tag: Gcp · 24 items
- Platform/SRE — Learn: AlloyDB AI query functions (ai.extract, ai.classify) are Preview-only, capping relevance at Learn. The Apigee API Health dashboard reaching GA in Cloud Monitoring is worth noting for teams running Apigee proxies, but it adds a monitoring view, not a required migration or operational change.
- CI/CD — Skip
- Leader — Skip
- Signals: GA announcement
- Platform/SRE — Plan: Teams using Cloud SQL Auth Proxy must upgrade to 2.26.3+, Cloud SQL Proxy Operator to 1.8.4+, and Ops Agent to 2.73.0 to address this Go root-certificate vulnerability (GCP-2026-071); managed Cloud SQL instances are auto-patched during scheduled maintenance. CVE is not KEV-listed and EPSS is 0.00, so no active exploitation warrants immediate emergency action, but proxy upgrades are concrete tasks to schedule this sprint.
- CI/CD — Skip
- Leader — Skip
- Signals: CVE-2026-111943 — CISA KEV: not listed, EPSS 0.00
- Platform/SRE — Learn: New console visibility into MIG instance distribution across zones and machine types — useful context for capacity planning and troubleshooting, but no operational change required.
- CI/CD — Skip
- Leader — Skip
- Platform/SRE — Plan: Apigee X instances below 1-18-0-apigee-6 will be automatically updated within 7–21 days (by ~Nov 10). Verify your instances are not excluded due to the DNS misconfiguration in Known Issue 445936920 or use of the deprecated Apigee Java Library (removed Oct 16, 2025); excluded instances will be left on older versions.
- CI/CD — Plan: Cloud SDK 590.0.0 (released Oct 20) removes cloud-build-local and the app-engine-grpc component as breaking changes. Audit pipelines for any use of cloud-build-local for local Cloud Build testing and plan migration to alternatives before auto-upgrading the Cloud SDK to this version.
- Leader — Skip
- Signals: breaking-change flagged
- Platform/SRE — Plan: Two items worth acting on this quarter: enable graceful shutdown on production Compute Engine instances where long-running tasks risk data loss on stop/delete, and review the Cloud Armor WAF v33-stable and v3.0-stable ruleset changes to assess whether updated detection rules alter behavior for existing tuned deployments.
- CI/CD — Skip
- Leader — Skip
- Signals: deprecation/EOL deadline mentioned: October 19, 2026 · GA announcement
- Platform/SRE — Plan: Regex matchers in Application Load Balancer URL maps are now GA, enabling more precise traffic routing via RE2 syntax on paths, headers, and query params — evaluate adopting this for routing rules this quarter.
- CI/CD — Plan: Cloud Build now lets you pin builds to a release channel controlling how often the underlying Debian and Docker versions are updated — review your build configs and private pools to choose an appropriate channel for reproducibility.
- Leader — Skip
- Signals: GA announcement
- Platform/SRE — Plan: Cloud SDK 588.0.0 has already removed the managed-flink-client component; audit any gcloud automation or IaC scripts that invoke it and update them before upgrading the SDK in your environment.
- CI/CD — Plan: If CI runners auto-update gcloud, pipelines using managed-flink-client commands will break on Cloud SDK 588.0.0; audit build scripts now and pin the SDK version or remove the dependency before runners pull this release.
- Leader — Skip
- Signals: GA announcement · breaking-change flagged
- Platform/SRE — Learn: External IPv6 on Standard Tier is Preview-only, so no action yet. The GA Chronicle API Terraform providers are worth noting for teams managing Google SecOps as code — evaluate adoption if SecOps is in your IaC portfolio.
- CI/CD — Skip
- Leader — Skip
- Platform/SRE — Act: The GCP Legacy Monitoring agent has officially reached end of support with all maintenance and bug fixes ceased; audit GCP VM instances for legacy agent usage and migrate to Ops Agent or OpenTelemetry now. A breaking Cloud Logging change also restricts billing account log writes to GCP platform services only — audit any custom logging pipelines writing to billingAccounts/…/logs/… paths and update them.
- CI/CD — Skip
- Leader — Skip
- Signals: deprecation mentioned (no explicit date found)
- Platform/SRE — Learn: AlloyDB Columnar Engine as an in-memory cache for HNSW vector indexes is now GA, offering a potential performance path for vector search workloads on AlloyDB — worth evaluating if vector databases are in the platform stack.
- CI/CD — Skip
- Leader — Skip
- Signals: GA announcement
- Platform/SRE — Act: P100 GPU instances are already shut down as of September 15, 2026 — audit immediately for any affected workloads. T4 and P4 GPUs reach EOS on August 1, 2027; begin migration planning to G2 (L4) or G4 (RTX PRO 6000) machine series before that date.
- CI/CD — Skip
- Leader — Plan: T4 and P4 GPU EOL on August 1, 2027 blocks new 3-year CUD purchases and renewals for those SKUs; review any committed GPU spend and revise infrastructure roadmap to target supported machine series (G2/G4) within this planning cycle.
- Signals: deprecation/EOL deadline mentioned: September 22, 2026; September 15, 2026; August 1, 2027
- Platform/SRE — Plan: Cloud SQL (MySQL and PostgreSQL) now auto-enables PITR after DR switchover/failover asynchronously, reducing RTO — review your DR runbooks to reflect the faster recovery path. The GA Z4D machine series (AMD EPYC Turin, up to 3 TB local storage) is worth evaluating this quarter for storage-heavy workloads like databases, search, or vector stores.
- CI/CD — Skip
- Leader — Learn: The Z4D GA instances and Cloud SQL PITR improvements are worth noting for teams standardizing on GCP for storage-heavy or analytics workloads, but neither forces a strategic decision or pricing change now.
- Signals: GA announcement
- Platform/SRE — Learn: Filestore small-capacity Regional tier reaching GA is worth noting for dev/test storage workloads; the auto-protection backup policy feature is Preview only, capping it at Learn.
- CI/CD — Skip
- Leader — Skip
- Signals: GA announcement
- Platform/SRE — Learn: The Batch workaround for job failures with outdated-kernel custom VM OS images is worth noting if you run GCP Batch with custom images; Conda support in Artifact Registry is pre-GA Preview, so track it but no action yet.
- CI/CD — Skip
- Leader — Skip
- Platform/SRE — Plan: The gcloud minikube component will be removed after January 31, 2027 — schedule migration to a standalone minikube install or alternative local-cluster tooling before that date. Cloud Monitoring’s per-chart time-range override is now GA and worth adopting for long-duration observability dashboards alongside high-frequency charts.
- CI/CD — Plan: If any pipelines invoke gcloud’s minikube component for integration-test clusters, plan migration to a standalone minikube binary before the January 31, 2027 removal; also audit scripts for the already-deprecated
gcloud storage buckets anywhere-caches pausecommand and replace it in affected workflows. - Leader — Skip
- Signals: deprecation/EOL deadline mentioned: 2026-09-09; January 31, 2027 · GA announcement · breaking-change flagged
- Platform/SRE — Plan: If your org uses GCP Compute Engine reservations across projects, you can now convert between single-project and shared reservations without recreating them — worth evaluating for multi-team capacity-sharing workflows this quarter.
- CI/CD — Skip
- Leader — Skip
- Signals: GA announcement
- Platform/SRE — Learn: Preview-stage feature that could simplify managing perimeters containing stale or deleted identities; worth evaluating if your GCP environment hits the invalid-email error today, but not actionable until GA.
- CI/CD — Skip
- Leader — Skip
- Platform/SRE — Skip
- CI/CD — Learn: Cloud Build now offers a UI path to rotate expired access tokens for 2nd-gen Bitbucket and GitLab host connections, useful if those credentials are aging. The Application Integration authorization change—scheduled/event-triggered runs will require an explicit run-as service account—could affect event-driven release workflows, but no enforcement deadline is given and the product is outside the standard CI/CD toolchain for most teams.
- Leader — Skip
- Platform/SRE — Plan: The M4N machine series is now GA on GCP, offering up to 400 Gbps network and 1M IOPS for memory/network-intensive workloads like vector databases and RAG layers — evaluate whether it fits high-memory workload placements this quarter. CVE-2026-12710 in Application Integration was already patched server-side on April 4, 2026; no customer action required.
- CI/CD — Skip
- Leader — Learn: GCP’s M4N instance family (GA) targets high-memory AI infrastructure workloads such as vector databases and in-memory RAG layers — relevant context for future GCP AI/ML platform architecture discussions, but no immediate strategic or budget decision is forced.
- Signals: GA announcement · CVE-2026-12710 — CISA KEV: not listed, EPSS n/a
- Platform/SRE — Plan: App Engine Images→Cloud Run migration support is now GA for Java and Python — schedule evaluation this quarter if you operate App Engine standard workloads. Cloud SDK 581.0.0 flags a breaking change (removal of
gcloud beta services mcpcommands), but those were already no-ops so real pipeline impact is minimal; worth verifying before upgrading the SDK. - CI/CD — Skip
- Leader — Learn: The GA availability of the App Engine→Cloud Run migration path is a strategic signal if the org still runs App Engine workloads; no forced deadline, but it clarifies the long-term migration route Google is offering.
- Signals: GA announcement · breaking-change flagged
- Platform/SRE — Plan: Cloud SDK 578.0.0 makes –auto-commit the default for gcloud database-migration seed/convert/import-rules commands; audit any automation or runbooks that call these operations and add –no-auto-commit explicitly before upgrading the SDK.
- CI/CD — Plan: If pipelines invoke gcloud database-migration commands, upgrading to Cloud SDK 578.0.0 silently changes commit behavior; pin the SDK version or add –no-auto-commit flags before the next runner/image update pulls this version in.
- Leader — Skip
- Signals: breaking-change flagged
- Platform/SRE — Act: GCP Batch will reject jobs whose
allowedLocations[]field lists regions or zones outside the job’s own location; the deadline is July 31, 2026 for most projects (June 30, 2027 for projects that already submitted a cross-region job before that date). Audit all Batch workloads for cross-regionallowedLocations[]entries and restrict them to the job’s location before July 31, 2026. - CI/CD — Skip
- Leader — Skip
- Platform/SRE — Plan: gcloud SDK 576.0.0 changes
gcloud storage rsyncto decompress gzip downloads by default, which can silently break any operational scripts relying on the previous behavior; audit usage and add--do-not-decompressor pin SDK version before upgrading. The bundled Python update for CVE-2026-34182 (EPSS 0.00, not KEV) adds low-urgency motivation to upgrade. - CI/CD — Plan: If pipelines invoke
gcloud storage rsyncto pull artifacts, the 576.0.0 default-decompress behavior change will alter what lands in the workspace; pin the gcloud SDK version or add--do-not-decompressbefore rolling out the upgrade across runners. - Leader — Learn: BigQuery conversational analytics now carries HIPAA compliance support, broadening Gemini-in-BigQuery eligibility for regulated-industry workloads — worth factoring into GCP data platform strategy for healthcare or other compliance-sensitive verticals.
- Signals: GA announcement · breaking-change flagged · CVE-2026-34182 — CISA KEV: not listed, EPSS 0.00
- Platform/SRE — Learn: Advanced Compute Images (Preview) offer pre-tuned AI/ML/HPC OS images with drivers and Slurm pre-installed, potentially simplifying GPU node provisioning; separately, BigQuery hybrid search has been temporarily disabled — check if any workloads depend on the VECTOR_SEARCH hybrid mode.
- CI/CD — Skip
- Leader — Skip