platformchangelog.dev
43 tools

Platform Changelog: what shipped in platform engineering, developer experience and DevOps

feedbefore 2026-10-08patches hidden unless security or breaking
newest entries
2026-10-0625 days ago
AWS Batch now publishes job metrics to Amazon CloudWatchannouncementAWS Batch now publishes job state transition and duration metrics to the AWS/Batch namespace in Amazon CloudWatch, with a JobQueueName dimension. The metrics cover job counts entering states such as submitted, running, succeeded, and failed, plus time spent between states. The metrics are available in all AWS Regions where AWS Batch runs.platform teamsPlatform teams running AWS Batch can track queue health, failure rates, and job durations in CloudWatch, the console, CLI, or SDKs. The source does not describe pricing, retention, or limits for these metrics.What's New with AWS ↗written by claude-haiku-5-5, prompt 7cbe154, checked by publish rules
2026-10-0229 days ago
Amazon EKS and Amazon EKS Distro now support Kubernetes version 1.37announcementAWS announced that Amazon EKS and Amazon EKS Distro now support Kubernetes version 1.37 in all EKS regions, including the GovCloud (US) Regions. Clusters can be created on 1.37 or upgraded through the EKS console, eksctl, or infrastructure-as-code tools. The source highlights Metrics API promotion to GA, DRA device taints and tolerations reaching GA, and Horizontal Pod Autoscaler scale-to-zero entering beta and enabled by default.platform teamsPlatform teams running EKS can plan upgrades to 1.37 and can use EKS cluster insights to check for issues before upgrading. Autoscalers with minReplicas: 0 using object or external metrics will scale idle workloads to zero by default, which teams should account for in their scaling policies.What's New with AWS ↗Amazon EKSwritten by claude-haiku-5-5, prompt 7cbe154, checked by publish rules
Integrate AWS DevOps Agent with third-party tools using Amazon EventBridgepostAWS published a walkthrough that uses AWS CDK, Amazon EventBridge, AWS Lambda, and DynamoDB to send AWS DevOps Agent investigation events to Jira Cloud. The Lambda function creates a Jira issue on Investigation Created, stores the mapping from task_id to issue key, and appends comments for later state changes. Credentials are kept in AWS Secrets Manager, and the post says the same pattern applies to ServiceNow and PagerDuty.platform teamsA platform team that runs AWS DevOps Agent can follow investigations in its existing ticketing tool by adding an EventBridge rule and a Lambda target, without changing the agent. The destination-specific logic lives in the Lambda code, so swapping tools mainly means rewriting that function.AWS DevOps Blog ↗AWS DevOps AgentAmazon EKSwritten by claude-haiku-5-5, prompt 7cbe154, checked by publish rules
Closed-loop incident response: connect AWS DevOps Agent to OpenSearchpostThe AWS DevOps Blog describes connecting AWS DevOps Agent to Amazon OpenSearch Service data through an MCP server, so that OpenSearch alerts trigger the agent to query logs and traces and correlate them with CloudTrail and CloudWatch. The post presents three hosting options: self-managed on ECS Fargate behind an NLB with VPC Lattice, Amazon Bedrock AgentCore, and the built-in MCP endpoint on OpenSearch 3.3+. The source text is cut off before the registration and verification steps are complete.platform teamsPlatform teams running OpenSearch alerting could route alerts to an agent that investigates automatically, but they would need to set up a hosted MCP server, a private network connection, and FGAC role mappings that control what the agent can read. The source describes this as a walkthrough and does not report measured results.AWS DevOps Blog ↗AWS DevOps AgentAmazon EKSwritten by claude-haiku-5-5, prompt 7cbe154, checked by publish rules
2026-09-301 month ago
Running production experiments with AWS AppConfig experimentationpostAWS describes AWS AppConfig experimentation, which adds A/B testing on top of AWS AppConfig feature flags so a change can be exposed to a controlled slice of production traffic. The AWS AppConfig Agent assigns each participant to a control or treatment, logs each assignment to stderr for shipping to CloudWatch Logs and Amazon S3, and leaves metric analysis to the customer's existing data warehouse. The post walks through a frontend Add to cart button experiment and a backend cache TTL experiment on Amazon ECS.platform teamsPlatform teams can run production A/B tests through existing AppConfig and Agent setups, but they must provision assignment logging, a warehouse for analysis, and CloudWatch alarms that stop runs. Entity IDs are logged verbatim, so PII must be hashed or pseudonymized.AWS DevOps Blog ↗Amazon EKSwritten by claude-haiku-5-5, prompt 7cbe154, checked by publish rules
2026-09-281 month ago
How Property Finder automated incident management with AWS DevOps Agentcase studyProperty Finder describes a pipeline built on AWS DevOps Agent that takes a CloudWatch ECS CPU alarm through root cause analysis, Slack posting, Jira ticket creation, phone escalation via Grafana IRM, and a Draft GitHub PR with a Terraform fix. The post reports a 14-minute end-to-end time for one incident, with investigation taking about 10 minutes. The post is written by AWS and the figures come from a single incident and the company's own account.platform teamsPlatform teams running ECS and Terraform can see a pattern for wiring an AI investigation agent to alerting, ticketing and paging, with a remediation agent restricted to a GitHub MCP server and always opening Draft PRs for human review. The source gives no independent measurements beyond the one incident and a vendor-attributed quote.AWS DevOps Blog ↗AWS DevOps AgentAmazon EKSTerraform MCP Serverwritten by claude-haiku-5-5, prompt 7cbe154, checked by publish rules
Audit trails for autonomous agents with AWS DevOps AgentpostAWS describes an audit pipeline for AWS DevOps Agent built on Amazon EventBridge, AWS Lambda, Amazon S3, Glue and Athena. The pipeline archives the agent's investigation journals, polls recommendations on a schedule, and correlates recommended fixes with resource changes recorded in AWS Config and CloudTrail. The post says the journal is immutable, and it points to sample code in a repository.platform teamsPlatform teams running DevOps Agent can capture per-investigation journals, recommendation versions, and lifecycle events into their own archive, and scope IAM access to that archive. Recommendations have no event, so they must be polled, and CloudTrail does not capture the agent's read queries.AWS DevOps Blog ↗AWS DevOps AgentAmazon EKSwritten by claude-haiku-5-5, prompt 7cbe154, checked by publish rules
Fix pod distribution drift in Amazon EKS with the Kubernetes deschedulerpostThe AWS Containers Blog describes how pods spread across three Availability Zones with soft topology spread constraints can drift into two zones after a node-availability gap and stay skewed. In a measured run on Amazon EKS with a 1,000-pod fleet, the post reports the skew persisted for over 16 hours after capacity returned. It then shows the Kubernetes descheduler's RemovePodsViolatingTopologySpreadConstraint plugin, configured to act on ScheduleAnyway constraints, restoring balance in about 183 evictions with PDB limits applied.platform teamsPlatform teams running soft topology spread on EKS should expect silent zone drift after node disruptions and can use the descheduler with ScheduleAnyway enabled, topologyBalanceNodeFit, and PDBs to correct it. Teams needing strict zone balance should consider DoNotSchedule and alert on pending pods instead.AWS Containers Blog ↗Amazon EKSwritten by claude-haiku-5-5, prompt 7cbe154, checked by publish rules
2026-09-231 month ago
Building a single-pane NOC dashboard for Amazon EKS with Amazon CloudWatchpostAWS Containers Blog describes a sample NOC dashboard for Amazon EKS built on Amazon CloudWatch, combining PromQL chart widgets with CloudWatch metric widgets. The sample comprises a 35-widget NOC and three drilldown dashboards for the retail store sample application, with a probe that validates each query against live telemetry before deployment. The post was validated with the Amazon CloudWatch Observability EKS add-on v6.4.0-eksbuild.1.platform teamsPlatform teams running EKS can reuse the pattern and the validation probe to check that dashboard queries resolve, and can add freshness alarms to separate missing telemetry from healthy zero counts. Auto Mode scale-in can produce blank widgets, so teams need the freshness alarms to tell expected scale-in from pipeline failure.AWS Containers Blog ↗Amazon EKSwritten by claude-haiku-5-5, prompt 7cbe154, checked by publish rules
2026-09-111 month ago
Kubernetes v1.37: Native Histograms Graduates to BetapostKubernetes v1.37 graduates native histogram support for Kubernetes metrics to Beta and enables it by default. Components expose classic histogram buckets alongside Prometheus native histogram spans in the same Protobuf payload, using a bucket factor of 1.1 and a cap of 160 buckets. Prometheus 3.x users can enable scraping per job with scrape_native_histograms and keep classic series with always_scrape_classic_histograms.platform teamsPlatform teams running Prometheus should plan to migrate dashboards and alerts from _bucket queries to native histogram queries while keeping always_scrape_classic_histograms true during transition. Rollback is available by setting scrape_native_histograms to false or disabling the feature gate.Kubernetes Blog ↗written by claude-haiku-5-5, prompt 7cbe154, checked by publish rules
2026-09-091 month ago
Kubernetes v1.37: Introducing Node Lifecycle ConditionspostKubernetes v1.37 reserves five well-known Node condition types: DrainInProgress, Drained, MaintenancePlanned, MaintenanceInProgress, and GracefulNodeShutdownInProgress. It also adds an Alpha NodeLifecycleConditions feature gate, disabled by default, which the release describes as a no-op in v1.37 since no core component reads the conditions yet. Administrators or authorized controllers are responsible for setting and clearing them.platform teamsPlatform teams can start publishing these conditions today from drain and maintenance automation to give dashboards, alerts and tooling a common status signal. No core controller changes behavior based on them in this release, so scheduling and eviction still rely on existing mechanisms.Kubernetes Blog ↗editorCorrection 2026-10-20: this was listed as a Amazon EKS v1.37 release by mistake. It is Kubernetes Blog coverage, not release notes from the project.written by claude-haiku-5-5, prompt 7cbe154, checked by publish rules
2026-09-041 month ago
Investigate DMS migration issues with AWS DevOps AgentpostAWS published a guide on extending AWS DevOps Agent with a sample MCP server that provides read-only, DMS-specific tools and 46 runbooks for investigating database migration issues. The server is deployed on AWS Lambda behind an IAM-authenticated function URL and registered with a DevOps Agent Space. The post describes five investigation scenarios run against a test migration from RDS MySQL to Aurora PostgreSQL.platform teamsPlatform teams running DMS migrations can evaluate a read-only MCP pattern for giving an agent migration-specific context, with tools restricted to Describe, Get, List, Lookup and TestConnection calls. Timing claims in the post come from the authors' own test runs and are not benchmarked.AWS DevOps Blog ↗AWS DevOps AgentAmazon EKSwritten by claude-haiku-5-5, prompt 7cbe154, checked by publish rules
2026-09-021 month ago
Kubernetes v1.37: Scale Workloads to Zero with HorizontalPodAutoscalerpostKubernetes v1.37 moves horizontal autoscaling to zero replicas to Beta and enables it by default. An HPA using an object or external metric can now scale a workload to zero and bring it back when the metric changes, a capability that previously required an add-on or the Alpha feature gate. The blog notes that Services do not buffer requests, so request-driven workloads need a separate buffering layer.platform teamsPlatform teams can scale idle queue consumers and batch workers to zero to cut costs on expensive CPU or GPU Pods, but must accept cold-start latency and make sure the external metrics adapter is reliable. Before upgrading, control plane components must both have the feature enabled, and HPAs with minReplicas 0 need an object or external metric.Kubernetes Blog ↗editorCorrection 2026-10-20: this was listed as a Amazon EKS v1.37 release by mistake. It is Kubernetes Blog coverage, not release notes from the project.written by claude-haiku-5-5, prompt 7cbe154, checked by publish rules
2026-09-012 months ago
Kubernetes v1.37: etcd RangeStream Cuts Memory Use on Large List ReadspostThe Kubernetes blog says etcd RangeStream is graduating to beta in Kubernetes v1.37 and pairs with etcd v3.7. RangeStream is a streaming version of the etcd Range read that sends results in chunks sized by bytes, so the API server and etcd hold less memory when reading large collections such as Pods. The EtcdRangeStream feature gate is on by default in v1.37, and the API server falls back to paginated Range reads if etcd is older than v3.7.platform teamsPlatform teams running clusters with many or large objects should see lower and more predictable kube-apiserver and etcd memory use during watch cache initialization and direct list fallbacks, provided etcd is upgraded to v3.7 or later. Teams can confirm use via the listStream operation label on etcd request duration metrics, or disable the gate with --feature-gates=EtcdRangeStream=false.Kubernetes Blog ↗written by claude-haiku-5-5, prompt 7cbe154, checked by publish rules
2026-08-312 months ago
Optimize EKS operations with agents: Reduce MTTR with AWS DevOps Agent and a Kubernetes OperatorpostAWS describes a Kubernetes Operator that watches EKS pods, collects logs, events, node-level data and manifests when a failure occurs, and sends that context to AWS DevOps Agent through a generic webhook. The post says the Operator stores evidence in S3 or CloudWatch Logs before pods are rescheduled or deleted. Setup requires an Agent Space, an IAM role with Pod Identity, and building the Operator image from source because no prebuilt image is provided.platform teamsPlatform teams running EKS can evaluate automated incident evidence capture and agent-driven investigation, but must set up IAM, Systems Manager registration on nodes, and build and host their own image. Node-level collection is unavailable on Fargate.AWS DevOps Blog ↗Amazon EKSAWS DevOps AgentK8sGPTwritten by claude-haiku-5-5, prompt 7cbe154, checked by publish rules
Kubernetes v1.37: Metrics API graduates to stablepostKubernetes v1.37 promotes the metrics.k8s.io API to stable v1. The v1 resource types and fields are identical to v1beta1, with NodeMetrics and PodMetrics unchanged. kubectl top prefers v1 and falls back to v1beta1, while the HPA controller still supports only v1beta1.platform teamsPlatform teams get stability guarantees for the API behind kubectl top and resource-metrics autoscaling, but no data changes. Implementations such as metrics-server should serve both v1 and v1beta1 during the transition, and HPA still needs v1beta1.Kubernetes Blog ↗written by claude-haiku-5-5, prompt 7cbe154, checked by publish rules
2026-08-262 months ago
AI-driven software delivery with Kiro, AWS DevOps Agent and Bluebox by DynatracepostAWS and Dynatrace describe a workflow in which Kiro retrieves runtime topology and traffic data from Bluebox via CLI before generating a pull request, and AWS DevOps Agent investigates a simulated traffic-spike incident. The agent's findings and mitigation plan are filed as a GitHub issue, and Kiro then opens a remediation pull request that a human reviews and merges. The post uses a travel-booking example, and the reported results come from that example rather than measured data.platform teamsPlatform teams evaluating AI incident agents can see a pattern where investigation output is written to GitHub issues and fixes flow through normal PR review and CI/CD. The post does not give benchmarks, adoption figures, or availability details for the release management capability, which is in Preview.AWS DevOps Blog ↗AWS DevOps Agentwritten by claude-haiku-5-5, prompt 7cbe154, checked by publish rules
2026-08-173 months ago
Streamline your GitHub journey with AWS CodePipeline and AWS DevOps AgentpostAWS published a walkthrough for connecting AWS DevOps Agent to GitHub repositories, AWS CodePipeline, CloudWatch alarms and a Lambda webhook. When a pipeline stage fails, a CloudWatch alarm or EventBridge rule triggers the agent, which investigates logs and recent commits. The post shows two sample failures, a bad DynamoDB table name and a misspelled npm dependency, where the agent identified the offending change.platform teamsPlatform teams running CodePipeline with GitHub sources can evaluate an agent that starts investigations from pipeline alarms and proposes remediation. The setup requires IAM roles, a webhook stored in Secrets Manager, and a Lambda bridge, and the post reports no measured results beyond the sample app.AWS DevOps Blog ↗AWS DevOps AgentGitHub ActionsAmazon EKSwritten by claude-haiku-5-5, prompt 7cbe154, checked by publish rules
2026-08-103 months ago
Automate SageMaker HyperPod incident triage and root-cause-analysis with AWS DevOps AgentpostAWS describes a solution that connects a SageMaker HyperPod cluster to AWS DevOps Agent. EventBridge events and a periodic Kubernetes-state audit feed the agent, which triages and root-causes conditions that still need a human decision and sends a verdict email. The agent is configured read-only, with no SSM, SSH, or action-taking permissions, and the solution deploys as one CloudFormation stack per cluster.platform teamsPlatform teams running HyperPod on EKS or Slurm can deploy an observe-and-report incident agent alongside existing self-healing, with custom skills that can be edited for their own conditions. Teams should check the read-only boundary and the cost of polling and investigations against their own operating model.AWS DevOps Blog ↗AWS DevOps AgentAmazon EKSwritten by claude-haiku-5-5, prompt 7cbe154, checked by publish rules
2026-07-233 months ago
PlatformEngineering.org: Self-service observability: Consistency, cost control, and scale with OpenTelemetrypostA platform team describes moving from having each team set up its own observability to offering self-service observability. They embedded OpenTelemetry into a NestJS Platform SDK and applied a four-pillar policy approach for consistency, cost guardrails and scale across more than 90 internal teams. The article, as published in the source abstract, reports a 4x TPS improvement, but the abstract gives no method or measurement details.platform teamsPlatform teams evaluating SDK-level instrumentation can see one approach to standardizing telemetry and enforcing cost limits without per-team setup. The source gives no detail on the policies or how the TPS figure was measured, so the result is hard to reproduce from this text alone.PlatformEngineering.org Blog ↗written by claude-haiku-5-5, prompt 7cbe154, checked by publish rules
2026-07-104 months ago
K8sGPTv0.4.36patchsecurity****!
daily email

One short email each morning: what shipped, what breaks, and a link to each change. Unsubscribe in one click. Privacy.

statuslast 30 days for lines
tools tracked43releases in registry964released in the last 30 days106entries published335breaking change lines18security lines84
change typestwo-column gutter, like git status
  • +addedNew functionality.
  • ~changedExisting behavior changed.
  • -deprecatedStill works, scheduled for removal.
  • -removedDecommissioned, gone in this version.
  • *fixedA defect corrected.
  • !securityA vulnerability closed or a dependency patched for one.
  • !breakingFlag in the first column; the change needs action before upgrading.
filter
categoryallidp & portalsai agents on platformci/cdobservabilitysupply chaincost & governancereports & dx dataenterprise app platforms
showmajor onlyreleasescase studiespostmortemsinclude patches