SREcon22 Europe/Middle East/Africa
2022-10-25..2022-10-27
Amsterdam, Netherlands
who it's for
site reliability engineers, platform engineers, observability engineers, incident managers, engineering managers, cloud architects, security engineers
topics
service level indicators and objectives; runtime observability and telemetry; on-call practices and rotations; incident analysis and writeups; dynamic Terraform configuration; feature flag evaluation in Terraform; high throughput logging pipelines; dependency tracing with OpenTelemetry; disaster recovery and game days; kernel and Linux upgrades for SRE; software carbon intensity measurement; regulatory compliance for cloud resilience
| title | author | length |
|---|---|---|
| Knowledge and Power: A Sociotechnical Systems Discussion on... | Laura Maguire, Lorin Hochstein | 39m |
Series of four cross-disciplinary interviews claim that knowledge is partial, tacit and unevenly distributed, and that incidents are the best method to surface… | ||
| SRE as She Is Spoke | Andrew Clay Shafer | 40m |
Argues SRE is a platform role that sits between infrastructure and product software, not just a label; traces origin to 2006 Vogel and 2009 devops, cites… | ||
| Oncall: An Equal Opportunity Waste of Time | Dave O'Connor | 30m |
Argues that on-call is a waste of time because gatekeeping and hero culture make incidents more complex and preserve headcount. Recommends two exercises: a… | ||
| Financial Regulators Worldwide Are Getting the Legal Right... | Andrew Ellam | 21m |
New laws like DORA and the UK Financial Services and Markets Bill give regulators the legal power to demand evidence of uptime, inspect game day results and… | ||
| Statistics for Engineers | Heinrich Hartmann | 32m |
Walks through concrete tactics: use histograms and heatmaps to preserve distributional detail, avoid reporting average latency (mean is non-robust), prefer… | ||
| Measuring Reliability: What Got Us Here Won't Get Us There | Štěpán Davidovič | 29m |
Argues that SLI/SLO metrics are useful but incomplete: a company with 200 SLOs sees Jan 3, Feb 4, Mar 12 violations and naive interpretation hides a 95% CI of… | ||
| Crayon Drawing Is a Vital Engineering Skill | Murali Suriar | 18m |
Claim: quick, hand-drawn system overview diagrams produce shared understanding that speeds onboarding, debugging and design decisions. Examples: three compact… | ||
| Building Dynamic Configuration into Terraform | Isabelle Miller, Hosh Sadiq | 08m |
LaunchDarkly builds a feature flag evaluation provider for Terraform that wraps the Go SDK and exposes four data sources (Boolean/string/number/JSON) so… | ||
| Hunting for Risky Dependencies in the World of Microservices | Theo Klein | 17m |
Describes a practical method that uses OpenTelemetry to record the original RPC origin and caller chain, aggregates traces into a central data lake, and… | ||
| How We Implemented High Throughput Logging at Spotify | Lauren Muhlhauser | 14m |
Fluentd is CPU-bound on a single core and causes massive silent loss — Spotify estimates ~80% of logs dropped as the deployment grows to ~1,700 namespaces. The… | ||
| Engineering for Sustainability | Namrata Namrata | 18m |
Claim: cloud and software choices drive measurable carbon costs and simple fixes scale. Use AWS Customer Carbon Footprint or GCP Carbon Footprint API and… | ||
| SLOs, SREs, and GHGs | Bill Johnson | 14m |
Provides a concrete method to measure software carbon: Software Carbon Intensity = energy (kWh) × carbon intensity (gCO2/kWh) + embodied carbon, normalized by… | ||
| The Biases Confronting SREs | David Owczarek | 15m |
SRE originates at Google in 2004 and the talk argues that real practice splits into three buckets: on-call or production reliability, collaborations with… | ||
| Market Data: Applying SRE Techniques to Legacy Designs | George Brighton | 20m |
Market data platform handles up to 700 billion ticks/day with a 5 ms latency target; initial wins come from Black box probes using the internal Market Data… | ||
| Life after The Chocolate Factory | Murali Suriar, Emil Stolarsky | 20m |
Large-company SRE practices often fail to translate because the friction is in small operational details not exotic tech: onboarding breaks from missing AWS… | ||
| Is Our Team as Resilient as Our Systems? | Effie Mouzeli | 16m |
Conference observes teams hiring faster than they can onboard, leaving people in production without training and a few heroes doing the work of three… | ||
| What SRE Could Be: Systems Reliability Engineering | Laura Nolan | 49m |
Argument: SRE is systems thinking applied to software and sociotechnical systems, not just SLOs or pager duty. Demonstrated tools include hierarchical task… | ||
| Diamonds with Flaws: Examining the Pressures, Realities, and... | Alex Hidalgo | 21m |
Argues that SRE is overloaded with responsibilities—cites a 26-item list of tasks and the market full of CI/CD, observability, APM and Kubernetes… | ||
| How We Drained Every Backbone Router Simultaneously | Francois Richard | 34m |
On October 4, 2021 a single operator command that listed all data planes triggers a safety-check bug and drained all eight data planes in seconds, knocking out… | ||
| Break Free of the Template: Incident Writeups They Want to Read | Laura Nolan | 29m |
Argues that written incident reviews are the lasting learning artifact and shows concrete ways to make them readable: turn incidents into narratives (example… | ||
| Making the Impossible Impossible: Improving Reliability by... | Chris Sinjakli | 28m |
Argues SLOs nudge teams toward percentage fixes but some outages are avoidable by design: using state machines to forbid invalid payment transitions, relying… | ||
| Deep Dive: Azure Resource Manager Outage | Benjamin Pannell, Brendan Burns | 39m |
A legacy configuration file missing in production (present in tests via a transitive dependency) causes a reflection-based static initializer to throw a… | ||
| Commas Save Lives, or at Least LinkedIn | Todd Palino | 29m |
A single missing comma in JSON entry to an anti‑abuse ACL causes LinkedIn’s front door to drop to zero requests, Apache Traffic Server enters a crash loop… | ||
| Passing the Torch - Building a New Grad Program to Mentor... | Chris Stankaitis | 32m |
Company designs a two-year roadmap to junior engineer that converts near-zero SRE/devops experience into a hireable junior role, hires two new grads for the… | ||
| Going from 30 to 30 Million SLOs | Alex Palcuie | 23m |
Describes engineering work that scales SLOs from ~30 to roughly 30 million by moving from zone-level to per-customer SLO budgets, building a daily analytics… | ||
| Disaster Recovery Testing at Booking.com | Yoann Fouquet, Paola Martinucci | 30m |
Booking.com documents a program that starts with 80,000 servers and ~2,000 services and moves from a manual region evacuation that takes 4+ hours to seconds… | ||
| Slack's DNSSEC Rollout: Third Time's the Outage | Rafael Elvira | 27m |
A large SaaS enables DNSSEC across dozens of domains, then trips on three failed attempts to turn it on for slack.com: first aborted for an ISP outage, second… | ||
| Meatbag Systems: How Our Reliability Culture & Practice... | Andrew Howden, Salomé Santos | 24m |
Zalando had >3,000 registered apps and a messy on‑call landscape; a Feb 2019 bug blocks checkout for ~2,000 customers and exposes noisy alerts, so teams define… | ||
| Principled Identification of "Root Causes" Using Techniques... | Laura de Vesine | 19m |
Argues that reliable root causes are system vs environment vulnerabilities, not upstream triggers; applies safety engineering tools to classify controllable… | ||
| A Case Study in Chaos Testing: Uncovering Kernel Scaling Issues | Gary Liku | 24m |
Chaos testing replicates a sporadic production scaling failure and pins it to /proc read paths and thread traversal in a 3.10 kernel; experiments use internal… | ||
| A Better Way to Manage Command Line Tools: What We Learned... | Bo Hou | 23m |
Criteo builds a lightweight launcher called CDT command launcher that runs packaged CLI tools described by a manifest (JSON/YAML), provides auto-completion… | ||
| Honey, I Broke the Things: Debugging Gray Failures... | Radha Kumari | 13m |
Slack migrates HAProxy to Envoy and hits multiple concrete regressions: a connect_failure typo that Envoy silently ignores until Envoy 1.10, missing retries… | ||
| The Repeat Incident Fallacy: What Jurassic Park Can Teach Us... | Emily Ruppe | 10m |
Claims that incidents can be made never to recur are false: complex socio-technical systems are in constant change so there are no true repeats. Two park… | ||
| SRE in Enterprise | Steve McGhee, James Brookbank | 41m |
Enterprise SRE lays out concrete moves: build reliable services from less reliable infrastructure using redundancy and software (raid analogy), expect a… | ||
| Unified Theory of SRE | Emil Stolarsky | 27m |
Argues that SRE at tiny startups looks nothing like the Google-origin model: the SRE book reflects teams with 70+ contributors and orgs sized like 1200 SREs /… | ||
| Dissecting the Humble LSM Tree and SSTable | Suhail Patel | 39m |
Practical walkthrough of an LSM implementation shows a Python memtable that flushes at 5 items into immutable SSTable files, uses commit logs for durability… | ||
| Caching Entire Systems without Invalidation | Peter Sperl | 34m |
Argues systems become cacheable by resolving all variable inputs up front and splitting into a state gathering layer and an immutable zone so downstream… | ||
| An SRE Guide to Linux Kernel Upgrades | Ignat Korchagin | 34m |
The kernel stable branch produces ~75 bugfix releases (example: 5.15 had 75), over half of them contain >100 commits and 55 of 74 releases carry CVE fixes… | ||
| The Math of Scalability | Avishai Ish Shalom | 43m |
Uses math to prove counterintuitive scaling effects: traffic at Burning Man shows adding lanes slows flow, the birthday paradox gives collisions at 23 people… | ||
| Schema-First Application Telemetry | Yuri Shkuro | 23m |
Meta builds a schema-first pipeline that defines telemetry as Thrift structs, generates strongly typed emitters, and ships compact binary events through the… | ||
| SRE Is Weird, Down the Stack | John P. Looney | 25m |
Infrastructure SRE work uncovers millions of lines of opaque firmware that break real systems: expect 30–50M lines of invisible firmware, mysterious 100 ms SMM… | ||
| SRE and ML: Why It Matters | Todd Underwood | 44m |
Machine learning systems are becoming part of core infrastructure: adoption rises, models run critical paths (recommenders, fraud, autoscaling), and operations… | ||
| Emotional Disaster Recovery: Debugging the Self with... | Tao Hansen | 31m |
Argues that venting doesn’t work — cites Professor Brad Bushman’s study of 600 college students where punching a bag kept anger higher — and recommends short… | ||
| Over Nine Billion Dollars of SRE Lessons - the James Webb... | Robert Barron | 31m |
Project costs reach $9.5B and launch happens December 2021 into L2 halo orbit; the telescope uses a foldable 18-segment mirror that must unfurl perfectly and… | ||
| Rock Fishing and Incident Analysis: Increasing Insight | Thai Wood | 27m |
Government coroner data shows 164 drownings (2004–2018) and five coded causes (inexperience, overtopping waves, slips/falls, freak waves, unknown), a mandate… | ||
| How Can SRE Help Security Governance?... | Mario Platt | 32m |
GRC functions are stuck in spreadsheeted, command‑and‑control workflows and often codify liabilities: ICO reports note 100% non‑adherence to written policies… | ||
| Navigating in the Dark | Nati Cohen | 23m |
Observability, paging and incident tools fail alongside production systems — even with correlated failures from shared infrastructure — illustrated by Air… | ||
| nothing matches | ||