"Operations"

You keep using that word, but I don't think it means what you think it means

Avishai Ish-Shalom (@nukemberg)

What's wrong with this picture?

1961 - Mercury-Redstone 3

Mercury-Atlas

Mercury-Atlas 6

The Designers & The Operators

Astronauts are not enough

  • Training
  • Mission planning
  • "Space vehicle CM"
  • Mission control 
  • Ground Ops
  • Systems engineering

And much more...

Of course, we don't fly to space

What is your org mission?

Hints:

Your mission is NOT to write code

Your mission is NOT to run servers

Your mission is NOT to store data

The "System" is

  • Servers
  • Code
  • Data
  • Clients

 

Individually they are meaningless

Operations: using components to achieve the mission

Alternate version

Extracting value from assets

Why do we need Operations Engineering?

(and operators)

The world is complex

Our systems are complex

We can't predict everything up front

  • No design is perfect
  • The inconceivable DOES happen
  • Systems need to adapt to a changing world

Duct tape & glue

Over time, Operational cost will trump all other costs

 

- Andrew Clay-Shafer

Short history of (modern) Operations Engineering

Taylorims

AKA "Scientific Management"

 

  • Gantt - scheduling
  • The Hawthorne studies
  • Erlang - Queueing Theory

WW2

  • The Allies are desparate to improve production and logistics
  • Allied governments recruit physicists, mathematicians and engineers to Operations Research groups
  • After the war these people go on to reform civilian manufacturing

The beast from the East

Taiichi Ohno and Edward Deming revolutionize manufacturing

  • Toyota Production System
  • Lean
  • JIT
  • Statistical Process Control
  • Total Quality Management
  • Six Sigma

And much more

The Computer age

  • The rise of Services organizations
  • ERP & IT systems
  • Computer simulations

Systems Engineering

With great power comes great complexity

What do Operations Engineers do?

Monitoring

  • Is the system working?
  • What do clients experience?
  • Where is the problem?

Business Continuity

  • Continued operation despite failure
  • Recovering from disasters

Lifecycle management

  • Scaling
  • Capacity planning
  • Maintain SLA/QoS
  • Add/remove/upgrade components
  • Adapt the service

Configuration & Asset Management

  • What do we own? where is it?
  • Parts need to be adapted to field conditions and other parts
  • Tune for a changing world
  • Multiple modes of operation

Blackboxes

  • Extract maximum value from blackboxes
  • Work around limitations
  • Configurations
  • Handle malfunctions
  • Build operational knowledge

In short, we have a pet

The birth and transformation of IT Ops

Corporate IT

Then came SaaS

  • Online services became the core business
  • Needed high SLA
  • Operational cost killed companies
  • Web scale

Operations is about running a business

  • Competitive edge
  • Must have business context
  • It ain't about servers

Questions?

Made with Slides.com