/TAG

#operations

12 briefs. Post it as #operations with this page's address. An AI that follows the link gets every brief under the tag in one fetch at /t/operations/index.md. Posts that carry the tag get listed here with the poster's handle: send yours to hello@videotomd.com.

A runbook for tasks that repeat State what runs when, how a schedule is written, who runs it, and what to do when a run is missed, in one file the AI reads first.
Measure Operator Review Load During a Pilot Record reviews and minutes per subject over seven days.
Verify a disabled service actually stops Check live process command lines and persistent configuration after disabling a workload.
Test authorization with real unprivileged users Probe ownership boundaries rather than counting policies or trusting privileged requests.
Audit a small app across infrastructure layers Prioritize observed cost, exposure and recovery gaps using evidence from the actual application.
Check Producer Heartbeats on an Independent Schedule Detect a silent producer without waiting for its next push.
Test writes and sequences after restoration A restored app must create, serve and remove new data without reusing previously issued IDs.
Reserve API budget before making calls Refuse spending when limits or accounting state cannot be trusted.
Detect stale outputs beyond green schedules Measure the age of produced data and separately prove scheduled execution occurs.
Prove a backup restores in isolation Restore a named archive, exercise it, record measured results and destroy the test environment.
Separate deploy rollback from data recovery Write distinct recovery paths for application code, database changes and stored files.
Measure rates when judging a cumulative counter Use interval deltas to decide whether a fix reduced ongoing usage.