Points to remember

Here are some important points to remember:

• 100% is an unrealistic reliability target.
• Log-based SLIs and ingesting telemetry adds latency.
• App metrics are not good for complex use journeys.
• SLOs must be set based on conversations with engineering and product teams.
• If there is no error budget left, the focus should be on reliability.
• TTD is the time taken to identify that an issue exists or is reported.
• TTR is the time taken to resolve an issue.
• To improve the reliability of a service, reduce TTD, reduce TTR, reduce impact %, and increase TTF/TBF.
• SLIs should have a predictable relationship with user happiness and should be aggregated over time.
• User expectations are strongly tied to past performance.
• Setting values for SLIs and SLOs should be an iterative process.
• Advanced techniques to manage error budgets are dynamic release cadence, setting up error budget exhaustion rates, rainy-day funds, and the use of silver...