Engineering for Reliability
Best practices for Cloud Operations in the enterprise
Twenty-three substantive episodes on SLOs, telemetry, alerting, investigation, and operations
Twenty-three 2021–2022 episodes connect user-facing reliability targets to metrics, logs, traces, and alerts. Examples show instrumentation, Prometheus collection, quota monitoring, GKE investigation, and observability cost controls.
Google Cloud Tech 8 of 23
In this collection Browse 23 summaries 8 of 23
As services spread across teams and projects, shared operational data helps engineers understand dependencies and investigate incidents. This episode recommends centralizing monitoring and logs while separately deciding who can read them and how long they should be kept 00:00–03:11(opens in a new tab).
Key Points Covered
- Make dependency health visible across teams: The presenter gives two reasons for shared data: a dependency's reliability history informs service design, and knowing whether it is currently having an incident helps with response 01:04–03:11(opens in a new tab).
- Keep an accessible but controlled audit trail: Audit logs should answer who did what, where, and when. The recommendation is to test configuration changes in a separate project before production and grant only the permissions each reader needs 02:07–04:13(opens in a new tab).
- Choose retention and access for each use: Compliance records, security investigations, and analytics can have different needs. Log views restrict which records readers see within a bucket; sinks route records to destinations such as BigQuery, Cloud Storage, or Pub/Sub, each with its own access and retention controls 03:11–05:08(opens in a new tab).
Full video: https://www.youtube.com/watch?v=rMOXWc5BUkc(opens in a new tab)