reliability

A KRAZAM comedy sketch in which a character recites reassuring affirmations to site reliability engineers, satirizing on-call and outage culture.

Bryan Cantrill's Uptime 2017 talk on the engineering challenges of the stateful data path, including storage, firmware and hardware bugs. It uses the 'zebras vs. horses' diagnostic heuristic to explain why rare failures keep appearing.

Dave Plummer tours the IBM z16 mainframe from design through factory assembly, test lab, accelerators and water cooling. He explains why mainframes still exist and what makes them reliable.

A dotScale 2014 talk by Robert Kennedy, a former healthcare.gov SRE who later worked at Google, about the engineering effort to rescue the troubled site. It explains the technical and organizational problems behind the failed launch.

A Google Tech Talk by Raymond Blum on how Google backs up massive amounts of data such as Gmail and Picasa. It covers scaling challenges, failure types, and approaches that did not work, including tape.

InRangeTV's controlled mud test of an AKM, covering it in mud and testing whether it still fires, in a series that also tested the M1, M14 and AR-15.

InRangeTV subjects an AR-15 to a mud test in the same mud and location as a preceding AK-47 test. It shows how the rifle runs when covered in mud.

Richard Cook's short Velocity 2013 talk on resilience in complex adaptive systems, arguing systems operate near the edge of failure and that the surprise is how rarely they fail. He presents a model of how pressures push systems toward failure boundaries.

Bryan Cantrill's GOTO Chicago 2017 talk on debugging production systems under pressure, centered on the incident where an operator's mistyped command rebooted an entire Joyent datacenter. It covers human factors, automation and architecture.

Rick Houlihan's AWS re:Invent 2018 deep dive on advanced DynamoDB design patterns. It covers index overloading, GSI sharding, composite keys, graph processing and transactions.

Howard Abrams demonstrates, in an EmacsConf 2015 talk, how he uses org-mode's Babel and Tramp to configure and operate remote servers from a single document. Code blocks, their results and the explanatory text all live together.

Liz Rice live-codes a minimal container runtime in a few lines of Go at GOTO Amsterdam 2018. She explains namespaces, control groups and what a host knows about its containers.

Overclocker der8auer tours Hetzner's Falkenstein data center in Germany, showing custom racks, own mainboards, passively cooled servers and ATX-style cases on shelves.

Bryan Cantrill of Oxide Computer Company gives a Stanford EE380 talk on rethinking the server. He argues that hyperscaler-style rack-scale machines with open firmware, RISC-V and Rust should be available to everyone, not just the giants.

Bryan Cantrill of Joyent presents Manta, an internet-facing object store that lets you run compute directly on the data. The talk is peppered with ZFS history and Cantrill's trademark commentary on the industry.

Daniel Azuma's ElixirConf 2018 talk on whether Docker and OTP conflict. He shows they can coexist and uses Horde and DeltaCrdt, with a live tank-game demo, to keep state alive across container kills.

James Hamilton's Tuesday Night Live keynote at AWS re:Invent 2016, touring AWS infrastructure at scale including networking, power and custom silicon. It also features NASA JPL's Tom Soderstrom.

Dropbox engineer Kevin Modzelewski gives a Stanford colloquium on Dropbox's history and the technical problems it faced scaling, including detailed database schema decisions.

Bryan Cantrill's 2007 Google Tech Talk on DTrace, showing how dynamic tracing improves debugging of both development and live production systems. It includes a vivid war story about tracking down a problem on a Sun Ray server.

Liz Rice builds a container from scratch in about 100 lines of Go, showing that a container is not a lightweight VM but a set of Linux kernel features and flags.

Nickolas Means recounts the 1979 Three Mile Island accident minute by minute, from the control rod insertion to the operator's decision to switch off emergency cooling. He argues that finding systemic causes is more useful than blaming individuals.

Dr. Richard Cook, a physician and patient safety researcher, presents his ideas on how complex systems fail, at Velocity 2012. He argues for designing for resiliency rather than trying to exhaustively handle every failure mode.

Benn Jordan investigates infrasound and low-frequency noise from data centers, including xAI's Colossus and a crypto mine, and reports his own experiments on the effects. The video covers the health and community impacts on nearby residents.

The author of copyparty introduces the FOSS file server he has built over five years, demoing its features such as resumable uploads, WebDAV and deduplication. The video itself is shown running on a demo server.

Jon Moore's Strange Loop 2017 talk argues that rate limiting is the wrong tool for protecting services from noisy clients. He proposes concurrency-based capacity management and quotas instead.

Jérôme Petazzoni explains how Linux containers are built from kernel features like cgroups, namespaces and SELinux. He demonstrates putting the pieces together by hand to make a container and compares runtimes and approaches like jails and zones.

DHH's opening keynote at Rails World 2024 in Toronto, where Rails 8 beta shipped. It covers built-in authentication, Propshaft, Solid Cache, Solid Queue, Solid Cable, Kamal 2 and Thruster, pitching the One Person Framework that needs no PaaS.

Eric Brandwine's AWS re:Invent 2015 session on how Amazon VPC networking works. It covers the problems the design solved, why traditional approaches do not scale, and the mechanisms used for customer isolation and VPC endpoints.

A podcast episode of Arrested DevOps with Andrew Clay Shafer and Bryan Cantrill, featuring Cantrill's rant on integrity in companies and technology.

Bryan Cantrill's Velocity NYC 2015 talk argues for running containers directly on bare metal rather than layering them on virtual machines. He covers the history of containers and how SmartOS runs Linux workloads via syscall emulation.

A Business Insider investigation into the US boom in AI data centers, mapping where they are, who owns them, and how much power and water they consume. It looks at the costs borne by communities, with a focus on Virginia.

Kyle Drake of Neocities explains the bandwidth pricing pitfalls of the big cloud providers. He describes how to run high-volume, bandwidth-heavy infrastructure far more cheaply, including through co-location.

Jared Mauch tells how he got fed up with poor internet options in rural Michigan and built his own fiber network to his town. The talk was given at NLNOG Live in September 2020.

SuperfastMatt looks at how the Tesla Model S retracting door handles were redesigned over time, from many parts to a motor plus sensor, with a focus on engineering and reliability.

Eager Space examines the practicality of putting data centers in orbit, walking through three scenarios with cost calculations and the launch cost reductions that would be required.

A walkthrough of installing Synapse, Riot and Jitsi from scratch on Debian to run your own Matrix communication service. The description warns it is now quite outdated and links to updated material.

Mikey Dickerson's keynote at Velocity NY 2014, one year after the healthcare.gov launch failure. He recounts how a small team rescued the site and what the episode showed about government IT contracting.

Pixar staff tell the story of how Toy Story 2 was almost deleted from the studio's computers and was saved by a home computer backup.

A fasterthanlime video essay arguing that GitHub Actions is a poorly designed system, with a history of personal computing and CI, and a look at how it works behind the scenes. It is sponsored by Brilliant.

Kelsey Hightower introduces Kubernetes to sysadmins at PuppetConf 2016, with a demo. He explains the declarative model and how it changes the contract between infrastructure and applications.