Lemur is a tool to launch hadoop jobs locally or on EMR, based on a
configuration file referred to as a jobdef. The jobdef file describes your
EMR cluster, local environment, pre- and post-actions (aka hooks) and zero
or more steps (jobs). Lemur reads your jobdef, at the end of your jobdef
you execute (fire! ...) to make things happen. Lemur is implemented as an
internal DSL.
What Lemur is Not: Lemur is NOT a replacement for the Ruby
elastic-mapreduce cli. There is some overlap, but I did not attempt to
reproduce functionality (like --list, --terminate, etc) that works
perfectly well in that tool. Also, Lemur is not a job scheduler (a la
Oozie, Quartz or Azkaban). You might decide to use one of those tools to
trigger Lemur which knows how to run your job.
A simple jobdef:
(add-validators
(val-opts :required :numeric :num-days))
(add-hooks
(when-local-test)
(diff-test-data ["RESULTS" "results"]))
(defcluster marcs-cluster
:num-instances 1
:master-instance-type "m1.large"
:my-root "/Users/mlimotte/projects/lemur/tmp"
:upload ["./marcs-input.txt" :to "${data-uri}/input.txt"]
:test-uri "${my-root}/work"
:keypair "my-keypair"
:jar-src-path "${my-root}/lemur-test-0.0.1-SNAPSHOT-standalone.jar")
(defstep marcs-step
:main-class "lemur_test.marcs"
:args.days #(:num-days %)
:args.data-uri true)
(fire! marcs-cluster marcs-step)
This tool was developed at The Climate Corporation (TCC). TCC has chosen
to release this project with an Apache 2.0 license.
Blog post:
http://entxtech.blogspot.com/2012/05/lemur-declarative-launching-of-hadoop.html
Github Project: https://github.com/TheClimateCorporation/lemur
Marc Limotte
--
You received this message because you are subscribed to the Google
Groups "Clojure" group.
To post to this group, send email to clojure@googlegroups.com
Note that posts from new members are moderated - please be patient with your
first post.
To unsubscribe from this group, send email to
clojure+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/clojure?hl=en
