Handling Poison Pills in Pub/Sub with Dead Letter Topics
Prevent 'poison pill' messages from blocking your Pub/Sub pipelines. Learn how to configure Dead Letter Topics, manage the required IAM permissions, and build a recovery workflow.
Loading…
ReadMeFeed / Community knowledge
#reliability
Real questions. Useful conversations. Find the people who know your stack.
Handling Downstream Outages with Logstash Persistent Queues
Learn how to implement Logstash persistent queues to prevent data loss during downstream outages and handle traffic spikes using disk-based buffering.
logstash
elasticsearchChoosing a Bash Error-Handling Strategy: A Decision Guide for Reliable Scripts
A decision guide comparing Bash error-handling strategies—set -euo pipefail, explicit checks, and ERR traps—with a validated combined pattern that works across Bash 3.2+ including macOS default.
bashArchitecture Note: Bash Strict Mode set -euo pipefail for Reliable Scripts
Architecture note on using Bash strict mode to force early exit on errors, treat unset variables as faults, and propagate pipeline failures, with minimal design, trust boundaries, checks and change triggers.
Raspberry Pi Hardware Watchdog for Unattended Recovery
Headless Raspberry Pi hangs are unfixable remotely. The bcm2835 hardware watchdog can force a defined reboot when software stops kicking the timer, giving unattended nodes predictable recovery at the cost of visibility.
raspberrypiUsing RabbitMQ Publisher Confirms for Reliable Message Publishing
Learn how RabbitMQ publisher confirms give publishers a way to know when a message is safely persisted, and how to enable, test, and trade‑off the feature.
Fine‑Tuning ROS 2 DDS QoS: Reliability, Durability, and Cross‑Node Compatibility
Learn how to adjust ROS 2 DDS Quality‑of‑Service (QoS) settings to balance latency and message reliability, verify correct profiles across nodes, and recover from common mis‑configurations.
rosRolling Updates That Don't Lie to You: Tuning maxSurge, maxUnavailable, and Rollback Depth in Kubernetes Deployments
Kubernetes rolling-update defaults encode assumptions about replica count and cluster headroom that may not match your workload. How to tune maxSurge, maxUnavailable, and revisionHistoryLimit deliberately.
kubernetesArchitecture Note: Designing Boundaries and Failure Isolation in a Network Policy Service
An architecture note on isolating policy decisions from data-plane forwarding in a network policy service, with modular components, trust boundaries, health checks, and failure-mode mitigations.
npss