5.2.15

Blog post on microservices, spark.

I've been playing lately with continous testing -> deployment of microservices via vagrant (test) and kubernetes (deployment).

I recently wrote up the basics of building simple microservices, because it seems there is a lot of confusion out there around how to really build a docker container properly.

My post is published on the red hat developer blog. 

If you're interested in Spark, docker, and micorservices in general, check it out !

http://developerblog.redhat.com/2015/01/20/microservice-principles-and-immutability-demonstrated-with-apache-spark-and-cassandra/


at

2 comments:

  1. UnknownMarch 3, 2015 at 5:26 PM

    thanks for this great post
    I wonder if there is some reason u r not chaining your run instructions
    chaining your run instructions has some good spin offs
    here is an example of what I mean
    https://registry.hub.docker.com/u/gkhachik/rhq-fedora.20/dockerfile/
    each run instruction spins up a intermediate container then commits as a layer in the image
    chained instructions => less layers => smaller images
    helps when you are pushing and pulling images and spinning up run-time containers
    https://docs.docker.com/articles/dockerfile_best-practices/

    ReplyDelete
  2. jayunit100May 26, 2015 at 8:38 AM

    hey nigel. just saw this. For STABLE images, your right --- but for dev work - unchained images are faster because changes you make can easily be layered on top of lower layers.

    ReplyDelete