SREcon17 Asia/Australia
2017-05-22..2017-05-24
Singapore, Singapore
| title | author | length |
|---|---|---|
| SRE From Inception to Global Scale | Bruno Connelly, Viji Nair | 43m |
LinkedIn pivots from biweekly monster releases to a trunk-based, self-service deployment model with canaries, metric-validation and A/B feature gating… | ||
| Next Generation of DevOps AIOps in Practice | Xianping Qu, Jingjing Ha | 33m |
Baidu builds an AIOps platform that standardizes operations across thousands of apps and hundreds of thousands of servers, claiming a unified operations… | ||
| How Could Small Teams Get Ready for SRE | Zehua Liu | 32m |
Zendesk Singapore scales from 10 to 40 engineers and dedicates a quarter to technical debt, then forms a tooling team that builds a deployment pipeline using… | ||
| How We Built TechLadies in Singapore | Elisha Tan | 26m |
Community-led initiative teaches 177 women in Singapore and Malaysia coding through an open-source Ruby on Rails curriculum and workshops, then runs a… | ||
| Focal Impact The Service Pyramid | Michael Elkin | 22m |
Framework claims: prioritize work by a six-layer Service Pyramid so lowest layers (hardware/provisioning, monitoring, lifecycle) must be solid before capacity… | ||
| Smart Monitoring System for Anomaly Detection on Business Trends | Zhaogang Wang | 20m |
Zhaogang Wang, Alibaba Group
Anomaly detection based on time series analysis approaches has been a focused theme in the field of monitoring, especially for… | ||
| Graphite@Scale or How to Store Millions of Metrics per Second | Vladimir Smirnov | 26m |
Booking.com pushes Graphite-compatible stack to handle millions of unique metrics/sec, ingesting thousands of gigabits/s and storing hundreds of terabytes in… | ||
| Data Checking at Dropbox | David Mah | 17m |
Dropbox runs periodic sanity checks that scan every storage unit (e.g., hundreds of thousands of hard drives) and re-read blocks to verify checksums, validate… | ||
| Managing Server Secrets at Scale with a Vaultless Password Manager | Ignat Korchagin | 33m |
Cloudflare describes using firmware flash to store a single 256‑byte master seed in UEFI variables and a deterministic KDF + DRBG pipeline to derive per-server… | ||
| Open Falcon A Distributed and High Performance Monitoring System | Yao-Wei Ou, Wei Lai | 21m |
Open-Falcon is a modular, microservices monitoring stack that companies use in production to collect, store and alert on thousands of metrics; it uses a push… | ||
| Talking to an OpenStack Cluster in Plain English | Wei Xu | 26m |
Wei Xu, Tsinghua University
Modern systems build on layers and layers of abstractions with tons of modules. These abstractions help the development but make… | ||
| Distributed Consensus Algorithms | Laura Nolan | 48m |
Distributed consensus is what prevents split-brain problem and lets services like Photon, Chubby, ZooKeeper and Spanner give strong consistency by ordering… | ||
| A Distribution Framework over ANSIBLE | AiZhen Chen | 21m |
Claims a distributed scheduling layer built on Ansible that scales to tens of thousands of nodes by splitting responsibilities and running many Ansible… | ||
| Draining the Flood—A Combat against Alert Fatigue | Yu Chen | 29m |
Baidu’s Argus alert stream averages >1.3k alerts per engineer per week (130/day, 100 daytime, 30 night), only ~15% are actionable and ~58% are duplicates… | ||
| Good, Better, Best, Mobile User Experience | Fred Wu | 29m |
Describes a deployable stack that combines injected SDK/JS, probe-based synthetic transactions, automatic application-topology discovery and code-level timing… | ||
| Reliable Launches at Scale | Sebastian Kirsch | 44m |
Google runs up to 180 production-impacting launches per week (7,000 total tracked), so a dedicated Launch Coordination Engineers team audits launches as… | ||
| How to Provide a Reliable Ridesharing Service | Ming Hua, Lin Tan | 33m |
Didi faces 500% annual request growth, 400+ cities, and 400+ releases per day and reduces downtime from about 40 to about 22 minutes per month. Capacity… | ||
| Measuring the Success of Incident Management | Gerry Millar | 26m |
Atlassian reports a 66% reduction in repeat incidents and steady improvement in median time to recovery by standardizing incident roles, control systems and… | ||
| Managing Changes Seamlessly on Yahoo's Hadoop Infrastructure Servers | Vineeth Vadrevu | 25m |
Yahoo runs Hadoop on ~45,000 machines and pushes ~15 changes per week across OS, Hadoop and infra. Automation enforces one-click deploy from sandbox → research… | ||
| Event Correlation: A Fresh Approach towards Reducing MTTR | Renjith Rajan | 46m |
Renjith Rajan and Rajneesh, LinkedIn
LinkedIn has hundreds of microservices spanning across different data centres and dependent on each other. Even though… | ||
| Automated Troubleshooting of Live Site Issues | Sriram Srinivasan | 22m |
PayPal builds an Auto Troubleshooting Platform that ingests tickets and alerts from a triage server, enriches them via an associative search and utility… | ||
| A Unit Test Would Have Caught This... | Andrew Ryan | 26m |
Argues that small, cheap tests give big quality gains and shows concrete tools and practices: use shellcheck and runtime flags (-u -e) for Bash, choose shunit2… | ||
| Testing for DR Failover Testing | Zehua Liu | 28m |
Zendesk Chat runs its backend in one datacenter and failover tests initially fail often (first exercise needs three tries, later one attempt succeeds). The… | ||
| Accept Partial Failures, Minimize Service Loss | Daxin Wang | 13m |
Baidu describes a pragmatic strategy that reduces user-impact during failures by isolating faults instead of full recovery. It enforces N+1 data center… | ||
| SREBot More than a Chatbot... | Cezar Alevatto Guimaraes | 24m |
Azure SREBot replaces tribal knowledge and automates incident response for Azure services, joining Skype for Business incident bridges, harvesting alerts, VMs… | ||
| Merou: A Decentralized, Audited Authorization Service | Luke Faraone | 20m |
Merou is Dropbox's open-source authorization service that manages access across corporate network, production data centers, and AWS and claims to remove… | ||
| Canary in the Internet Mine | Brook Shelley | 13m |
Argues for incremental, reversible releases using an edge router so traffic shifts, not deploys, control exposure; implements a reverse-proxy that routes a 30%… | ||
| InnoDB to MyRocks Migration in Main MySQL Database | Yoshinori Matsunobu | 50m |
Facebook migrates massive, latency‑sensitive MySQL shards from InnoDB to MyRocks (RocksDB inside MySQL), and reports reduced DB size by half and ability to… | ||
| Golang's Garbage | Andrey Sibiryov | 48m |
Uber measures Go nodes spending ~75% of runtime in background GC on a ~96GB heap; perf top and flamegraphs plus go debug gctrace reveal long background… | ||
| Capacity Planning and Flow Control | Jun Zhang, Jiajiang Lin | 38m |
Alibaba lays out a four‑stage capacity plan: traffic estimation, single‑machine stress testing, system‑wide Full-link Pressure Test, and runtime flow control… | ||
| Managing Capacity | Anuprita Harkare | 20m |
LinkedIn describes a capacity system serving 500 million members across 95,000 servers and 29 MW, with Hadoop at 170 PB, Espresso holding 4.2 PB across 237… | ||
| Distributed Scheduler Hell | Matthew Campbell | 33m |
DigitalOcean moves hundreds of VMs into containers and tests Mesos+Marathon, Kubernetes, Nomad, Swarm and manual scheduling, finding Mesos+Marathon easy to… | ||
| SRE Your gRPC—Building Reliable Distributed Systems Illustrated with gRPC | Gráinne Sheerin, Gabe Krabbe | 48m |
Adds a translator backend to a simple C++ gRPC greeter and shows concrete reliability fixes: set and propagate explicit deadlines (configurable via a flag)… | ||
| Operationalizing DevOps Teaching | John Contad, Matt Witherow | 22m |
Claim: giving developers concrete operational knowledge leads them to operate services. Experiments include 50-person classes (SNS), one-to-one mentorship… | ||
| Scaling Reliability at Dropbox Our Journey towards a Distributed Ownership | Sat Kriya Khalsa | 33m |
Dropbox shifts reliability into distributed ownership by building a ticketed incident pipeline: memorable drop subs names, four severity levels up to sub-zero… | ||
| nothing matches | ||