Skip to content

HTTPS clone URL

Subversion checkout URL

You can clone with HTTPS or Subversion.

Download ZIP
Browse files

Merge pull request #21 from mdxp/master

adding ignite proposals
  • Loading branch information...
commit 86de43bec2504ed1ea3b9f7342a23dd78472564a 2 parents 1a7c171 + febe420
@endemics endemics authored
View
25 site/content/events/2012-mountainview/proposals/ignite - Don't pay for a dumb broker/index.txt
@@ -0,0 +1,25 @@
+---
+extension: html
+filter:
+- erb
+- markdown
+dirty: true
+proposal: true
+ignite: true
+author: Patricio Arvizu
+title: "Don't pay for a dumb broker."
+---
+**Abstract:**
+
+A lot of Continuous Integration tools today (both free and licensed) offer functionality out of the box or through plugins to do a variety of things. They all have pros and cons and are a great tool for startups and to quickly get on the Continous Delivery train.
+
+But consider the case of a company with multiple products. More than likely the build plans for both will be identical. So why would you need to have two separate projects? To further complicate matters, as your build/release needs change, you will have to maintain and update each plan separately. It becomes exponentially worse as the number of products increases.
+
+A better solution? Encapsulate the DSL for your build/release process in a single project, including all your build scripts. With this, two things can be achieved:
+
+1. The build system can be versioned and optionally released as an artifact. Either way, it becomes the first dependency for any product.
+2. More importantly, your CI servers only need to be ""task brokers"" that invoke these scripts with no knowledge about the build process. Do you really want to pay for a "dumb" task broker?
+
+As a result of this, the build system artifact becomes the central repository of the build/release knowledge and saves you from update headaches in the future.
+
+**Proposer:** Patricio Arvizu, Build Automation Engineer for TheLadders.com
View
16 site/content/events/2012-mountainview/proposals/ignite - EverybodyOps/index.txt
@@ -0,0 +1,16 @@
+---
+extension: html
+filter:
+- erb
+- markdown
+dirty: true
+proposal: true
+ignite: true
+author: Stefan Zier
+title: EverybodyOps
+---
+**Abstract:**
+
+Why and how all our developers carry the pager and deploy the system. We've practiced this here at Sumo Logic for the past year or so, with pretty good success, and I'd like to share some of the choices we made and what effect they've had. Somewhat similar to the NoOps practiced at Netflix, but at a much smaller scale.
+
+**Proposer:** Stefan Zier
View
20 site/content/events/2012-mountainview/proposals/ignite - Monitor All the Things/index.txt
@@ -0,0 +1,20 @@
+---
+extension: html
+filter:
+- erb
+- markdown
+dirty: true
+proposal: true
+ignite: true
+author: Jos Boumans
+title: Monitor All the Things
+---
+**Abstract:**
+
+At Krux, we run at a pretty high traffic load from multiple sources (we provide infrastructure for big publishers), so we live and die by our monitoring and graphing.
+
+We use some clever tricks to measure everything we can get our hands on and are pushing about 300k metrics per seconds.
+
+I'd like to take 5 minutes and point folks at some of the tricks they could use to make their lives easier.
+
+**Proposer:** Jos Boumans
Please sign in to comment.
Something went wrong with that request. Please try again.