The Site Reliability Podcast with Fexingo: SRE, Uptime, and Production Engineering

Fexingo

Lucas and Luna cut through the noise around site reliability engineering to examine how real-world SRE teams balance uptime, incident response, and production change. Each episode takes a single concept — error budgets, toil automation, postmortem culture, capacity planning — and grounds it in a specific case: how a major streaming service reduced paging noise, how a payments platform rebuilt its incident command structure, or how a cloud provider manages multi-region failover. Lucas brings the numbers — latency percentiles, MTTR trends, SLO burn rates — while Luna pushes on the human and organizational trade-offs: What does a junior SRE need to know about on-call? How do you measure reliability without crushing innovation? Why do some blameless postmortems actually work? Together they treat SRE not as a certification topic but as a living practice, citing real outages, open-source tools, and engineering blogs. This show is for engineers, ops leads, and platform teams who already know the basics and want to debate the hard edges: Is 99.999% uptime always worth the cost? When should you deliberately degrade service to improve reliability? How do you design for resilience when your system is already in production? Lucas and Luna don't pretend to have final answers — they build the conversation so you can draw your own. If you've ever argued about whether a page was necessary or whether an SLO should be tightened, this is your show. #SiteReliabilityEngineering #SRE #Uptime #ProductionEngineering #IncidentResponse #ErrorBudgets #SLOs #Postmortem #ToilAutomation #CapacityPlanning #Observability #DevOps #PlatformEngineering #Resilience #OnCall #FexingoBusiness #BusinessPodcast #Technology Keep every episode free: buymeacoffee.com/fexingo

  1. 1d ago

    How SRE Teams Use Blameless Postmortems to Improve Reliability

    Episode 129 of The Site Reliability Podcast explores how SRE teams use blameless postmortems to drive systemic improvements without creating a culture of fear. Lucas and Luna dive into the specific practices at a major tech company that reduced repeat incidents by 40 percent after implementing a structured postmortem process. They discuss the key elements of an effective postmortem, the role of data collection, and how to ensure action items get implemented. The episode also covers common pitfalls, such as focusing on individuals instead of systems, and how to foster psychological safety. Listeners will learn concrete steps for running postmortems that actually prevent future outages, including how to write a clear timeline, how to identify contributing factors without blame, and how to track follow-through. Whether you're a seasoned SRE or just starting with on-call, this episode offers practical advice for turning incidents into learning opportunities. #SRE #Postmortem #Blameless #IncidentResponse #Reliability #SiteReliabilityEngineering #DevOps #IncidentManagement #LearningFromFailure #PsychologicalSafety #Google #Technology #FexingoBusiness #BusinessPodcast #ProductionEngineering #Uptime #OnCall #RootCauseAnalysis Keep every episode free: buymeacoffee.com/fexingo

    How SRE Teams Use Blameless Postmortems to Improve Reliability

  2. 1d ago

    How SRE Teams Use Runbooks to Reduce Incident Response Time

    Episode 128 of The Site Reliability Podcast dives into the unsung hero of incident response: runbooks. Lucas and Luna break down how GitLab reduced their mean time to mitigate by 40% by replacing tribal knowledge with structured runbooks, and why many teams fail by treating runbooks as documentation instead of executable playbooks. They cover the anatomy of a great runbook, the difference between runbooks and playbooks, and how to keep runbooks from going stale. If you're in SRE, DevOps, or platform engineering, this episode gives you a concrete framework to make your on-call team faster and less stressed. #SRE #SiteReliabilityEngineering #Runbooks #IncidentResponse #MTTR #OnCall #GitLab #DevOps #PlatformEngineering #IncidentManagement #ProductionEngineering #Uptime #Observability #Automation #Fexingo #FexingoBusiness #BusinessPodcast #Technology Keep every episode free: buymeacoffee.com/fexingo

    How SRE Teams Use Runbooks to Reduce Incident Response Time

  3. 1d ago

    How SRE Teams Use Cost-to-Serve Analysis to Optimize Infrastructure

    In this episode of The Site Reliability Podcast, Lucas and Luna explore how SRE teams are applying cost-to-serve analysis to optimize cloud infrastructure spending. Using a real case from a mid-size fintech company that saved $2.3 million annually by mapping compute costs to specific revenue-generating services, they break down what cost-to-serve actually means in an SRE context—allocating shared infrastructure costs to individual products or features. Lucas explains the technical challenges: tagging chaos, shared databases, and the tension between granularity and overhead. Luna pushes back on whether this is just accounting theater or a real operational lever. They discuss practical implementation: starting with a pilot service, using tools like AWS Cost Explorer and Kubernetes cost allocation, and building a culture where engineers see cost data alongside latency and error rates. The episode also covers common pitfalls like over-aggregation and the risk of optimizing costs at the expense of reliability. Listeners will walk away with a concrete framework for starting cost-to-serve analysis in their own teams. #CostToServe #CloudCostOptimization #FinOps #SRE #SiteReliabilityEngineering #InfrastructureCost #AWS #Kubernetes #CostAllocation #CloudEconomics #Technology #FexingoBusiness #BusinessPodcast #Podcast #DevOps #Observability #CostEfficiency #EngineeringCulture Keep every episode free: buymeacoffee.com/fexingo

    How SRE Teams Use Cost-to-Serve Analysis to Optimize Infrastructure

  4. 2d ago

    How SRE Teams Use Rolling Backups to Recover From Ransomware

    Lucas and Luna explore how site reliability teams are reframing backup strategy in an era of ransomware and silent data corruption. Lucas breaks down the 3-2-1-1 rule — three copies, two media types, one offsite, one immutable air-gapped copy — and explains why recovery testing matters more than backup completion. He contrasts snapshot-based approaches with continuous database archiving using tools like WAL-G and pgBackRest, citing how one fintech found latent corruption in 12 percent of supposedly healthy backups. Luna pushes back on backup frequency dogma and asks whether teams should focus on mean time to recovery instead. The episode drills into a real example: a mid-sized e-commerce company that recovered from a cryptolocker attack in four hours because they had practiced restore drills monthly. The hosts also touch on the cost of immutability — S3 Object Lock adds latency and storage overhead — and why backup SLAs need explicit recovery point objectives and recovery time objectives. A practical, case-driven conversation for engineers who want their backups to actually work when it counts. #SRE #BackupStrategy #RansomwareRecovery #DataCorruption #ImmutableBackups #3-2-1-1Rule #RecoveryTesting #SiteReliability #ProductionEngineering #Uptime #IncidentResponse #WAL-G #pgBackRest #S3ObjectLock #RPO #RTO #FexingoBusiness #BusinessPodcast Keep every episode free: buymeacoffee.com/fexingo

    How SRE Teams Use Rolling Backups to Recover From Ransomware

  5. 3d ago

    How SRE Teams Use Synthetic Monitoring to Catch Outages Before Users Do

    In this episode of The Site Reliability Podcast with Fexingo, Lucas and Luna dive into synthetic monitoring — a proactive approach where SRE teams simulate user traffic to detect issues before they impact real customers. They break down how tools like automated browser scripts and API health checks can catch regressions in staging and production, using a real example of a major e-commerce platform that reduced its mean time to detection by 70 percent. Lucas explains the difference between synthetic monitoring and real user monitoring, when to use each, and common pitfalls like over-alerting. The hosts also discuss how to design effective synthetic checks, from critical user journeys to geographic distribution, and why synthetic monitoring is a key part of a robust observability strategy. This episode is packed with actionable insights for SREs, DevOps engineers, and anyone responsible for keeping digital services reliable. #SyntheticMonitoring #SiteReliabilityEngineering #SRE #Uptime #Observability #ProactiveMonitoring #IncidentResponse #ProductionEngineering #DevOps #Reliability #Technology #Podcast #FexingoBusiness #BusinessPodcast #Monitoring #Alerting #Automation #UserJourney Keep every episode free: buymeacoffee.com/fexingo

    How SRE Teams Use Synthetic Monitoring to Catch Outages Before Users Do

  6. 3d ago

    How SRE Teams Use Capacity Planning to Prevent Outages Before They Happen

    Lucas and Luna dive into the science of capacity planning for site reliability engineering. They break down how Netflix uses predictive modeling to scale infrastructure ahead of demand spikes, avoiding the kind of cascading failures that hit other streaming services during major events. The episode explores real-world examples of capacity planning failures—like the AWS outage that took down half the internet in 2021—and explains the difference between reactive scaling and proactive capacity planning. Lucas shares a concrete framework for setting capacity thresholds using the '80/20 rule' and discusses how SRE teams at companies like Google and Uber use historical traffic patterns to forecast resource needs. Luna challenges the assumption that more capacity is always better, pointing to the 2024 Google Cloud outage caused by overprovisioning. The hosts also touch on the role of automation in capacity planning and the importance of regular load testing. If you've ever wondered why some services stay up during traffic surges while others crash, this episode explains the engineering behind the scenes. #CapacityPlanning #SRE #SiteReliabilityEngineering #Netflix #AWS #GoogleCloud #Uber #PredictiveModeling #InfrastructureScaling #OutagePrevention #LoadTesting #Automation #Tech #Engineering #FexingoBusiness #BusinessPodcast #Technology #SREPodcast Keep every episode free: buymeacoffee.com/fexingo

    How SRE Teams Use Capacity Planning to Prevent Outages Before They Happen

  7. 4d ago

    How SRE Teams Use Feature Flags to Control Risk in Production

    Lucas and Luna explore how feature flags give SRE teams a kill switch for new code, reducing blast radius and enabling canary-style testing without full redeploys. They break down the case of a major e-commerce platform that used feature flags to roll back a catastrophic pricing bug in under 90 seconds, saving millions in potential losses. The episode covers the difference between boolean and multivariate flags, the role of flag management platforms, and common pitfalls like flag debt and stale conditions. Lucas explains how flags shift deployment risk from a binary event to a gradual rollout, and why this approach aligns with error budget policies and incident response runbooks. Luna challenges the operational overhead of flagging everything, and they discuss when feature flags become a crutch rather than a tool. The conversation ties to SRE principles of toil reduction, observability, and safe production experimentation. #FeatureFlags #SRE #SiteReliabilityEngineering #ProductionEngineering #ControlledRollouts #KillSwitch #FlagDebt #CanaryDeployments #IncidentResponse #BlastRadius #LaunchDarkly #ToggleSystems #GradualRollout #RiskManagement #Technology #FexingoBusiness #BusinessPodcast #SiteReliabilityPodcast Keep every episode free: buymeacoffee.com/fexingo

    How SRE Teams Use Feature Flags to Control Risk in Production

  8. 4d ago

    How SRE Teams Use Graceful Degradation to Keep Services Running

    In this episode of The Site Reliability Podcast, Lucas and Luna explore how SRE teams design systems to degrade gracefully rather than fail catastrophically. They examine Netflix's Hystrix library, which introduced circuit breakers and fallbacks to protect microservices, and discuss how Stripe applies similar principles today with feature-level degradation during payment processing surges. Lucas explains the difference between fail-fast, fail-slow, and partial degradation, using the example of a degraded search feature that returns cached results instead of crashing. The conversation covers practical patterns like capping, queuing, and static responses, and highlights the trade-offs between user experience and engineering complexity. By the end, you'll understand why graceful degradation is a first-class reliability pattern, not just a nice-to-have. #GracefulDegradation #SiteReliabilityEngineering #NetflixHystrix #CircuitBreaker #Microservices #Stripe #FailGracefully #ResilienceEngineering #ProductionEngineering #SLO #FallbackPattern #CachedResults #DegradedMode #HighAvailability #FaultTolerance #SRE #Technology #FexingoBusiness #BusinessPodcast Keep every episode free: buymeacoffee.com/fexingo

    How SRE Teams Use Graceful Degradation to Keep Services Running
See All (128)

About

Lucas and Luna cut through the noise around site reliability engineering to examine how real-world SRE teams balance uptime, incident response, and production change. Each episode takes a single concept — error budgets, toil automation, postmortem culture, capacity planning — and grounds it in a specific case: how a major streaming service reduced paging noise, how a payments platform rebuilt its incident command structure, or how a cloud provider manages multi-region failover. Lucas brings the numbers — latency percentiles, MTTR trends, SLO burn rates — while Luna pushes on the human and organizational trade-offs: What does a junior SRE need to know about on-call? How do you measure reliability without crushing innovation? Why do some blameless postmortems actually work? Together they treat SRE not as a certification topic but as a living practice, citing real outages, open-source tools, and engineering blogs. This show is for engineers, ops leads, and platform teams who already know the basics and want to debate the hard edges: Is 99.999% uptime always worth the cost? When should you deliberately degrade service to improve reliability? How do you design for resilience when your system is already in production? Lucas and Luna don't pretend to have final answers — they build the conversation so you can draw your own. If you've ever argued about whether a page was necessary or whether an SLO should be tightened, this is your show. #SiteReliabilityEngineering #SRE #Uptime #ProductionEngineering #IncidentResponse #ErrorBudgets #SLOs #Postmortem #ToilAutomation #CapacityPlanning #Observability #DevOps #PlatformEngineering #Resilience #OnCall #FexingoBusiness #BusinessPodcast #Technology Keep every episode free: buymeacoffee.com/fexingo

Information

You Might Also Like