OffNet Newsroom

Daily topic roundup

AWS

Wednesday, July 15, 2026 · 8 stories, curated & summarized — click any story for the source.

AWS has extended Amazon GuardDuty to monitor AI-specific workloads in Bedrock and SageMaker. The service analyzes CloudTrail events to identify threats like prompt injection and anomalous model invocations. It also detects cost harvesting attacks where actors exhaust GPU time and tokens. This provides visibility into AI security without requiring custom tooling or manual configuration.

  • GuardDuty now covers Bedrock and SageMaker for AI threat detection
  • Detects prompt injection, anomalous invocations, and cost harvesting
  • Analyzes CloudTrail management and data events automatically
  • Reduces need for custom security tooling for AI workloads
CHECKLISTGuardDuty AI Threat DetectionMonitor Bedrock and SageMaker workloadsDetect prompt injection attacksIdentify anomalous model invocationsSpot GPU cost harvestingAnalyze CloudTrail events automatically

AWS Security Hub now includes an AI inventory feature to provide centralized security teams with a continuous, organization-wide view of AI assets and their security posture. This capability addresses the visibility gap that arises as organizations rapidly deploy AI agents, models, and pipelines. The feature automatically discovers and catalogs AI workloads across the AWS environment to help secure assets that might otherwise go unnoticed.

  • Centralized visibility into AI assets eliminates blind spots in security posture.
  • Automatically discovers AI workloads without requiring additional configuration.
  • Catalogs managed AI services via AWS Config resources from Amazon Bedrock.
  • Includes Bedrock AgentCore and Amazon SageMaker in the inventory scope.
  • Connects AI assets to active threats and misconfigurations for better defense.
CHECKLISTSecure AI Assets NowGain organization-wide visibility into AI assetsAutomatically discover AI workloads without extra configCatalog services via AWS Config resourcesConnect assets to active security threats

Amazon Redshift Serverless now automatically maintains zero-ETL and Amazon S3 event integrations when restoring a namespace from a snapshot or recovery point. Previously, these integrations were marked as failed during restore, forcing manual recreation and risking data ingestion gaps. This update ensures integrations resume operation immediately after restore, simplifying disaster recovery and testing workflows.

  • Snapshot restores now keep zero-ETL and S3 event integrations active without manual reconfiguration.
  • Eliminates data ingestion gaps and administrative overhead during disaster recovery or testing.
  • Reduces risk of errors associated with rebuilding complex data pipelines post-restore.
HOW IT WORKSSeamless Snapshot Restore Flow1Create namespace snapshot2Restore from recovery point3Integrations resume automatically4Zero manual reconfiguration

Google Cloud's VMware service has suffered a resilience degradation triggered by a faulty update. This issue coincides with VMware's own warning regarding a critical flaw in its load balancer components. The incident highlights the fragility of managed VMware offerings when underlying software patches introduce regressions.

  • Google Cloud VMware service resilience is currently compromised due to a recent update.
  • The outage aligns with VMware's disclosure of a critical load balancer vulnerability.
  • Practitioners should monitor Google Cloud status pages for VMware service health.
  • Verify load balancer configurations if running on affected VMware infrastructure.
  • Treat pending VMware patches with caution until resilience is restored.

Amazon Managed Service for Apache Flink now integrates AI Agent Skills that provide coding assistants with expert, up-to-date guidance for building and operating Flink applications. These skills cover common tasks like troubleshooting, scaling, monitoring, and version upgrades, transforming specialized knowledge into a guided experience. This allows developers to independently manage application health and accelerate development using existing tools like Kiro.

  • AI coding assistants now access expert Flink guidance for tasks like scaling and monitoring
  • Simplifies upgrades to latest Apache Flink versions like 2.2 without deep expertise
  • Reduces dependency on specialized knowledge by turning complex ops into guided workflows
  • Compatible with existing AI agents like Kiro to boost developer productivity
HOW IT WORKSFlink AI Skill Workflow1Developer invokes AI agent2Agent accesses Flink skills3Guided scaling or monitoring4Version upgrade to 2.25Independent app management

New York has become the first state to pause construction of data centers exceeding 50 megawatts while regulators draft new rules. The moratorium aims to address environmental concerns and shield ratepayers from the financial and infrastructural strain of massive power demands. This move signals a tightening regulatory landscape for large-scale AI and cloud infrastructure projects in the region.

  • New York pauses all datacenter builds over 50MW pending new regulatory framework
  • Focus is on protecting the electrical grid and reducing costs for ratepayers
  • Environmental impact assessments will likely become stricter for large deployments
  • Other states may follow suit as power constraints become a critical bottleneck
  • Project timelines for major hyperscaler expansions in NY are now uncertain
AWS Database Blog awsdatabase ↺ since 07-14

Aurora PostgreSQL: Rebuild large indexes via Blue/Green and NVMe

AWS describes using Blue/Green Deployments to rebuild large indexes on Aurora PostgreSQL without impacting production. The process performs the reindex on the green staging environment using NVMe-backed instances, allowing the sort phase to leverage fast local storage instead of networked EBS. This approach isolates the heavy I/O and compute costs from the active blue production cluster.

  • Isolate expensive reindexing workloads by targeting the green staging environment exclusively.
  • Select NVMe-backed instance classes to utilize local storage for the sort phase, reducing network overhead.
  • Use Aurora Optimized Reads to enhance performance during the index rebuild process.
  • Avoid production latency spikes by keeping the heavy I/O off the active blue cluster.

Fleet impact: For DBAs managing Aurora PostgreSQL, this offers a zero-downtime path to handle massive index maintenance. Verify that your NVMe instance classes have sufficient local storage capacity for the sort operations, and plan your Blue/Green deployment cycles to absorb the staging rebuild time before cutover.

InfoQ generaldevops ↺ since 07-14

Eliminating Latency in Multi-Region AWS APIs by Dropping Discovery Calls

Regional outages exposed a hidden latency bottleneck in a multi-region AWS API architecture. The root cause was a pre-flight discovery call embedded in every client session, originally implemented as the only available option at the time. The engineering team successfully removed this round trip to enable viable global failover, detailing the implementation steps and rollout costs.

  • Pre-flight discovery calls add unavoidable latency, hindering rapid global failover.
  • Legacy design choices can become critical bottlenecks during regional outages.
  • Removing hidden round trips is essential for low-latency multi-region architectures.
  • The team documented the specific costs and steps required to execute this change.