Last month, I joined a group of software practitioners and other enthusiasts at the 2015 ALM Forum in Seattle, Washington. I was invited to speak about two topics that I’m passionate about: Continuous Delivery and monitoring.
Before I spoke, I got to listen – while the term “Application Lifecycle Management” has been around for a while, the concepts and patterns being discussed at the conference were well ahead of the curve. Speakers from organizations like Netflix, Google, Amazon, Nordstrom, Moz, HP, ActiveState, Etsy, CenturyLink, Accenture, ThoughtWorks, Rally Software, Riverbed, RedHat, and Gartner talked about topics ranging from Agile, Lean, DevOps, and Microservices to Hybrid Cloud, Continuous Delivery, and Continuous Deployment.
Based on the relative “maturity” of people I was interacting with, I was unsure that the idea of introducing production level monitoring solutions earlier in the process would be “new” or “interesting enough” for folks to show up.
In fact, Roy Rappaport (monitoring architect from @Netflix who also gave a fantastic presentation) was shocked when he said to me “Is this really something people are not aware of? I’m can’t believe people AREN’T doing this already!”
After my presentation, I asked the audience if they actually did any pre-production monitoring – and only one person raised their hand! which makes me think there is a real market need and desire by practitioners for this type of information, and that many have not yet taken the plunge for introducing Monitoring earlier in their Continuous Delivery pipeline.
Our latest #c9d9 piggybacked off of this topic, where we put the following questions to our panelists:
- What should you measure? Infrastructure? Performance? App usage? Business logic?
- Where should you measure it? Which environment? Why?
- And finally, how do you use the data you collect?
Watch the episode to find out how you can make Ops teams more comfortable with Continuous Delivery by taking advantage of the benefits of monitoring earlier.