ReadMeFeed / Community knowledge
Real questions. Useful conversations. Find the people who know your stack.
An architecture note on achieving external consistency with Google Cloud Spanner's TrueTime: minimal multi-region design, trust boundaries, operational signals, failure modes, and design-change triggers.
gsutil rsync falls apart at multi-terabyte scale. Storage Transfer Service gives you incremental, scheduled, agent-based on-prem migrations to Cloud Storage — with honest trade-offs on bandwidth, metadata fidelity, and cost.
Learn how to protect internal HTTP services with Cloud IAP, set up OAuth credentials, attach IAP to a Cloud Run backend, and verify access with identity tokens.
Partitioning cuts BigQuery costs for date-filtered queries, but selective filters on tenant or event type need clustering. Here's how to combine both, with a worked example and the trade-offs.
Enable Cloud SQL point-in-time recovery correctly (backups plus transaction logs), clone an instance to a UTC timestamp, verify the clone, and run a drill that proves the window is real.
Learn how to configure Google Cloud Pub/Sub push subscriptions with OIDC tokens so that external webhooks receive authenticated events without managing custom token services.
Goal Enable uniform bucket-level access on an existing Google Cloud Storage bucket while ensuring that no objects become publicly readable and that existing ACL‑based permissions are preserved. Constraints & Uncertainty When uniform bucket‑level access is turned on, ACLs are disabled and only IAM permissions apply. Existing object‑level ACLs that granted
Goal: assess whether existing service account keys that lack an explicit expiration date will be automatically treated as expired once an organization policy enforces a maximum key lifetime, and understand the impact on workloads that continue to use downloadable keys versus those migrated to Workload Identity Federation. Constraints include the version‑sens
When optimizing a low-traffic workload on Cloud Run, setting min-instances to 0 and selecting "CPU allocated only during request processing" is the standard approach to eliminate idle costs. However, this configuration introduces a specific risk during the transition from zero to one active instance. If a container requires a heavy initialization phase—such
google
json
cloudrun