SREcon22 Asia/Pacific
2022-12-07..2022-12-09
Sydney, Australia
as the organisers put it
SREcon is a gathering of engineers who care deeply about site reliability, systems engineering, and working with complex distributed systems at scale. SREcon strives to challenge both those new to the profession as well as those who have been involved in it for decades. The conference has a culture of critical thought, deep technical insights, continuous improvement, and innovation. SREcon22 Asia/Pacific will be held 7–9 December, 2022 at the Sheraton Grand Sydney Hyde Park in Sydney, Australia.
who it's for
site reliability engineers, platform engineers, observability engineers, database engineers, engineering managers, platform architects
topics
Prometheus TSDB internals; OpenTelemetry collector and OTLP; chaos engineering at scale; eBPF and XDP observability; capacity planning for cloud databases; incident impact calculation for payments; continuous production profiling; adaptive distributed tracing controls; designing observable stateful systems; SRE principles applied to CI/CD; kubernetes cluster safety and operators; synthetic monitoring at global scale
| title | author | length |
|---|---|---|
| Computing Performance 2022: What's on the Horizon | Brendan Gregg | 52m |
Survey of near-term platform changes: processors hit ~3.5 GHz plateau with recent beta 3.9 GHz parts, chiplet and 3D stacking and rising core counts (e.g., AMD… | ||
| Move Fast and Learn Things: Principles of Cognition, Teaming... | Laura Maguire, Nora Jones | 57m |
Compares two reviews of the same search outage and shows what changes matter: a quick templated writeup that takes 45 minutes vs one week yields two action… | ||
| How to Not Destroy Your Production Kubernetes Clusters | Qian Ding | 21m |
A custom operator accidentally deletes ~20k Service objects because of a label length mismatch (63 vs 127) between cluster libs, and missing etcd snapshot… | ||
| The Math behind the Incident Aftermath: A Practical Guide to Measuring... | Ashish Patel, Sriram Srinivasan | 35m |
PayPal builds an automated Incident Impact Calculation Framework that pulls per‑minute payment counts from observability APIs, computes a six‑week trimmed… | ||
| OpenTelemetry and Observability: What, Why, and Why Now? | Greg Leffler | 18m |
OpenTelemetry provides vendor-agnostic instrumentation and a single collector binary that runs as agent or gateway, using the OTLP wire format (gRPC/HTTP)… | ||
| Principles of Safety and Reliability Learned from US Navy Landing Signal... | Matthew Brahms | 45m |
Aircraft carrier recovery runs every 45–60 seconds with planes stacked by altitude and timed to the minute; Landing Signal Officers use a gyroscopic meatball… | ||
| Infra Eng to Staff SRE: A Tale of Developing Yourself in an Ever Evolving... | Jess Belliveau | 25m |
An infrastructure engineer recounts moving into a staff SRE role by adopting observability practices (golden signals, error budgets), learning golang and… | ||
| Lifecycle of a Sample in the Prometheus TSDB | Ganesh Vernekar | 30m |
Prometheus TSDB stores each sample as int64 timestamp + float64 value, writes every mutation to a Write-Ahead Log (WAL) for crash durability, compresses on the… | ||
| Metrics Stream Processing Using Riemann | Pradeep Chhetri | 24m |
Riemann is an event stream processor written in Clojure that treats each metric as a map with a TTL per event so expiry drives down-detection; it keeps a… | ||
| Lifecycle of Reusable Automations: Track, Maintain, Deprecate | Renisha Fernandes | 31m |
An in-house automation platform manages 2000+ automations and ~5 million executions per month by ingesting telemetry into a data lake, running SQL queries to… | ||
| Dashboards and Runbooks: Scrapbooking for Engineers | Colin Douch | 21m |
Claim: dashboards and runbooks usually hurt incidents because teams create knee‑jerk, hyper‑specific artifacts that pile up into an unsearchable scrapbook… | ||
| Observability Is Not Analytics! | Andrew Cowie | 42m |
Observability drives a six‑month reengineering to untangle a DB‑centric, tightly coupled system and yields 300+ PRs a month, deployments under 10 minutes and… | ||
| Lessons Learned Building a Global Synthetic Monitoring System | Surajnath Sidh | 35m |
Built a global synthetic monitoring network using blackbox exporter, deploying 22 probe locations across 9 Grafana Cloud regions (≈198 probes total) on… | ||
| Sustaining Everything, Everywhere, All at Once! | Fanjing Meng, Robert Barron, Hua Ye | 43m |
IBM builds a full‑stack sustainability platform running in a 20‑year Beijing data center that collects environmental sensors and agentless telemetry from… | ||
| Introducing the Reliability Map – r9y.dev | Aaron Bowden | 24m |
An open source reliability map at r9y.dev maps concrete capabilities to target SLO tiers (90–99.999…), with each column representing a step often requiring… | ||
| Chaos Engineering at Scale | Sharath Reddy, Venkatesh Maligireddy | 40m |
PayPal builds an internal chaos pipeline on Mangle (VMware), adds ticketing and ACL integration, then runs staged and production game days that inject packet… | ||
| The Multi Layered Cake of Resilience | Joe Chop | 21m |
Shopify handles peaks like 76 million requests per minute and 14 million MySQL queries per second and builds resilience by accepting failure at every layer… | ||
| Capacity vs Efficiency: Building a Globally Scalable Cloud Database | Daniel Marshall | 42m |
Google Cloud Firestore runs as a fully scalable pay-per-request NoSQL database and the operational claim is that efficiency comes from shrinking paid idle… | ||
| Improving Observability, Reliability, and Security of Relational Database... | Sundar Raman Ganesh | 40m |
LinkedIn moves hundreds-to-thousands of MySQL instances to a standardized client stack by shipping custom wrappers around open-source connectors that emit… | ||
| Real-Time Adaptive Controls for Resilient Distributed Systems | Praveen Yedidi | 37m |
CrowdStrike adapts hundreds of tunables in real time for Falcon by sampling errors and latencies from Kafka/Cassandra paths and inferring limits without… | ||
| Improving Machine Learning Development Reliability | Brian Hansen, Yan Yan | 42m |
Meta builds a production model service that catalogs every model, shows owners and on‑call rotations, and serves thousands of users monthly; training backlog… | ||
| How Can We Make Data Integrity Easy? | Adrian Ratnapala | 20m |
Argues for building a reusable managed framework that orchestrates backup → restore image → verification pipeline → back to production so recovery is a tested… | ||
| Cognitive and Self-Adaptive System for Effective Distributed-Tracing... | Susobhit Panigrahi | 11m |
Applications keep existing Jaeger/Zipkin instrumentation while libraries send spans to a span aggregator that merges partial spans, clusters traces by… | ||
| Site Reliability Evangelism: Practice Start-up within an Established... | Piers Chamberlain, Catherine Matheson | 23m |
A yearlong SRE starter replaces one hire with a practical program: custody of an existing telemetry platform, synthetic checks wired to PagerDuty, and… | ||
| Deploying Humans at the Edge of SRE | Jan Peuker | 18m |
Argues that SRE must expand beyond platform teams to include support, TAMs, deployment and integration engineers, and devrel, applying SRE practices to… | ||
| Challenges, Best Practices, and Solutions for Monitoring and Alerting... | Daniel O'Dea | 21m |
Atlassian monitors Jira for millions of customers with a multi‑tenant, shard‑based design and limits per‑tenant metrics because per‑tenant cardinality… | ||
| A Better Way to Manage Stateful Systems: Design for Observability and Robust | Kazuki Higashiguchi | 18m |
Built a business-critical WebSocket connect server that holds long‑living tunnels and proxy connections, discovers blue‑green fails because sessions survive… | ||
| Reliability Reviews in the Wild: Using Data to Drive Production Health | Karthik Nilakant | 22m |
Argues that routine reliability reviews must be driven by measurable signals and describes a four‑phase rollout: DIY templates, facilitated reviews, tooling… | ||
| Leveraging Continuous Production Profiling for Providing Insights into... | Saurabh Badhwar | 21m |
LinkedIn runs always-on Java profiling using async-profiler in 30-minute chunks, generating ~48 TB raw data per day before compression; retention is three days… | ||
| Applying SRE Principles to CI/CD | Mel Kaulfuss | 27m |
Buildkite’s CI/CD is analyzed through SRE practices with concrete measures: an observed 9413 days retrying failed steps across users in one month, agent-backed… | ||
| Gremlins Exposed: Shining a Light on Mischievous Systems | Thomas Cuthbert | 24m |
Shows a practical, repeatable triage: use a concise slip book or A3 problem document to capture clues, start top‑down with HAProxy logs and pipes, spot a… | ||
| Burnout at Scale: What to Try When You Just Can't | Courtney Eckhardt | 41m |
Explains why “be more empathetic” fails and gives concrete, repeatable moves: prioritize active listening only where it’s sustainable, take Mental Health First… | ||
| Backend API Design for SREs | Sam Dunster | 23m |
Good backend API design reduces operational toil and prevents false SLI drops by using clear error classes, predictable IDs, and client-friendly flows. Use… | ||
| Online Database Reliability, Performance, and Consistency Engineering | Yoshinori Matsunobu | 49m |
Operating at Meta scale uses single-primary, multi-replica clusters with sharded replica sets and a Tau cache in front of MySQL; UDB uses LSM/RocksDB to save… | ||
| Migrating Datastores | Tessa Bradbury | 25m |
Buildkite migrates a single-threaded Redis cache to clustered Redis, splitting hot keys across three shards after hitting an 85% CPU cliff on a single… | ||
| Our Experience Tracking and Driving SLO Adoption at Goldman Sachs | Jordan Li, Ivan Ryabov | 43m |
Goldman Sachs builds a configuration-driven dashboard called SLO Repository (internal name SLO Poetry) that exposes SLIs as short-window status-first tiles… | ||
| Operationalizing ML Training Infra at Meta Scale | Shivam Bharuka | 21m |
Meta re-architects AI infrastructure end-to-end: custom GPU-heavy servers with layered lossless networking, PyTorch-based distributed training, and procurement… | ||
| Advanced Linux Kernel Networking Monitoring | Jizhong Jiang, Shane Xie | 44m |
Built a kernel-level monitor called net-exporter that uses eBPF tracepoints (kfree_skb, napi/softirq entry/exit) to stream labeled events and stack traces into… | ||
| Using the Internet as Your Load-Balancer | Martin Barry | 29m |
Compares DNS-based steering versus BGP anycast and gives operational recipes: use EDNS Client Subnet is patchy so resolver IPs mislead, prefer deterministic… | ||
| A Post Incident Review Review | Tom Partington | 44m |
Describes a post-incident practice that drops no root cause and no action items and instead runs a blame aware after-action review plus either a one-hour… | ||
| nothing matches | ||