Engineering for Reliability
Best practices for Cloud Logging
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 6 of 23
In this collection Browse 23 summaries 6 of 23
Useful logs need to be easy to query, kept for the right amount of time, and accessible to the right people. This episode explains how structured records and centrally stored logs support those goals in Cloud Logging 00:00–03:07(opens in a new tab).
Key Points Covered
- Structure makes records queryable: JSON fields can be filtered directly instead of requiring regular expressions over text payloads 00:00–02:04(opens in a new tab).
- Route related logs to shared storage: Log buckets hold records; sinks route records to a destination. The presenter recommends using sinks to collect logs from multiple projects into a shared bucket, with audit logs as an example. Retention and location should follow the business's requirements 02:04–03:07(opens in a new tab).
- Central storage need not mean identical access: Identity and Access Management (IAM) controls permissions, while log views narrow which records a user can see within a bucket. The example gives one user access to a single project's logs and another access to all projects' logs in the same central store 03:07–04:09(opens in a new tab).
Full video: https://www.youtube.com/watch?v=hh_z8qEwGfQ(opens in a new tab)