Practical APM & DevOps By @SolarWinds | @DevOpsSummit [#DevOps]
DevOps is a development methodology based on continuous integration and continuous delivery
By: Gerardo A Dada
Nov. 19, 2014 08:00 PM
A Practical Approach to Application Performance & DevOps
Considered by many as the next step beyond Agile, DevOps has proven to be effective at accelerating development cycles, improving performance, reducing bugs and overall improving the innovation and velocity of development teams.
There are a couple ways to look at DevOps: first, DevOps is a development methodology based on continuous integration and continuous delivery supported by set of configuration management tools such as Chef, Puppet, Salt and Ansible. We can also think of DevOps as a simpler set of principles that guide development and deployment practices - automate everything, monitor everything and log everything. All with the goal to get visibility into how every change in a fast-paced iterative process impacts performance.
The challenge - aside from the complexity involved in implementing these tools, building scripts and morphing the entire development process - is that DevOps requires a culture change and a new set of skills. The key question then is: Where to start? How can teams start enjoying the benefits of DevOps without having to wait months or years for new skills, tools and processes to be in place and for a new culture to take root?
While this sounds simple, the reality is that more often than not, the different teams working on applications are organized in silos, with each team running their own dashboard and having their unique view into their particular portion of the application stack. The result is finger pointing and issues that hide in the space between silos and the interdependencies between components of the application stack.
A powerful first step towards DevOps is working toward providing a single version of truth for these teams - a common framework where all team members can understand what goes on in the application, database, OS, hypervisor, host server and storage systems. Such a system eliminates finger pointing by making it very clear where issues are, in a way there is no more ambiguity, just action.
How to Implement
But That's Not All
Second, many DevOps teams try to monitor everything. The result is gigabyte-plus logs that require advanced tools and lots of time to analyze. While the graphics produced can be very cool, their usefulness is limited to their ability to produce insights. A terabyte of log data is useless unless it can pinpoint what is wrong with the system.
In the end, development and operations teams can work better together when:
Even without the complexity of a full-blown configuration management system, having the right goals and effective performance tooling can get you one step closer to DevOps nirvana, which means faster applications and innovation.
Reader Feedback: Page 1 of 1
SOA World Latest Stories
Subscribe to the World's Most Powerful Newsletters
Subscribe to Our Rss Feeds & Get Your SYS-CON News Live!
SYS-CON Featured Whitepapers
Most Read This Week