Queue

Posts, talks, or essays planned. No particular order.

Google Doc Backlog

Why to use python virtualenvs

My journey into immutable infra

Building with ansible and packer

• Deploying on DigitalOcean
• Managing changes
• Drawbacks

Oh god kubernetes!

• This could be a book. Or a therapy session.

Building a storage API for Legobot

• Motivation: needed state
• Legobot philosophy: universal APIs
• Design problems: REST and CRUD interfaces
• Problems: Masking unique features of certain storage backends

Zero to consul and consul-template with Legobot

Bang commands?

“Ops for hackers”

• C&C clusters
• Bitcoin 101
• Opsec for bitcoin
• Mixers and tumblers
• Payment systems
• Hot and cold wallets
• Deanonymizing users
• Tracking tools

Building a quick-deploy hackbox with Packer and CoreOS

• superseded by axiom

Hackathon Series

~Why bother~

• What can you actually, meaningfully accomplish in 24 hours?

Moving the Needle

• A practical look at what “moving the needle” looks like and means
  • Hackathons build communities, not products

The sprint and the marathon

• A look at the role of the hackathon in the lifecycle of a product
  • Products born from hackathons
    • Many die from lack of investment (time and money)
    • Many die from lack of leadership
• Products augmented by hackathons
  • Many are not ready for the sprint that hackathons bring
  • Hackathons often create more problems than they solve
  • It takes time to get contributors used to your products

Building communities

• As a followup to “moving the needle”, a look at what goes into a community
  • Communities happen on purpose
  • Communities must be welcoming and inclusive
  • Communities must offer value to participants
  • Points (ii) and (iii) can override one another if strong enough. this is not desirable.

You should not host a hackathon

• Why and why not to host a hackathon

Other Essays

Update to Hacker Ethos (2020)

DevOps Archaeology

• What to do when you inherit an undocumented infrastructure

Facebook, dead trees, and ephemerality

Trees evolved before organisms who specialized in breaking them down, so for millions of years, we had piles of dead trees everywhere. They were recycled and cleaned up through massive conflagrations which wiped out whole areas rather than building a sustainable ecosystem and returning nutrients to the soil.

In the same way, our development and adoption of social media outpaced our notion of ephemerality. So much content we have created will be around forever because we simply didn’t consider the consequences of systems that didn’t clean up or age out old content. Do your facebook statuses about going to the beach or eating a sandwich need to live forever? Do you want those idiotic things you said as a teenager lingering for the rest of that platform’s life, waiting to surprise you decades later?

Ephemerality is quickly becoming a core feature of products today, and some existing products are adopting the concept. This essay explores the need for ephemerality as well as the benefits and dangers of a history that cleans up after itself. This talk is about being forgotten, for better or worse.

The Best Year I Ever Regretted: My intro to eng management

The “red wedding” at COTA.

The Dwight Schrute Guide to Threat Models

virtually all of Dwight Schrute‘s behavior can be explained through poor threat modeling. he is zealous about security but a poor threat modeler, so he ends up over fitting for some imaginary problem and misses the obvious one that ends up literally smacking him in the face (Jim, over and over).

Umwelt and DevOps

Umwelt - (in ethology) the world as it is experienced by a particular organism.

As applied to organizations, the idea that different specialists working on the same project or even team can have radically different experiences of the world based on their specialties, past experience, and current role

Quick example: a DevOps engineer, web dev, and a data scientist all working on the same project will have very different views of the same stack. To the ops engineer, she will have a large, detailed, and nuanced view of the infrastructure that supports the apps, the deployment models, the failure modes, and the data is just “data” and the web stack are just a handful of processes running on her infrastructure.

Staged in git

Communicating your limits, part 1: technical reality

Communicating your limits, part 2: personal limitations

DevOps in a day

If I only had one working day with someone to pass on the most important principles of DevOps, what would I say?

First Wave DevOps

When did it stop becoming a culture and start becoming “just a sysadmin”?

Defenders Paradox: SecOps edition

Visualizing chat data with Prometheus and Grafana

The Whiteboard Ideas

Mostly problems I want to solve for myself

Effective onboarding through automation and culture

Dependency injection for Flux and Terraform

CIS benchmarking TF and EKS

IAM management via Service Control Policies

Software Supply Chain concerns, gaps, tooling landscape

Productive chatops

• What makes chatops successful and adopted in an organization? Should you even pursue it?
• What is the difference between “chatops” and having some push notifications in a channel?