Our on-call rota for six people
How our Leeds team shares the pager, hands over open work and checks that a backup is available.
Journal
Notes on monitoring, incidents and running a small software company.
Journal
What a 30-second interval costs us
Faster checks mean more requests, more stored results and more work keeping the monitoring system reliable.
Writing a status update people can use
A useful incident update names the impact, the current work and the time of the next update.
Why we confirm from a second region
A failed request is evidence, but a second regional check helps us decide when to wake someone up.
