Practical writing on monitoring, incident response and keeping the pager quiet for the right reasons.
Teams do not ignore alerts because they are careless. They ignore them because the alerts are badly designed. A way to audit a month of pages and cut the noise.
Read postA status page earns trust by being accurate, specific and boring. What to publish during an incident, what to leave out, and a template for updates.
Read postA check from a single place cannot tell your outage from its own network trouble. How a second vantage point settles it, and the rules for combining two opinions.
Read postPage one person first, wait, then widen. How to design tiers, timings and a last resort so an alert always reaches someone without paging everybody.
Read postAn expired certificate is a full outage, not a degradation. Where renewal fails, which clients break first, and how to set a warning that leaves time to fix it.
Read postOne failed check is a weak signal. Here is the arithmetic behind confirmation thresholds, what they cost in detection time, and how to pick a number.
Read post