SlideShare a Scribd company logo
Xtreme Deployment
                         / develop in production; push to dev



Tuesday, October 2, 12
Who are we?


          Brian Clapper
          CTO of Circonus
          Whipping boy of the modern SaaS stack.

          Theo Schlossnagle
          CEO of Circonus
          Repeat denier of reality.




Tuesday, October 2, 12
How the world used to look (less xtreme)




          Architectures driven by user interaction:
          (e.g. web page load ➠ web page response)




Tuesday, October 2, 12
How the world used to look (less xtreme)


          isolated
          user A’s page load doesn’t effect user B.


          latent
          user A’s affect shows up “later” in user B’s view.



Tuesday, October 2, 12
How the world used to look (less xtreme)




          Problem reproducibility is achievable
          (even if sometimes painful)




Tuesday, October 2, 12
How the world looks now (more xtreme)




          Architectures driven by
          high volume data and users




Tuesday, October 2, 12
How the world looks now (more xtreme)




          every piece of data entering the architecture
          can effect a user’s subsequent experience.




Tuesday, October 2, 12
How the world looks now (more xtreme)




          sometimes
          the data must effect a user’s
          subsequent experience.




Tuesday, October 2, 12
How the world looks now (more xtreme)




          Social systems cause indirect impact
          via user action eroding isolation.




Tuesday, October 2, 12
How the world looks now (more xtreme)




          A more responsive web is eroding
          users’ acceptance of latent state.




Tuesday, October 2, 12
How the world looks now (more xtreme)




          Highly distributed systems
          (due to complexity and scalability) means
          no single world state




Tuesday, October 2, 12
How the world looks now (more xtreme)




          Highly distributed systems
          (due to complexity and scalability) means
          difficult repeatability of state and failures




Tuesday, October 2, 12
How the world looks now (more xtreme)




          Highly distributed systems
          (due to complexity and scalability) means
          treacherous debugability




Tuesday, October 2, 12
Despair all ye’ who debug here




          I studied distributed systems long enough
          to understand “the suck” you do not wish to
          have in your lives.




Tuesday, October 2, 12
Despair all ye’ who debug here




          Expedient debugging requires a quick
          realization of a precipitating state of failure.




Tuesday, October 2, 12
Despair all ye’ who debug here



          In distributed systems,
          post-mortem techniques are less useful
          because recreating possible global state is a
          mind-bending (or brain-damaging) exercise.




Tuesday, October 2, 12
Despair all ye’ who debug here



          Applying older techniques
          to small controlled distributed systems
          is still feasible,
          but complex, asynchronous architectures
          can be nearly impossible.




Tuesday, October 2, 12
Some context about us




          Circonus ingests telemetry data
          across an entire business




Tuesday, October 2, 12
Some context about us




      alerting • correlation • visualization • analytics




Tuesday, October 2, 12
How we see the world today




          High velocity non-user-generated data inflow.




Tuesday, October 2, 12
How we see the world today




          Users expect sub-second reflection of
          all input data and drawn-conclusions.




Tuesday, October 2, 12
How we see the world today




          Users expect extremely high availability
          (100%, you can only disappoint)




Tuesday, October 2, 12
How we see the world today



          I was surprised how
          impatient and demanding users are...

          until I used the tool...
          and felt exactly the same way.




Tuesday, October 2, 12
Circonus architecture

          Voltron




Tuesday, October 2, 12
Circonus architecture

          Voltron
             • distributed data collection (via brokers)




Tuesday, October 2, 12
Circonus architecture

          Voltron
             • distributed data collection (via brokers)
             • distributed data ingestion (via aggregators)




Tuesday, October 2, 12
Circonus architecture

          Voltron
             • distributed data collection (via brokers)
             • distributed data ingestion (via aggregators)
             • distributed data distribution! (via message queueing)




Tuesday, October 2, 12
Circonus architecture

          Voltron
             • distributed data collection (via brokers)
             • distributed data ingestion (via aggregators)
             • distributed data distribution! (via message queueing)
             • distributed data storage (via snowth)




Tuesday, October 2, 12
Circonus architecture

          Voltron
             • distributed data collection (via brokers)
             • distributed data ingestion (via aggregators)
             • distributed data distribution! (via message queueing)
             • distributed data storage (via snowth)
             • distributed streaming fault detection (via muppet love)




Tuesday, October 2, 12
Circonus architecture

          Voltron
             • distributed data collection (via brokers)
             • distributed data ingestion (via aggregators)
             • distributed data distribution! (via message queueing)
             • distributed data storage (via snowth)
             • distributed streaming fault detection (via muppet love)
             • ... distributed nightmare.



Tuesday, October 2, 12
Despair actualized




          I shall not memorialize this in slides.




Tuesday, October 2, 12
Despair actualized




          I shall only cry in front of you now.




Tuesday, October 2, 12
Hope (keys to the city)



      • Leveraging
             • stream mirroring and
             • idempotent behavior.




Tuesday, October 2, 12
Hope




          Allows for the recklessness that we all
          appreciate in a no-consequences
          development environment...




Tuesday, October 2, 12
Hope



          with
          production data,
          production volume, and
          production “personality.”




Tuesday, October 2, 12
Hope actualized




                         Development In Production

Tuesday, October 2, 12
Hope actualized




          Ex #1: Dual storage backends




Tuesday, October 2, 12
EXPLORATION




                         PostgreSQL
Tuesday, October 2, 12
EXPLORATION




                         Snowth
Tuesday, October 2, 12
EXPLORATION




Tuesday, October 2, 12
                         {
                         Safety & Happiness
EXPLORATION




                         Safety & Happiness
Tuesday, October 2, 12
Hope actualized




          Ex #2: Dual streaming analysis tools




Tuesday, October 2, 12
Hell




Tuesday, October 2, 12
Hell




Tuesday, October 2, 12
Hell




Tuesday, October 2, 12
Hell




Tuesday, October 2, 12
Hell




Tuesday, October 2, 12
Hell




Tuesday, October 2, 12
Hope actualized




          Ex #3: Dual messages queueing systems




Tuesday, October 2, 12
The Future




          This is the foundation of
          real responsive design
          going forward.




Tuesday, October 2, 12
Thank You




          Questions?




Tuesday, October 2, 12

More Related Content

PDF
Monitoring and observability
PDF
PDF
What's in a number?
PDF
Building Data Driven Products With Ruby - RubyConf 2012
PDF
London Shared Bicycles: Measuring Intervention Impact
PDF
The math behind big systems analysis.
PDF
Understanding Slowness
PDF
SRECon Coherent Performance
Monitoring and observability
What's in a number?
Building Data Driven Products With Ruby - RubyConf 2012
London Shared Bicycles: Measuring Intervention Impact
The math behind big systems analysis.
Understanding Slowness
SRECon Coherent Performance

Viewers also liked (11)

PDF
Omnios and unix
PDF
Adaptive availability
PPTX
Craftsmanship
PDF
Monitoring is easy, why are we so bad at it presentation
PDF
Project reality
PDF
OmniOS Motivation and Design ~ LISA 2012
PDF
It's all about telemetry
PDF
Monitoring and observability
PDF
A Coherent Discussion About Performance
PDF
Monitoring the #DevOps way
PDF
Operational Software Design
Omnios and unix
Adaptive availability
Craftsmanship
Monitoring is easy, why are we so bad at it presentation
Project reality
OmniOS Motivation and Design ~ LISA 2012
It's all about telemetry
Monitoring and observability
A Coherent Discussion About Performance
Monitoring the #DevOps way
Operational Software Design
Ad

Similar to Xtreme Deployment (20)

PDF
Apache Cassandra - A gentle introduction
PDF
Ruxcon Finding Needles in Haystacks (the size of countries)
PDF
Availability, The Cloud and Everything (version 2, Surge2010)
PDF
Municipal Government Meets NoSQL
PDF
Bio-IT for Core Facility Managers
PDF
Optimizing the Mobile Search Experience
PDF
Using Ampere for Super-Fast Data Caching Awesomeness
PDF
Peeling back your Network Layers with Security Onion
PDF
Twitter Storm
PDF
Big Data - architectural concerns for the new age
PDF
Super secure clouds
PPTX
Big Data Tutorial V4
PDF
Performance Schema in MySQL (Danil Zburivsky)
PDF
Batch Indexing & Near Real Time, keeping things fast
PDF
Kotlin: Incompetence * Motivation = Innovation?
PDF
there is no talent shortage - Velocity 2013
PDF
D3.js capita selecta
PDF
Silverchair Strategies
PDF
OSSCON: Big Search 4 Big Data
PDF
Flowcon - Mixing Lean UX & Agile Development
Apache Cassandra - A gentle introduction
Ruxcon Finding Needles in Haystacks (the size of countries)
Availability, The Cloud and Everything (version 2, Surge2010)
Municipal Government Meets NoSQL
Bio-IT for Core Facility Managers
Optimizing the Mobile Search Experience
Using Ampere for Super-Fast Data Caching Awesomeness
Peeling back your Network Layers with Security Onion
Twitter Storm
Big Data - architectural concerns for the new age
Super secure clouds
Big Data Tutorial V4
Performance Schema in MySQL (Danil Zburivsky)
Batch Indexing & Near Real Time, keeping things fast
Kotlin: Incompetence * Motivation = Innovation?
there is no talent shortage - Velocity 2013
D3.js capita selecta
Silverchair Strategies
OSSCON: Big Search 4 Big Data
Flowcon - Mixing Lean UX & Agile Development
Ad

More from Theo Schlossnagle (13)

PPTX
Adding Simplicity to Complexity
PPTX
Put Some SRE in Your Shipped Software
PPTX
Monitoring 101
PPTX
Distributed Systems - Like It Or Not
PDF
Applying SRE techniques to micro service design
PDF
Commandments of scale
PDF
Is this normal?
PDF
Social improvements in monitoring
PDF
Building Scalable Systems: an asynchronous approach
PDF
Webops dashboards
PDF
Web Operations Career
PDF
Http front-ends
PDF
Applying operations culture to everything
Adding Simplicity to Complexity
Put Some SRE in Your Shipped Software
Monitoring 101
Distributed Systems - Like It Or Not
Applying SRE techniques to micro service design
Commandments of scale
Is this normal?
Social improvements in monitoring
Building Scalable Systems: an asynchronous approach
Webops dashboards
Web Operations Career
Http front-ends
Applying operations culture to everything

Xtreme Deployment

  • 1. Xtreme Deployment / develop in production; push to dev Tuesday, October 2, 12
  • 2. Who are we? Brian Clapper CTO of Circonus Whipping boy of the modern SaaS stack. Theo Schlossnagle CEO of Circonus Repeat denier of reality. Tuesday, October 2, 12
  • 3. How the world used to look (less xtreme) Architectures driven by user interaction: (e.g. web page load ➠ web page response) Tuesday, October 2, 12
  • 4. How the world used to look (less xtreme) isolated user A’s page load doesn’t effect user B. latent user A’s affect shows up “later” in user B’s view. Tuesday, October 2, 12
  • 5. How the world used to look (less xtreme) Problem reproducibility is achievable (even if sometimes painful) Tuesday, October 2, 12
  • 6. How the world looks now (more xtreme) Architectures driven by high volume data and users Tuesday, October 2, 12
  • 7. How the world looks now (more xtreme) every piece of data entering the architecture can effect a user’s subsequent experience. Tuesday, October 2, 12
  • 8. How the world looks now (more xtreme) sometimes the data must effect a user’s subsequent experience. Tuesday, October 2, 12
  • 9. How the world looks now (more xtreme) Social systems cause indirect impact via user action eroding isolation. Tuesday, October 2, 12
  • 10. How the world looks now (more xtreme) A more responsive web is eroding users’ acceptance of latent state. Tuesday, October 2, 12
  • 11. How the world looks now (more xtreme) Highly distributed systems (due to complexity and scalability) means no single world state Tuesday, October 2, 12
  • 12. How the world looks now (more xtreme) Highly distributed systems (due to complexity and scalability) means difficult repeatability of state and failures Tuesday, October 2, 12
  • 13. How the world looks now (more xtreme) Highly distributed systems (due to complexity and scalability) means treacherous debugability Tuesday, October 2, 12
  • 14. Despair all ye’ who debug here I studied distributed systems long enough to understand “the suck” you do not wish to have in your lives. Tuesday, October 2, 12
  • 15. Despair all ye’ who debug here Expedient debugging requires a quick realization of a precipitating state of failure. Tuesday, October 2, 12
  • 16. Despair all ye’ who debug here In distributed systems, post-mortem techniques are less useful because recreating possible global state is a mind-bending (or brain-damaging) exercise. Tuesday, October 2, 12
  • 17. Despair all ye’ who debug here Applying older techniques to small controlled distributed systems is still feasible, but complex, asynchronous architectures can be nearly impossible. Tuesday, October 2, 12
  • 18. Some context about us Circonus ingests telemetry data across an entire business Tuesday, October 2, 12
  • 19. Some context about us alerting • correlation • visualization • analytics Tuesday, October 2, 12
  • 20. How we see the world today High velocity non-user-generated data inflow. Tuesday, October 2, 12
  • 21. How we see the world today Users expect sub-second reflection of all input data and drawn-conclusions. Tuesday, October 2, 12
  • 22. How we see the world today Users expect extremely high availability (100%, you can only disappoint) Tuesday, October 2, 12
  • 23. How we see the world today I was surprised how impatient and demanding users are... until I used the tool... and felt exactly the same way. Tuesday, October 2, 12
  • 24. Circonus architecture Voltron Tuesday, October 2, 12
  • 25. Circonus architecture Voltron • distributed data collection (via brokers) Tuesday, October 2, 12
  • 26. Circonus architecture Voltron • distributed data collection (via brokers) • distributed data ingestion (via aggregators) Tuesday, October 2, 12
  • 27. Circonus architecture Voltron • distributed data collection (via brokers) • distributed data ingestion (via aggregators) • distributed data distribution! (via message queueing) Tuesday, October 2, 12
  • 28. Circonus architecture Voltron • distributed data collection (via brokers) • distributed data ingestion (via aggregators) • distributed data distribution! (via message queueing) • distributed data storage (via snowth) Tuesday, October 2, 12
  • 29. Circonus architecture Voltron • distributed data collection (via brokers) • distributed data ingestion (via aggregators) • distributed data distribution! (via message queueing) • distributed data storage (via snowth) • distributed streaming fault detection (via muppet love) Tuesday, October 2, 12
  • 30. Circonus architecture Voltron • distributed data collection (via brokers) • distributed data ingestion (via aggregators) • distributed data distribution! (via message queueing) • distributed data storage (via snowth) • distributed streaming fault detection (via muppet love) • ... distributed nightmare. Tuesday, October 2, 12
  • 31. Despair actualized I shall not memorialize this in slides. Tuesday, October 2, 12
  • 32. Despair actualized I shall only cry in front of you now. Tuesday, October 2, 12
  • 33. Hope (keys to the city) • Leveraging • stream mirroring and • idempotent behavior. Tuesday, October 2, 12
  • 34. Hope Allows for the recklessness that we all appreciate in a no-consequences development environment... Tuesday, October 2, 12
  • 35. Hope with production data, production volume, and production “personality.” Tuesday, October 2, 12
  • 36. Hope actualized Development In Production Tuesday, October 2, 12
  • 37. Hope actualized Ex #1: Dual storage backends Tuesday, October 2, 12
  • 38. EXPLORATION PostgreSQL Tuesday, October 2, 12
  • 39. EXPLORATION Snowth Tuesday, October 2, 12
  • 40. EXPLORATION Tuesday, October 2, 12 { Safety & Happiness
  • 41. EXPLORATION Safety & Happiness Tuesday, October 2, 12
  • 42. Hope actualized Ex #2: Dual streaming analysis tools Tuesday, October 2, 12
  • 49. Hope actualized Ex #3: Dual messages queueing systems Tuesday, October 2, 12
  • 50. The Future This is the foundation of real responsive design going forward. Tuesday, October 2, 12
  • 51. Thank You Questions? Tuesday, October 2, 12