DevOps and Infor’s Journey to the Cloud

 (also posted on devops.com)

Infor, a leading provider of micro-vertical application suites, has introduced DevOps as part of an initiative to migrate its enterprise application suites to a cloud platform. The significance of this should not be lost.  Infor is not a young, Web 2.0 unicorn that was born with DevOps DNA.   Infor is a 14 year old privately held software supplier with over 70,000 customers across multiple micro-vertical markets.  Infor’s industry-specific applications are mostly deployed and managed on-premise.  The move to a cloud based AWS platform is as much of a paradigm shift for Infor as it is for its diverse customer base.

Infor’s leadership team launched the company’s annual user conference last week with keynote details about Infor XI’s cloud strategy and partnership with Amazon Web Services.   CEO Charles Phillips explained the reasoning behind the move while COO Pam Murphy recounted the technical efforts behind the scenes.   Both acknowledged the necessity of creating a function that would be dedicated to cloud operations. Standing in front of thousands of customers, Ms. Murphy described Infor Labs and how its DevOps practices would ensure that the right operating procedures were inserted into the development lifecycle.  According to Ms. Murphy, DevOps would bring “fully automated hands-free cloud operations that improves efficiency, effectiveness, consistency and security while reducing manpower requirements. “

Infor’s recognition of DevOps as a critical business success factor for its next generation of industry suites is impressive.  Brian Rose, Senior Vice President of Infor Labs and a key member of Infor’s management team for the past nine years provided some insight.  “The motivation behind developing and growing the DevOps team was driven by our ‘Cloud First’ strategy.  Over the last 18 months, Infor underwent a significant transformation to deliver industry solutions in the cloud.   This required development teams to gain a new understanding of the technical operations so that our products operate optimally and cost effectively in the cloud.”

“Our DevOps team worked closely with the development and cloud operations teams to develop standards for the cloud.  Further, they were responsible for setting and maintaining the cloud standards including the development of architectural components  that allowed for the automation of provisioning, patch processing, upgrades, backups, restores, monitoring and the development of a portal to provide a self-service option for these functions,” Rose explains.

Infor Labs has already seen results from DevOps. “The benefits that we have experienced include the reduction of the provisioning of an industry suite from weeks to minutes, decreased implementation time and costs for our customers, and solutions that are optimized from a technical ops perspective that allow our cloud infrastructure to scale so that we can support the thousands of customers that we have today in the cloud and the ability to scale to support the thousands more that we see moving to the cloud in the future, ” says Rose.

While DevOps is not intended solely for applications in the cloud, software providers such as Infor that publicly promote the value of their own DevOps culture increase proof of concept while serving as role models for their enterprise customers.    If the primary objective of DevOps is to improve speed to market and time to value, who better than a trusted commercial software provider to demonstrate real world insight and results.  

DevOps and the Theory of Constraints

The Theory of Constraints was first introduced in the business novel The Goal by Eliyahu Goldratt.   The main character of The Goal is Alex, a plant manager who is challenged to increase the output of his factory in a short period of time or face shutdown.  Alex embarks on a journey to understand systems thinking and relies on a mentoring relationship with an elusive Jonah to help him understand how to improve his production processes.

One of Alex’s key realizations occurs during a hike with his son’s scout troop.   Herbie, a less than athletic scout, is having a difficult time keeping up with the other boys.  Alex observes that it is actually not Herbie who has to keep pace with the troop, but rather the troop that has to keep pace with Herbie.   Poor Herbie is a constraint that affects the rest of the system. 

This example illustrates the basis of the Theory of Constraints – a well-respected model for systems management.  Every process is an end-to-end system that has one or more constraints affecting its outcome.  The process will therefore only meet the capacity of its weakest link.  If the constraint is improved or removed, the flow of the entire end to end system will also be improved.   The key is being able to analyze, identify and improve the constraints. 

Gene Kim’s best-selling novel, The Phoenix Project, builds on the same premise in an IT context.  Constraints are interwoven throughout the story and form the basis for improving flow in the First Way.  As more bottlenecks are overcome, the entire end to end flow between Dev and Ops is improved.

Bottlenecks may not always be apparent since work slows down but does not necessarily stop.   Good practices for process analysis and identification of constraints are starting to emerge.   The soon-to-be-published DevOps Cookbook by Gene Kim and colleagues will provide better guidance.

I believe that understanding and applying the Theory of Constraints will be a key practice to help transform a DevOps culture.  Analyzing the bottlenecks and constraints that exist in and through Dev, Ops and ITSM will help to eliminate waste, identify collaborative and opportunities and streamline the flow of work downstream and upstream.

A Culture of Trust

(also posted on devops.com)

Trust is by far the most critical element of a DevOps culture.  It is also the most difficult to achieve.  IT was not built for trust.  The silo structure that is common in most IT organizations was built for specialization and territorial ownership.  The silo walls are thick.

While DevOps may not be able to break down the silos, but it can encourage and groom a culture of trust.

Trust is difficult to define.  We know when we feel it, we know when we don't.   In 1993, Dr. Duane C. Tway, Jr. published a dissertation called "A Construct of Trust".  In it, Dr. Tway defines trust as

“The state of readiness for unguarded interaction with someone or something.”

Dr. Tway believes that trust is actually "constructed" from three basic elements:

• The capacity for trusting
• The perception of competence
• The perception of intentions

The capacity for trusting is how your total life experiences have developed your current capacity and willingness to risk trusting others.

The perception of competence is your perception of your ability and the ability of others with whom you work to perform competently at whatever is needed in the current situation.

The perception of intentions is your perception that the actions, words, direction, mission and decisions are motivated by mutually-serving rather than self-serving motives.

While trust is individual, a culture of trust is organizational.   Both are earned.  DevOps is a great opportunity to look at Dr. Tway's construct of trust and identify small but meaningful opportunities to steadily increase the trust between Dev and Ops including

• Improved and honest communications
• Increased personal interaction
• Honored commitments
• More listening than talking
• Frequent knowledge sharing
• Welcomed input and feedback
• Admitted mistakes
• Intolerance of blame
• Mutual respect
• A celebration of mutual achievements

What else can you do to create a “state of readiness for unguarded interaction with someone or something” in your organization?  How can you increase a mutual capacity for trusting, perception of competence and perception of intentions?   Perhaps you can start with an honest dialog with and between your Dev and Ops teams.  The answers may be surprising and insightful.

Service Management Fusion Conference 2014

As we ring in September, we also ring in the start of the fall conference season.

It's hard to believe but the itSMF/HDI Fusion 2014 conference is next month!   

As always, members of the ITSM crew will be among the many experienced service management professionals sharing insight, knowledge and experiences at the event.

Donna Knapp will be co-presenting with Anthony Orr at a Breakfast Briefing: How to Pass Your MALC Exam on Monday, October 20th at 7:30 am

I will be leading Session 606: Practical Problem Management: The Courage to Ask Why on Tuesday, October 21 at 2:45 PM.

I will also be participating in Session 704: Expert Focus on Practical DevOps with Brandon Gillis of Bank of NY Mellon on Wednesday, October 22 at 9:00 am.

You can register for Fusion14 at https://www.servicemanagementfusion.com/purchase.

Hope to see you there.

Join me at the DevOps IT Culture Cocktail Party session at Fruition Partners' FruDevCon

I am very flattered to have been asked to present at Fruition Partners’ fruDevCon14 conference in Chicago, IL October  5-9.  In it’s second year, fruDevCon offers the unique perspective of uniting development and process on a ServiceNow platform.   Attendees will not only learn more about “what” and “how” to develop within ServiceNow, they also will get insight into “why” they are doing what they do and how good process contributes to overall business success.   

  

I am particularly excited about this approach since the unity of development and process lies at the heart of the emerging DevOps movement.  DevOps does not rely solely on automating tasks – it is as a much a cultural initiative as it is a technical opportunity. Culture is nurtured from the top down and bottom up.  It starts with changing the way individuals think and behave.  

Culture is actually the focus of my presentation, “The IT Culture Cocktail Party” which will be part of the fruDevCon opening keynote session.  During our time together, we will explore the basics of DevOps and discuss how the integration of service management processes with Agile and Lean concepts can foster a culture of better collaboration and faster deployments between Dev and Ops.  We will also have a little fun mixing up a DevOps Culture Cocktail with best practice ingredients from multiple frameworks.  Since DevOps touches everyone in IT, it is my hope that the message of cultural unity will resonate equally with developers, operators and executives.

If your organization is utilizing a ServiceNow platform, I highly encourage you to attend fruDevCon.    In fact, if you register quickly, there is a 20% discount using promo code ANRBJ3.    I hope to meet you there.

Me and My Kanban

(also posted on www.devops.com)

There has been a trending interest in Kanban Boards, a tool that helps teams visualize their workflows, limit work in progress and get more “done”.   Kanban Boards are mostly associated with Agile Scrum teams as a collaborative means of seeing the burndown of a sprint.   In fact, Scrum and Kanban are considered to be so interdependent that a new framework called Scrumban is starting to emerge.   However, Kanban is not limited to Agile.  A Kanban board can be very useful for any project or scope of work that would benefit from visualization. 

The premise behind the Kanban is almost deceptively simple.  A board is built with three columns:  “To Do”, “In Progress” and “Done”.   At the start of each sprint, all of the work that needs to be completed is captured on individual sticky notes and placed in the “To Do” column.  When the task or work is actually being undertaken, the note moves into the “In Progress” column.  Upon completion, it is moved to the “Done” column.   If the team has successfully completed the work that was planned for a sprint, all of the cards that began in the “To Do” column will end in the “Done” column.  Any other result opens the opportunity for questions about impediments, team velocity (the ability to absorb work) and sprint planning.

You will often see team kanban boards built on whiteboards and peppered with multi-colored sticky notes.  There is also a great online tool (www.leankit.com) that offers a virtual team Kanban board with many of the same characteristics (including virtual sticky notes.  If the three columns are too basic, you can certainly add more columns because of the nature of the project.   The only caution is that it is easy to inadvertently make deceptively simple into overly complex.  Beware!

I have recently started to use a personal Kanban board to manage my own workload.  While not perfect, I have seen a significant improvement in my ability to get things done.   It also helps to keep me organized and focused.   I like moving the sticky notes around.

To begin, I created a backlog column with a colored note for each of my outstanding tasks.  The list is never-ending so notes are added to the backlog just about every day.  At the start of each week, I re-prioritize the backlog to ensure that deadlines and other considerations are addressed.  I then take a reasonable chunk of tasks off the top and add each note to the “To Do” column.  These are my goals for the week.  When I start to work on a particular task, I move its note into the “In Progress” column.  Upon completion, the note is moved to “Done”.   At the end of the week, I celebrate what I have accomplished and analyze what was left behind.   

If a note stays in the “In Progress” column for too long or moves back to “To Do”, I have to question why.  Did I overestimate my velocity or time it would take to complete a task?  Have I been faced with extensive interruptions or unplanned work?  Was I waiting on someone else?   The visualization of the Kanban board helps me understand my own patterns of work and recognize how I contribute to or impact the results of my team.  

I highly recommend that everyone to set up a personal Kanban, whether for work tasks or home tasks.  You can use the whiteboard in your office or set up a free basic online board through LeanKit  (www.leankit.com).   The results were be very interesting.

The DevOps Diet (Getting Lean)

(Also posted on devops.com)

IT has gotten fat – fat silos, fat processes, fat procedures.  Adopting DevOps is a great opportunity to look at some of our unhealthy habits and identify where we can eliminate unproductive waste from our process diets.  DevOps will ultimately teach IT how to exercise more with less effort, trim the fat and get lean.

According to lean principles, the seven main areas of waste are:

• Defects –  variations from requirements that result in interruptions and re-work
• Overproduction – delivering something more or before it is required
• Inventory  –  carrying excess raw materials, work in progress (WIP) or finished goods
• Over-processing – doing more work than is required
• Motion – moving people or equipment more than is required
• Transportation – moving products from one location to another
• Waiting – doing nothing or moving slowly while waiting on a previous step

What are the greatest sources of waste in your IT organization?  Are defects being regularly passed downstream? Are ITSM processes overproducing through complexity and bureaucracy?  Are Agile teams  dealing with excessive work in progress and less finished product?  Are unresolved bottlenecks resulting in frequent delays while waiting for something to get done or someone to be available?  Could automation reduce "motion sickness" and make transportation nonstop? These questions will hopefully facilitate dialog between Dev and Ops about analyzing and streamlining existing workflows.

While waste is fattening, it is not necessarily ineffective.  Projects and other tasks are still getting done and therefore some areas of waste may be difficult to recognize.  To become leaner, start simple by identifying and eliminating one unhealthy habit from your IT diet.   Engage stakeholders from Dev, Ops, customers and suppliers (if appropriate).  Discuss ways to improve the work with healthier habits.  Agree on what to do next or what not to do anymore.

Every diet should be supported by recipes for success.  The upcoming DevOps Cookbook by Gene Kim, et al. will provide more tangible guidance and recipes for healthy IT habits.  More on that in future blogs.

Most of all,  identifying and eliminating waste through Lean DevOps is a great opportunity for collaboration.  After all,  it's always better to diet with a friend.