Maturity Model for SLOs

Transcript

3. Agenda SLO Maturity Model * Cloud Native * Numbers * Others Journeys Inspiration Next: Framework * Two axis: * Implementation * Adoption * Our story. * How to start. * Automate the framework.
4. What says the numbers? Where are you at?
5. Numbers SRE Adoption according Gartner www.yurynino.com
6. Numbers Respondents working with SLOs www.yurynino.com
7. Numbers Large companies have more experience using SLOs www.yurynino.com
8. Numbers How many services have SLOs www.yurynino.com
9. How can WE start?
10. Cloud Native www.yurynino.com
11. Learning from Others
12. Cloud Native www.yurynino.com Schulumberger’s SLO Journey Schlumberger engaged the Google CRE team to learn SRE. Defining SLOs and how to measure them. Implementing reports and insights on SLIs using dashboards. LEARN DEFINE Iterating its existing SLOs to improve them. Instrumentation, collection, monitoring, and alerting of SLIs. Holding workshops for introducing and educating developers. IMPLEMENT ADAPT ALERTING EVANGELIZE
13. SLOS Maturity Model High Implementation Initial High Adoption SLOs Champion Master Padawan Knight Knight www.yurynino.com
14. Moving through to Adoption
15. Appropriation Initial Investment Expectation Adoption
16. Initial Adoption
17. • No system is covered by SLOs. • There is low or no organizational awareness. • But organizations decide to adopt SLOs. • Early adopters infrequently perform SLOs. Initial www.yurynino.com
18. Investment Adoption
19. • Establishing SLOs is officially sanctioned. • Resources are dedicated to the SLOs. • Multiple teams are interested and engaged. • A few critical services are supported with SLOs. Investment www.yurynino.com
20. Appropriation Adoption
21. • SLOs are incorporate into the shared responsibility. • SLOs are part of new features and fixes backlog. • Observability dashboards report insights into how SREs identify and build reliability. • SLOs help in improving the reliability of the services. Appropriation www.yurynino.com
22. Expectation Adoption
23. • All critical and noncritical services are frequently supported with SLOs. • SLOs, SLIs, and error budgets provide a data-driven change management strategy. • Evangelizing about SLIs, SLOs and SLAs. • SLOs and SLIs are part of the engineers onboarding process. Expectation www.yurynino.com
24. Moving through Implementation
25. Chaos Maturity Model Implementation Initial Usage Automated Evolution
26. Implementation Initial
27. Initial • Organizations decide to adopt SLOs. • Choosing some basic SLIs to identify SLOs. • The process is administered manually. • SLOs are not supporting services in production. • An observability tool could no be a good idea here. www.yurynino.com
28. Implementation Usage
29. Usage • SREs use SLOs, SLIs, and error budgets to mitigate outages and maintaining quality to customers. • SLO and SLI data will become much more meaningful and rich. • SLIs indicate how well certain features penetrate the market. • You are using an observability tool and take advantage to make decisions. www.yurynino.com
30. Implementation Automated
31. • Tooling supports available observability tools. • Setup, provisioning dashboards and result analysis are automated. • IaC and GitOps are used here. • SLOs framework is integrated with continuous delivery. • Results are tracked over time. Automated www.yurynino.com
32. Implementation Evolution
33. • Development teams correlate SLOs insights with new feature releases. • Apps can autohealing with the support of SLOs. • Results are tracked over time. • SLO results are compared between releases and control groups. Evolution www.yurynino.com
34. Our story
35. Our status www.yurynino.com NewRelic/Datadog dashboards Templates by service
36. @yurynino www.yurynino.com