LISA18
| title | author | length |
|---|---|---|
| The Beginning, Present, and Future of Sysadmins | Tameika Reed | 39m |
Argues that classic sysadmin skills remain valuable and portable while tooling shifts: virtualization (KVM, VMware) and cloud migrations use Terraform… | ||
| Introducing Reliability Toolkit: Easy-to-Use Monitoring and Alerting | Janna Brummel, Robin van Zijll | 37m |
A team of eight SREs supports 1,700 DevOps engineers by shipping a preconfigured monitoring stack called the Reliability Toolkit that uses Prometheus… | ||
| Incident Management at Netflix Velocity | Dave Hahn | 55m |
Netflix reports that injecting failures with Chaos Monkey and a new Latency Monkey breaks assumptions: cranking latency to 500 ms across 100% of calls… | ||
| SLO Burn—Reducing Alert Fatigue and Maintenance Cost in Systems of Any Size | Jamie Wilkinson | 43m |
Argues that paging on user-facing symptoms and SLO error budgets cuts alert fatigue, shows a Prometheus implementation that measures SLIs at the load balancer… | ||
| Serverless Data Processing and Machine Learning | Sunil Mallya | 32m |
A MapReduce-style pipeline uses Amazon S3 as shared storage and AWS Lambda for mappers, reducers and a coordinator, processing 420 GB (3.5 billion lines, 4,200… | ||
| Designing for Failure: How to Manage Thousands of Hosts Through Automation | Brandon Bercovich | 28m |
Uber runs thousands of hosts with Apache Mesos and Aurora and builds a Cluster Lifecycle Manager to automate safe upgrades, drainage and remediation. CLM… | ||
| Familiar Smells I’ve Detected in Your Systems Engineering Organization and How to Fix Them | Dave Mangot | 33m |
Groups get paged 90 times in six hours because alerts are noisy, runbooks are untested Word docs and short‑term manual fixes (like outsourced ops deleting… | ||
| Code-Yellow: Helping Operations Top-Heavy Teams the Smart Way | Michael Kehoe, Todd Palino | 29m |
Code Yellow is a five-part rescue process with a clear problem statement, concrete exit criteria, extra resources and project management, prioritized… | ||
| How Bad Is Your Toil?: Measuring the Human Impact of Process | Kurt Andersen | 28m |
LinkedIn measures toil by monthly 3‑minute surveys that split time into categories and finds ~37% reactive work, 21% overhead and 42% engineering time against… | ||
| How to Be Your Security Team's Best Friend | Emily Cole | 27m |
Summarizes concrete steps: adopt the CIS Critical Security Controls; maintain searchable inventories of instances, containers and installed packages so alerts… | ||
| Isolation without Containers | Tyler McMullen | 29m |
Claims a lightweight isolation layer that avoids processes or containers by compiling WebAssembly to native AOT code and running thousands of sandboxes per… | ||
| Pass the Torch Without Dropping the Ball: Lessons in Community Management | Rich Bowen | 29m |
A year‑long handover produces a 30‑page transition document that spells out daily/weekly/monthly tasks, a cron‑based daily to‑do list, the cast of community… | ||
| Containers and Security on Planet X | Michael Jennings | 30m |
HPC needs container runtimes that preserve bare‑metal performance and low jitter, not general cloud defaults. Charliecloud uses unprivileged user namespaces so… | ||
| Getting the Most Out of Your Mesos | David Morrison | 27m |
Yelp runs Apache Mesos in production since 2014, hosting about 400 services and ~600 batch jobs and running 100k unit/integration tests hundreds of times a day… | ||
| SRE (and DevOps) at a Startup | Craig Sebenik | 26m |
Split is a feature‑flagging SaaS at Series B with a few dozen engineers, and the claim is that startups must choose between developers wearing the SRE hat or… | ||
| Cross Atlantic: Scaling Instagram Infrastructure from US to Europe | Sherry Xiao | 22m |
Instagram partitions its stateful stores to avoid cross‑Atlantic quorum traffic: it uses partitioned Cassandra (replication reduced from eight total copies to… | ||
| Delete This: Decommissioning Servers at Scale | Anirudh Ra | 25m |
Facebook automates server decommissions with a migration handler framework plus a UI and ticket state machine that take retiring and replacement sets, spawn… | ||
| Anatomy of a Crime: Secure DevOps or Darknet Early Breach Detection | Sarah Lewis Cortes | 42m |
Walkthrough of the Roman Seleznev-era carding economy: 1.7 million stolen cards found on seizure, darknet bazaars selling packs of 10,000 cards for $10–$25… | ||
| Do the Right Thing: Building Software in an Age of Social Responsibility | Jeffrey Snover | 44m |
Software now drives most top companies and compute trends project compute power ≈ all human brains per $1k in ~40 years, so engineering choices shape society… | ||
| The History of Logging @ Facebook (Abridged) | KC Braunschweig | 42m |
Scribe evolves from a C++ thrift collector into a multi-language pipeline that scales to thousands of categories and peaks of hundreds of GB/s; Calligraphers… | ||
| Container Security | Daniel Walsh | 47m |
Containers share the host kernel so isolation sits between bare metal and VMs: claim backed by examples — only two KVM breakouts in ten years, but common… | ||
| How Houghton Mifflin Harcourt Went from Months to Minutes with Infrastructure Delivery | Robert Allen | 26m |
Claim: infrastructure delivery shrinks from months to minutes by building a Mesos-centered self-service platform. Implements Apache Mesos + Apache Aurora… | ||
| Kafka on Kubernetes: From Evaluation to Production at | Shrinand Javadekar | 30m |
Shrinand Javadekar, Intuit, Inc.
Abstract:
Kubernetes is fast becoming the platform of choice for running distributed, containerized applications in the… | ||
| Overcoming the Challenges of Centralizing Container and Kubernetes Operations | Oleg Chunikhin | 28m |
Centralizes Kubernetes operations by building an enterprise ops layer on open-source pieces—federated Prometheus for metrics, central ELK with per-cluster… | ||
| Automating Multi-service Deployments on Kubernetes | Michael Hrivnak | 32m |
Shows how a containerized Ansible Playbook Bundle (APB) functions as a meta‑container installer that runs to completion inside a secure sandbox namespace… | ||
| Debugging Linux Issues with eBPF | Ivan Babrou | 27m |
A Kafka cluster sees a 10x CPU increase after a Debian/kernel upgrade; perf and flamegraphs show spinlock contention and many timer callbacks. Using eBPF/BCC… | ||
| Debugging and Optimizing User Experience with Deep Visual Analytics | Sarvesh Nagpal, Toby Walker | 30m |
Clarity captures DOM layout, viewport and pointer events, batches them into compressed JSON, and runs ML on those features to detect anomalies and cluster… | ||
| What Breaks Our Systems: A Taxonomy of Black Swans | Laura Nolan | 30m |
Conference postmortems group outages into six actionable failure modes and show concrete defenses. Examples include an ext3 2 TB filesystem ceiling that halts… | ||
| Change Management for Humans | Tiffany Longworth | 27m |
Uses the ADKAR steps: get people aware of the real risk, build desire by showing the personal wins, delay detailed knowledge until after buy‑in, make the… | ||
| Operations Reform: Tom Sawyer-ing Your Way to Operational Excellence | Thomas A. Limoncelli | 31m |
A spreadsheet-based, 4-question gut check scales to hundreds of services and produces monthly live roll-ups so management sees which services—not people—need… | ||
| Managing OS Release Transitions at Netflix Scale | Edward Hunter | 27m |
Netflix migrates over 150k+ Ubuntu instances across multiple AWS accounts by building a controlled BaseAMI pipeline and a Bakery that produces immutable… | ||
| Solving All the Problems with systemd | Alvaro Leiva Geisse | 31m |
Demonstrates how systemd plus kernel features replace containers or init scripts by doing concrete things: using socket activation and SO_REUSEPORT to achieve… | ||
| Unikraft: Unikernels Made Easy | Simon Kuenzer | 35m |
Unikraft presents an automated unikernel build framework that produces Xen/KVM/container/bare-metal images from composable OS libraries using make + Kconfig… | ||
| How Our Security Requirements Turned Us into Accidental Chaos Engineers | Paul Carleton | 27m |
Built a Lifespan Management system that enforces a seven-day VM lifespan using AWS Auto Scaling Groups, a local Terminator graceful-shutdown agent, and a… | ||
| Securing a Security Company | Patrick Cable | 21m |
Describes concrete steps taken at Threat Stack to tighten production security: replacing long‑lived AWS keys with short‑lived credentials via Vault and a… | ||
| We Already Have Nice Things, Use Them! | Sabice Arkenvirr | 12m |
Consulting across many companies reveals pervasive custom ops tools that cause single-point-of-failure risk, undocumented code, missing tests and slower… | ||
| Five-sigma Network Events (and How to Find Them) | John O'Neil | 28m |
Code implements anomaly detectors in Python using HyperLogLog sketches, incremental mean/std and online linear regression to track per-IP cardinalities and… | ||
| Reducing Chaos in Production: Testing vs. Monitoring | Robert Treat | 30m |
Claim: testing can’t catch scale and user-driven failure modes, so pair targeted tests with instrumentation that surfaces business impact. Case study: a client… | ||
| Engineering an IP Portfolio | Matthew Horton | 46m |
Outlines a concrete engineer workflow: run quarterly invention mining meetings, file short invention disclosure forms (IDFs) when a technical hurdle is… | ||
| Operational Excellence in April Fools’ Pranks: Being Funny Is Serious Work! | Thomas A. Limoncelli | 31m |
Stack Overflow’s “stack egg” prank floods load balancers, causing a self‑inflicted DoS that is disabled in three minutes and fully restored within an hour… | ||
| Keeping the Balance: Load Balancing Demystified | Murali Suriar, Laura Nolan | 46m |
Explains concrete choices and tradeoffs: an example server handles five queries per second, DNS round‑robin adds a TTL delay (short TTLs multiply DNS load… | ||
| Apache Kafka and KSQL in Action: Let’s Build a Streaming Data Pipeline! | Robin Moffatt | 43m |
Live demo builds a full streaming pipeline that mirrors MySQL rows into Kafka, enriches review events with customer data, filters low-star reviews for platinum… | ||
| Mastering Near-Real-Time Telemetry and Big Data: Invaluable Superpowers for Ordinary SREs | Ivan Ivanov | 31m |
Netflix Open Connect runs a purpose-built CDN on thousands of sites, peers directly with ISPs for >95% of traffic and supports 137 million subscribers… | ||
| Datastore Axes: Choosing the Scalability Direction You Need | Nicolai Plum | 30m |
Argues datastores must pick which axis to optimize because hardware and protocols prevent scaling all ways at once, and names four orthogonal axes: read rate… | ||
| Make Your System Firmware Faster, More Flexible and Reliable with LinuxBoot | Andrea Barberio, David Hendricks | 35m |
LinuxBoot replaces proprietary firmware by embedding a Linux kernel (handed off from coreboot) plus u-root/systemboot userland to initialize devices, enable… | ||
| Modern HTTP Routing | Sandor Szücs | 26m |
Zalando runs more than 80 Kubernetes clusters and uses an HTTP router called skipper that implements predicate route matching plus filters that modify headers… | ||
| Taking Over & Managing Large Messy Systems | Steve Mushero | 29m |
A managed‑service team in China documents concrete tactics for taking over live messy systems serving 1M–300M users across 50–100 datacenters: emergency… | ||
| Reborn in IT | Lori Barfield | 29m |
Based on a Southern California nonprofit program, the session gives concrete steps to jump into or pivot within IT: build a master resume and make targeted… | ||
| Serverless Ops: What to Do / This is What We Do, When the Server Goes Away | Tom McLaughlin | 23m |
Describes replacing host work by wiring cloud services together: take JSON from one API, trigger AWS Lambda via API Gateway, store in DynamoDB, emit events to… | ||
| Chaos Engineering | Nora Jones | 38m |
Claims that chaos experiments increase confidence in system resilience by validating failure modes in production, not by counting fixes. Netflix builds and… | ||
| nothing matches | ||