How SRE works, for teams without an SRE team.
The practices our SREs use daily, explained in plain English with what we recommend and how to start. They apply whether or not you work with us.
-
6 min
What does 24/7 on‑call really take?
Which services need night cover at all, how fast someone must respond, and how many people a sustainable rota needs.
Read the guide -
6 min
How reliable does each service need to be?
Reliability targets and error budgets: what they mean, who should set them, and what to do when the budget runs out.
Read the guide -
7 min
Why your alerts wake people for nothing
Alert on what users feel, page on how fast the error budget is going, and keep quiet services from paging on bad luck.
Read the guide -
5 min
Why monitoring isn’t enough
Where dashboards stay green while users struggle, and the order to add user-journey checks, traces and labelled logs.
Read the guide -
7 min
Incident reports that make the next one less likely
Mitigate first, look for contributing factors rather than one cause, write without blame and make sure the actions close.
Read the guide -
6 min
How to write your first runbooks
The five safe first moves, writing for whoever is paged at 2am, testing before you rely on them, and automating in stages.
Read the guide -
6 min
Your backups only count once you’ve restored them
Set how much data you can lose and how fast you must recover, keep copies apart from live data, and test restores automatically.
Read the guide -
6 min
Cloud capacity: what to check, not what to forecast
Autoscaling, headroom to lose a zone, cloud quotas, deciding what breaks first, and planning for sudden surges.
Read the guide -
6 min
How much should AI do when production breaks?
The five levels of AI autonomy, the controls to set first, and a better way to judge an AI responder than hours saved.
Read the guide -
Reference
SRE terms in plain English
About 30 terms, from error budgets and burn rates to runbooks, triage and toil, each with a link to the guide that goes deeper.
See the glossary
How we write these
Each guide explains every term it uses. Where a practice has a longer treatment elsewhere, the guide points to it under “Further reading”.
Checking a provider?
Our security and access page lists the questions to ask any provider about access to your production, and how we answer them.
Our sample monthly report shows what we send you each month.