Omnythos
Practical engineering writing for people who keep systems alive.
Zero-Downtime Deployments Without the Drama
May 23, 2026
The folklore version of zero-downtime deployment involves blue-green environments and instant cutovers. The operational version is mostly about boring details: health checks that test real work, connection draining that actually finishes, and database migrations that respect the running version.
The rule that saves the most pain is expand-and-contract for schema changes. Add the new column, deploy the code that writes both, backfill, then remove the old one later. Every step is safe to roll back because no deploy ever depends on a migration that the previous version cannot tolerate.
Reading Latency Percentiles Without Fooling Yourself
May 13, 2026
Arithmetic averages hide the extreme tail latencies that percentiles make evident. Even if an endpoint posts an average duration of fifty milliseconds, one out of twenty calls might experience a grueling two-second delay; customers subjected to cold paths and overloaded database shards are the ones …
Why Edge Caching Still Matters in 2026
July 26, 2026
Periodic declarations that edge caching is dead consistently overlook networking realities: fiber lines still obey physical constraints. Even with cheap bandwidth and fast runtimes, a round trip between Sao Paulo and Frankfurt easily consumes hundreds of milliseconds. No amount of runtime tuning can…
A Field Guide to Graceful Degradation
June 7, 2026
Architectures need an explicit hierarchy of failure. Recommendation engines should yield before purchasing flows break; query auto-complete should disable before search fails; generated previews should drop before source content vanishes. Formulating this priority map and reinforcing it through isol…
Understanding TLS 1.3 Session Resumption
June 1, 2026
TLS 1.3 collapsed the handshake to a single round trip, but the real latency win for repeat visitors comes from resumption. A client presenting a valid pre-shared key can skip straight to encrypted application data, which on mobile networks can be the difference between a page that feels instant and…
More reading
- HTTP/3 and QUIC: What Changed for Operators — Networking, June 7, 2026
- Structuring DNS for Reliability — Networking, May 27, 2026
- The Hidden Cost of Chatty Microservices — Engineering, August 26, 2026
About us
Our contributors have spent years on-call for large platforms. This site collects the playbooks, postmortems and reference material we wish someone had handed us earlier.