SREday London 2025 Q3
as the organisers put it
Conference on Site Reliability Engineering, Cloud and DevOps. Probably the most awesome SRE conference in London, UK!
| title | author | length |
|---|---|---|
| 7 Deadly Traps of SRE | Miko Pawlikowski | 22m |
Seven concrete failure modes: Hero trap (all‑nighters create single points), Tooling trap (adding Kubernetes and dozens of shiny stacks increases learning… | ||
| What Observability Can Learn From BI | Peter Marshall | 24m |
Argues that observability must decouple ingestion, storage, compute and UIs the way analytics did, using Kafka and OpenTelemetry for transport, S3/column… | ||
| Strategic AI Agents for K8s SRE: Beyond Ad-hoc Prompting | Andrei Pokhilko | 33m |
Built an internal AISR agent that starts from pseudo CLI tool calls and later moves to Agno tool calling, using parallel data querying and deterministic… | ||
| Optimizing K8s and unlock blockers with Pod Live Migration | Kim-Norman Sahm | 24m |
Cast AI presents container live migration that moves pods between nodes inside a cluster without restart by creating a shadow pod, mirroring memory and… | ||
| Beyond Single-Cloud SRE | Dheeraj Bandaru, Francois Martel | 23m |
Demonstration shows an Azure SRE Agent and Hawkeye (Newbird) collaborating across an Azure frontend and AWS backend: a ServiceNow incident triggers parallel… | ||
| Why We Built a Custom Exporter to Replace Our OTel Pipeline | Rory Crispin | 25m |
ClickHouse-to-ClickHouse native transfers replace the OTEL JSON pipeline, copying system tables byte-for-byte with a Go client to avoid parsing, batching… | ||
| The Airgap Paradox: Adversarial Tactics & Defensible Design | Sean Behan | 15m |
Air gaps only buy time: past breaches like Agent BTZ (year‑long cleanup of a Supernet worm) and Stuckset (air gap jump via USB using four zeroday exploits and… | ||
| Automating Kibana Alerting: When GitOps meets UI | Ayd Asraf | 26m |
Company consolidates logs and metrics into Elastic and moves alerting to Kibana to tame 3 billion logs daily and 2,500 alerts per day across 1,300+… | ||
| Breaking Bad (Systems): a journey through chaos engineering | Simon Hanmer | 23m |
Shows a practical path from history to hands‑on: Netflix tools inspiring Chaos Monkey style failures, then using AWS Fault Injection Simulator (FIS) to run… | ||
| Making AI Accountable: Azure AI Foundry | Stratos Kourtzanidis | 22m |
Demo shows Azure AI Foundry running an evaluation SDK and a GitHub Action to catch regressions in CI/CD and production. It measures coherence, fluency… | ||
| From Blind Spots to Total Vision: O11y at Massive Scale | Bharav Patel | 25m |
Argues that lowering mean time to detect is the key goal and shows a working pipeline: OpenTelemetry agents and Fluent Bit collect traces and logs, metrics go… | ||
| Echoes in the Core: Designing Resilient Platforms | Prithvi Raj Venkat Raj | 23m |
Claims an open-source prototype called Cordon state manager that acts as a super control plane to manage multi-cloud, multi-cluster fleets and add-on services… | ||
| Beginner's Guide to Terragrunt | Ceyda Duzgec | 21m |
Terragrunt acts as a thin wrapper around Terraform, generating remote backends and orchestrating modules so a VPC deploys first and its output feeds an RDS… | ||
| Who Owns Your SRE Stack? The Rise of AI Agents | Cynthia Akiotu | 15m |
Claims that agents already monitor and remediate incidents now with Azure SRE agent (preview) and pilots from Fuzzy Labs, capture workflows into GitHub and can… | ||
| Cutting Through Metrics Cardinality Noise with VictoriaMetrics | Virginia Diana Todea | 19m |
Explains why high metric cardinality drives CPU, memory, storage and query latency, shows a live cardinality explorer demo in the VictoriaMetrics UI that lists… | ||
| Empowering Developers through Open-Source AI | Amit Kushwaha | 13m |
Samanova runs open-source models on proprietary reconfigurable data units that use a three-tier memory and 16‑chip nodes to cut inference latency; a single… | ||
| From Git Push to Exit | Dima Malyshenko | 25m |
A startup builds continuous deployment from day one using GitHub with a single master branch, feature flags via Launch dly, AWS CDK (Python), AWS Secrets… | ||
| Driving Platform Adoption: Embedded SRE | Leena Mooneeram, Jorge Lainfiesta | 21m |
Company adopts Backstage as a catalog and self‑service layer but faces low uptake, so a central platform team of 10–15 platform engineers supporting ~120… | ||
| 10 Common Things People Do Wrong in K8s Environments | Kim-Norman Sahm | 20m |
Lists 10 concrete Kubernetes mistakes and fixes: Pod Disruption Budget drain-lock causes stuck one-replica deployments during node drains; privileged… | ||
| Flexible Kubernetes Multitenancy with vCluster | Piotr Zaniewski | 32m |
Creates a fully functional Kubernetes control plane as Kubernetes control plane inside a pod (deployed as a StatefulSet or HA mode) and runs a small sync… | ||
| There's no AI without APIs | Pooja Mistry | 23m |
Argument that modern AI agents require APIs to act, not models alone, and that making APIs machine-ready fixes hallucination and broken workflows. Postman… | ||
| OTel You It's Not Just for Backend! | Carly Richmond | 29m |
Live demo shows a small record store demo app instrumented with the OpenTelemetry JavaScript/TypeScript SDK sending logs, metrics and traces through a proxy to… | ||
| Platform Engineering and AI - Two Buzzwords Finally Meet! | Michael Coté | 24m |
Collects industry practices (not original field data) and argues the one concrete architecture choice is a gateway/broker choke point to observe, apply policy… | ||
| Move to Integrated SRE Ops at a Retail organization | Mitul K. Jain | 21m |
16-week proof embeds an SRE into three product teams across two brands on a microservices stack (Azure, Kubernetes, Azure DevOps) and builds a 27-capability… | ||
| PaaS isn’t enough: your internal platform in Multi Cloud | Ashish Upadhyay | 15m |
Argues that a homegrown internal developer platform replaces PaaS limitations by delivering a composable control plane that exposes a single provisioning API… | ||
| Using OpenTelemetry to Improve Service Level Objectives | Martin McLarnon | 26m |
Argues that OpenTelemetry supplies the actual instrumentation developers need to implement SLOs, turning hope-based deployments into data-driven releases… | ||
| Aggregating Metrics In-Flight: Challenges, Opportunities | Roman Khavronenko | 20m |
Claims that a stream-aggregation collector (VictoriaMetrics agent) can precompute and drop raw series to prevent cardinality explosions, reduce query cost and… | ||
| From experimentation to continuous verification | Joris Bonnefoy | 26m |
Defines chaos engineering as formalizing resilience not breaking things and shows Datadog building a usable chaos platform that ties Gremlins/Litmus/Chaos… | ||
| Kepler - Optimized power usage monitoring for Kubernetes | Mayank Goyal | 20m |
Kepler reports real‑time power at process, container and pod granularity and exports metrics to Prometheus (scraped from port 9102) so dashboards show values… | ||
| What's Happening AMBA? | Jamie Pullinger | 27m |
Azure Monitor Baseline Alerts (Amber) automatically applies alerting policies at management‑group level, creating NSG delete alert and UDR update alert rules… | ||
| Igniting Sociotechnical Systems with Reliability Advocacy | Stephan Mousset | 24m |
A grassroots reliability advocacy movement at ING builds a global SRE hierarchy, turns local fixes into a scaled observability platform, and embeds reliability… | ||
| Why internships and entry-level roles matter in Tech | Alecia Cotterell | 26m |
Tech internships matter because they plug clear hiring gaps: industry data shows a 50% decline in tech adverts (2019–2024) and a 70% fall in programmer roles… | ||
| Spotify's Growth Exposed the Abstraction Paradox | Stuart Clark | 20m |
Rapid growth makes clean APIs hide crucial operational detail, causing new hires to take 60 days to 10th PR while abstractions work for 90% vs 10% of cases and… | ||
| State in a stateless world: data serving in the cloud | Harel Safra | 23m |
Argues that state still matters in cloud-native stacks and gives concrete choices: use distributed SQL (CockroachDB, YugabyteDB, Aurora DSQL) to get… | ||
| How We Built ClickStack - an open source O11y stack | Dale McDiarmid | 24m |
ClickStack is an opinionated, open-source observability stack built from ClickHouse (Apache-2) for storage, HyperDX (MIT) for UI, and OpenTelemetry for… | ||
| O11y in One: Exploring ClickHouse as a Unified Telemetry Database | Josh Lee | 26m |
Argues for one unified telemetry store by using ClickHouse to replace separate logs/metrics/traces stacks, claiming it handles petabyte-scale and “trillions of… | ||
| Observability Made Simple: Monitoring Plants with IoT & Grafana | Marie Cruz | 21m |
Built a soil‑moisture monitor using an ESP32 devkit, Arduino IDE and a moisture probe that reads analog values every 5s, sends metrics via the Prometheus… | ||
| 15 NGINX Metrics to Monitor | Dave McAllister | 27m |
Open source NGINX exposes roughly 10 metrics via the status module (active connections, request rate, request_time, upstream response time, error rates… | ||
| TLA+: Your Secret Weapon Against Concurrency Hell | Frederic G. Marand | 24m |
TLA+ models actual system states and finds concurrency bugs before code lands: Amazon credits an S3 model with catching real faults, Kafka’s replication… | ||
| How Cloud AI Leaks What It Learns | Meletius Mgbeodichimma Igbokwe | 13m |
Cloud-hosted LLMs leak memorized training data via crafted prompts and side channels; plain examples include a Stanford prompt ignore previous instructions… | ||
| nothing matches | ||