Rapid Adoption of Technologies and the Resulting Complexity
76% of tech industry CIOs are worried that IT complexity will make it impossible to manage performance effectively. Read more here.
Join the DZone community and get the full member experience.
Join For FreeAn interesting panel discussion between Steve Tack, S.V.P. of Product Management at Dynatrace, Mark Tomlinson, Performance Engineer at PayPal and Mark Kaplan, Director of IT at BARBRI.
Mark Kaplan began the discussion by pointing out that BARBRI's mission is to always deliver a better experience to students and being able to deploy that technology is his responsibility. It can be difficult to know what he is going to need on the infrastructure side to run the technology and maintain a great UX with added compute power. Dynatrace allows us to measure UX and helps him keep an eye on our needs.
Mark Tomlinson pointed out that things are getting fragmented, with increased complexity, diversity and autonomy as developers tear down the monolith. Virtual teams are more independent and more autonomous. The "Brent Model" doesn't exist today. No longer is there one person that knows all and can fix all, now you have the entire team to know everything. If you're just relying on one guy and he leaves you’re in trouble, it’s not efficient.
What About the Pressure to Innovate Faster?
Mark Tomlinson:
- PayPal went independent as a company and the pressure is to continuously innovate away from the monolith has grown ever since.
- We've created autonomous, durable teams to deliver value with the freedom to monitor and deliver KPIs.
- The new generation learns to iterate faster, so they don’t build another monolith. They are developing new principles guide feeding across the enterprise.
Mark Kaplan:
- I refer to this as meta-complexity:
- Teams are more diverse and distributed;
- Greater diversity in how they think;
- More services, containers, microservices;
- How can we simplify the complexity?
- Complexity aids in the destruction of the monolith
- The newer technology you bring in the more it replaces the monolith
- People learn to manage the new technology
- Too much time is spent resolving problems
- The “Brent Model” doesn’t work
- I'm focused on building a stronger team
- My "Brent" was with the company for 20 years
- He knew the processes, the code, and the enterprise.
- He was someone to rely on who can guide rest of team.
- When we lost that person, it left a huge hole.
- But, when he left the rest of the team stepped up and flourished
- The team is living up to their own expectations and exceeding them
- Working collaboratively to solve the problems
- Part 2 of the DevOps handbook optimize for speed
- Every member of the team needs to be able to see information horizontally across the pipeline
- Make connections Brent never did by changing the paradigm
What About the Cost of Trying to Iterate Faster?
Mark Kaplan:
- Even solutions can be problematic
- Does anyone else have this problem?
- Finding solutions is not that big a deal
- You must scope the solution so someone has a view of the total cost of operation
- Give people the opportunity to see the total cost of the solution
Mark Tomlinson
- Figuring out the TCO involves more than the cost of the solution and the maintenance
- New solutions can create problems that were not there
- Don’t "build it and they will come" - identify the best solution to the problem taking TCO into consideration
- Partner with the business and propose a solution to the problem
What About the Role of AI Today?
- 81% of CIOs feel AI will be critical to mastering increasing IT complexity
- 83% of CIOs either have or will deploy AI in the next 12 months
Mark Tomlinson:
- I can’t wait to automate myself out of a job
- Getting to the point of predicting
- AI will help take blobs of information and analyze and present in an easily digestible way
Mark Kaplan:
- We integrate AL/ML into predictive analytics and adjust courses based on the students’ need
Opinions expressed by DZone contributors are their own.
Comments