SREcon18 Asia/Australia
2018-06-06..2018-06-08
Singapore, Singapore
as the organisers put it
SREcon18 Asia/Australia, a gathering of engineers who care deeply about site reliability, systems engineering, and working with complex distributed systems at scale, will be held June 6–8, 2018, at the Grand Copthorne Waterfront Hotel in Singapore.
| title | author | length |
|---|---|---|
| The Evolution of Site Reliability Engineering | Benjamin Purgason | 51m |
Five generations of SRE evolution: starts as firefighters with 1500 minutes MTTR and manual scripts, slides into a harmful service guard whitelist gatekeeper… | ||
| Safe Client Behaviour | Ariel Goh | 32m |
Shows concrete fixes: mobile apps that poll every five minutes should add jitter (example: uniform between 2.5–7.5 minutes) to avoid synchronized spikes, and… | ||
| Service Monitoring Manual—2018 Edition | Nikola Dipanov | 22m |
Modern monitoring treats observability as first-class: separate agents emit host, service, mesh and aggregate data; counters and latency histograms go to a… | ||
| Introduction to Alibaba Monitoring System | Ren Xinchi | 20m |
Alibaba maps hundreds of KPIs into a business CMDB called Hammurabi CMDB and monitors five canonical signals — total count, success count, success rate… | ||
| Building SRE: Culture from the Outside In | Todd Palino | 22m |
Describes LinkedIn’s concrete SRE culture: require engineers to code for infrastructure and monitor everything so incidents are measurable; automate fixes so… | ||
| Quantifying Empathy with Service Level Objectives | Ketan Gangatirkar | 45m |
Argues that Service Level Objectives must map to real user outcomes and gives concrete rules: use six flavors of user happiness (availability, responsiveness… | ||
| Doing Things the Hard Way | Chris Sinjakli | 20m |
Company grows from ~25 to ~215 people and engineers from 8 to 60, and three concrete failure modes emerge: splitting product and infra early breaks the… | ||
| Achieving Observability into Your Application with OpenCensus | Emil Mikulic | 20m |
OpenCensus provides a consistent, language‑specific library (seven languages supported) that automatically propagates trace context, collects RPC latency and… | ||
| Know Thy Enemy: How to Prioritize and Communicate Risks | Matt Brown | 37m |
CRE/Customer Reliability Engineering builds a simple spreadsheet that converts each risk into expected cost by multiplying estimated frequency (ETBF) and… | ||
| Data Visualization for SREs—an Essential Skill for Quick Debugging | Yash Shah | 20m |
Yash Shah, LinkedIn
SREs are software engineers with a broad skill set who work with systems in general. Depending on the type of work and teams, our usual… | ||
| You Can't Stop Fires with an Ambulance | Piers Chamberlain | 19m |
Alerts double during cloud migration and incident count keeps rising despite reduced MTTR; 30 SRE engineers build a symptom-first alerting stack, a Klaxon… | ||
| Comprehensive Container-Based Service Monitoring with Kubernetes and Istio | Fred Moyer | 49m |
Shows how Istio service mesh + Kubernetes wires telemetry to monitoring without code changes, and how a Circonus mixer adapter (about ~20 lines of handler… | ||
| How to Make Releases Safer in Baidu | Pingping Xue, Yu Chen | 21m |
Baidu finds 54% of faults come from changes and moves to staged, package-based rollouts (sandbox → gray → blue) that deploy 1% then 99% of instances to limit… | ||
| Cultural Nuance and Effective Collaboration for Multicultural Teams | Ayyappadas Ravindran | 27m |
Ayyappadas Ravindran, LinkedIn
What is considered "good-communication" is different for different cultures. In some cultures "good communication" is being as… | ||
| Automatic Datacenter and Service Deployments... | Xiaoxiang Jian | 39m |
Alibaba builds Apsara Stack as an image-driven, immutable bare-metal infrastructure with a central image builder, per-datacenter image caches and local… | ||
| From Monitoring to Automated Testing of Your Infrastructure Code | Jesse Reynolds | 28m |
Argues that infrastructure code needs integration/acceptance tests that run in CI, shows how to spin VMs, apply Puppet/Chef/Ansible and assert state so… | ||
| Shopify's Move from the Data Centre to the Cloud | Scott Francis | 41m |
Shopify migrates over 600,000 shops and ~80,000 rps from colo to cloud by sharding shops into 100+ pods, moving shops one-at-a-time with a shop mover that uses… | ||
| Ensuring Reliability of High-Performance Applications | Anoop Nayak | 34m |
LinkedIn Lite is engineered to load under 6 seconds by keeping page weight to ≈75 KB for 100 kbps networks, using server-side rendering, early flush, and… | ||
| Smarter Disasters: End-to-End Automation for Incidents | Karthik Nilakant | 22m |
Xero scales incident handling for 1.3 million orgs and >1,000 releases a year by turning customer reports into alerts via an issue report pipeline, routing… | ||
| Debugging at Scale—Going from Single Box to Production | Kumar Srinivasamurthy | 21m |
Covers concrete techniques used at Bing and Cortana: Zipkin-style distributed tracing with trace ID based log retrieval, Kafka + Storm pipelines to compute… | ||
| Productionizing Machine-Learning Services: Lessons from Google SRE | Salim Virji, Carlos Villavieja | 40m |
Claim: models must be retrained continuously because accuracy decays in days or weeks; Google sees production models age fast and YouTube’s “watch next” needs… | ||
| Pro Tip: Save Money on Outages by Having a Bot Do the Heavy Lifting | Cezar Alevatto Guimaraes | 21m |
Cezar Guimaraes, Microsoft
Humans are slow, unreliable and hard to train. Azure has saved many millions of downtime minutes by using a knowledgeable and… | ||
| Evolution of SRE and Rising Need of SRE Catalyzers | Isha Ganeriwal | 28m |
LinkedIn recounts SRE rising from a seven‑person site‑ops crew in 2010 to backing product growth to 5M members by 2017 at ~30% YoY, and credits concrete… | ||
| How to Serve and Protect (with Client Isolation) | Frances Johnson | 24m |
Maps SREs implement client isolation by adding layered defenses: coarse global protection at the edge, per-client quotas (one component was oversubscribed… | ||
| A Tale of One Billion Time Series | Ruiyao Yao | 42m |
Baidu stores and serves 1 billion time series with a three-tier TSDB: Redis+HBase+HDFS for hot, warm and cold data; ingestion hits ~40M points/sec with 50… | ||
| Isolation without Containers | Tyler McMullen | 20m |
Fastly describes a compiler+sandbox that compiles WebAssembly to native code (Katana / Frittata) to run many thousands of sandboxes concurrently inside one… | ||
| Automatic Traffic Scheduling for Internet Connectivity Failures | Liuqing Zhang | 23m |
Baidu’s system detects 3–5 PoP outages and 5–10 backbone breaks weekly and classifies failures as PoP, backbone, or local ISP by aggregating probe failures… | ||
| Lessons Learned from Our Main Database Migrations at Facebook | Yoshinori Matsunobu | 47m |
Facebook builds MyRocks, a MySQL storage engine on RocksDB, and migrates the UDB from compressed InnoDB to MyRocks, cutting storage roughly 50% space reduction… | ||
| PV Monitoring Based on Linear Regression | Wang Bo | 18m |
Proposes replacing moving averages with robust linear regression using least absolute deviations solved by iterative reweighted least squares, initializing… | ||
| Do Docs Better: Practical Tips on Delivering... | Riona MacNamara | 20m |
Documentation is framed as core engineering work with concrete practices: move docs into source control as simple markdown and integrate them with code reviews… | ||
| Characterizing and Understanding Phases of SRE Practices | Kurt Andersen | 33m |
A five-stage Shu‑Ha‑Ri model adapted from martial arts maps SRE maturity across incident response, postmortems, SLOs, monitoring, capacity and performance… | ||
| Scaling Yourself for Managing Distributed Teams... | Paul Greig | 43m |
Distributed SREs span three geographies and 18 engineers supporting JIRA and Confluence, and rapid growth creates friction, duplication and cost. Practical… | ||
| Interviewing for Systems Design Skills | Sebastian Kirsch | 47m |
Google SRE’s NALSD format tests concrete system design work: pick a simple problem, scale it to create real tradeoffs, and probe for numbers, failure modes and… | ||
| Scaling a Distributed Stateful System: A LinkedIn Case Study | Sai Kiran Kanuri | 37m |
Sai Kiran Kanuri, LinkedIn
A system is called scalable if it manages to take additional users and requests without losing any noticeable performance. Scaling… | ||
| Mentoring: A Newcomer's Perspective | Leoren Tanyag | 21m |
REA Group runs an 18-month graduate rotation that embeds mentoring into team commitments, uses ticket vouchers to claim teaching sessions (used here to learn… | ||
| You Get What You Measure—Why Metrics Are Important | Kumar Srinivasamurthy | 23m |
Shows concrete failures and fixes: unit mismatches cost a Mars probe and pounds vs kilograms causes an Air Canada dual-engine flameout; availability expressed… | ||
| Blame. Language. Sharing: Three Tips for... | Lindsay Holmwood | 55m |
Argues that learning is blocked by three concrete inhibitors — blame, language, sharing — and shows practical fixes: use the framing hierarchy what > how > why… | ||
| A Theory and Practice of Alerting with Service Level Objectives | Jamie Wilkinson | 41m |
Argues for alerting on SLOs and error budgets instead of many cause-based pages, with concrete math: count requests inside the app or at the load balancer… | ||
| Production Engineering: Connect the Dots | Espen Roth | 24m |
Explains how Facebook hires and trains new grads into a hybrid SRE/software role that sits embedded production engineers on software teams, shares a 30‑person… | ||
| Mental Models for SREs | Mohit Suley | 25m |
Connects specific cognitive errors to concrete SRE practices: WWII plane bullet‑hole analysis illustrates survivorship bias; a production outage shows… | ||
| nothing matches | ||