How ops can prepare for DevOps
The devops solution is more practical than you might think, according to IT ops leaders from CA Technologies customers.
Making DevOps work in the enterprise may sound technical, but the key to making it work is actually more practical than you might think.
A few weeks ago, CA Technologies convened a group of IT operations leaders from our enterprise customers and service providers and, not surprisingly, DevOps dominated a large portion of the agenda.
One area of great discussion between the group was the advice they would give to other operations leaders on the pre-requisites to successfully adopt or participate in DevOps.
The advice from these leaders to their peers was very practical. Often based on their own personal experiences with their enterprises, their advice was primarily focused on “people and process” aspects of the change. Devops tool chains and service practices fell much lower on their priority list.
Since DevOps is fundamentally about organization change management, one of the key pieces of advice regarding the people part of the equation is to make sure there is “executive buy-in.” Below are some other main points from the discussion:
Have a clear definition and vision on what DevOps really means for your business and the implication on the ops functions.
Evangelize the mindset among ops staff that they are a third party competing for the business and that they should act accordingly – ultimately the business is paying you for results.
Continue to bolster the skillsets on ops side to build credibility with dev.
Get ops people involved in the beginning of the application lifecycle and participating on Agile dev teams.
Develop empathy: if you don’t care about the other group(s), the silos or walls will never drop.
Listen to counterparts in dev or, better yet, get dev and ops to walk in each other’s shoes to better understand issues and challenges the other is facing.
Develop shared priorities, without jeopardizing specialization. Ask yourself, “What is the primary goal that everyone shares and can align around?”
Be willing break everything and rebuild – but with customer at the center.
Stop perpetuating the “hero” culture with ops.
Stay current of dev tools and languages and why dev is using a certain tool or platform.
On the process side of the equation, here’s what the IT operations leaders had to say:
Streamline your ops processes: understand risk levels and align processes accordingly.
Figure out what ops shouldn’t be doing: ask what are the things that make the business more competitive versus things the company could care less about and can be easily outsourced. Justify this decision based on business value.
Deploy technologies that enable more agile response: automate repeated tasks and rationalize tools.
Remove silos within ops: share metrics, tools, dashboards and reduce intra-ops finger pointing.
Shift to measuring customer experience of critical business services that align with business priorities.
Look at goal sharing and alignment: make global priorities everyone’s priority.
Link compensation to business relevant metrics and away from the availability of ops silos.
From my vantage point as a facilitator of the discussion, not only was the preceding practical advice helpful, it was also the process of discussing these and sharing examples with peers from other businesses.
If the advice from these ops leaders leaves you asking for more detail, you should reach out to your peers from other companies and see how they are implementing some of the advice above.