Engineering writeups and incident post-mortems for #sre.
Articles: 15In-depth engineering analysis: Zero-downtime kernel live patching with kpatch in enterprise Linux environments. Practical diagnostics, logs and production configuration.
In-depth engineering analysis: Post-mortem: TLS certificate renewal failure due to stale redirect in HTTP-01 challenge. Practical diagnostics, logs and production configuration.
In-depth engineering analysis: Linux kernel memory fragmentation: compact_memory and memory watermark tuning. Practical diagnostics, logs and production configuration.
In-depth engineering analysis: Post-mortem: PostgreSQL connection leak in background worker pool during upstream timeouts. Practical diagnostics, logs and production configuration.
In-depth engineering analysis: Post-mortem: systemd-journald log drop causing silent monitoring blindness. Practical diagnostics, logs and production configuration.
In-depth engineering analysis: Post-mortem: RAID controller battery failure disabling write-back cache. Practical diagnostics, logs and production configuration.
In-depth engineering analysis: Post-mortem: RabbitMQ memory alarm deadlock during incoming queue spike. Practical diagnostics, logs and production configuration.
In-depth engineering analysis: Post-mortem: DNS TTL 0 storm crashing internal HashiCorp Consul cluster. Practical diagnostics, logs and production configuration.
In-depth engineering analysis: Zero-downtime database schema migrations: the Expand and Contract pattern in practice. Practical diagnostics, logs and production configuration.
In-depth engineering analysis: systemd cgroups v2: practical differences between MemoryHigh and MemoryMax in production. Practical diagnostics, logs and production configuration.
In-depth engineering analysis: Kafka consumer rebalance storm: why rolling deployments caused cascading partition lags. Practical diagnostics, logs and production configuration.
In-depth engineering analysis: PostgreSQL autovacuum freeze storm: production downtime due to transaction ID wraparound. Practical diagnostics, logs and production configuration.
In-depth engineering analysis: OOM killer priority tuning: how a bad oom_score_adj killed a critical monitoring daemon. Practical diagnostics, logs and production configuration.
How short-lived inter-service HTTP calls exhausted ephemeral ports and triggered cascading 500 errors.
A classic war story: why 400 GB of free disk space could not prevent 'No space left on device' when ext4 inode table hit 100% capacity.
