected] Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/upstart-devel
