AWS (Amazon Web Services) favicon

AWS outage on 2026-11-06

amazon.com
November 2026 6 Minor incident

AWS's US-WEST-2 (Oregon) region had elevated EC2 launch failures for 2 hours and 23 minutes on 6 November 2026. The fault sat in a subsystem supporting EBS volumes needed at instance launch, and it was AWS's problem, not anything in your own stack.

Started

22:51 UTC

Duration

Lasted 2h 23m

Source

IsDown

Next time

AWS (Amazon Web Services) is degraded. Is your product working?

An incident on their side does not always mean an outage on yours, and the only way to know is to already be watching your own endpoints. Set that up in under a minute and you will be notified the next time your own site is affected. And there will be a next time.

50 monitors for your own websites. No card.

Affected components: VM Import/Export, VM Import/Export (Oregon), Amazon Elastic Compute Cloud, Amazon Elastic Compute Cloud (Oregon), Amazon Elastic Load Balancing, Amazon Elastic Load Balancing (Oregon), Amazon Elastic MapReduce, Amazon Elastic MapReduce (Oregon), Amazon EMR Serverless, Amazon EMR Serverless (Oregon), AWS Batch, AWS Batch (Oregon), Amazon Relational Database Service, Amazon Relational Database Service (Oregon), Amazon WorkSpaces, Amazon WorkSpaces (Oregon), AWS DataSync, AWS DataSync (Oregon), AWS Glue, AWS Glue (Oregon), Amazon EventBridge, Amazon EventBridge (Oregon), AWS CodeBuild, AWS CodeBuild (Oregon), AWS Elastic Beanstalk, AWS Elastic Beanstalk (Oregon), AWS Elemental, AWS Elemental (Oregon), AWS Step Functions, AWS Step Functions (Oregon), Amazon CloudWatch, Amazon CloudWatch (Oregon), AWS CodePipeline, AWS CodePipeline (Oregon), Amazon DocumentDB, Amazon DocumentDB (Oregon), AWS HealthImaging, AWS HealthImaging (Oregon), Amazon Elastic Container Service, Amazon Elastic Container Service (Oregon), Amazon SageMaker, Amazon SageMaker (Oregon), AWS CloudShell, AWS CloudShell (Oregon), AWS Device Farm, AWS Device Farm (Oregon), AWS Fargate, AWS Fargate (Oregon), AWS Backup, AWS Backup (Oregon), AWS HealthOmics, AWS HealthOmics (Oregon), Amazon Managed Workflows for Apache Airflow, Amazon Managed Workflows for Apache Airflow (Oregon), Amazon MQ, Amazon MQ (Oregon)

What happened?

AWS reported the first signs of trouble at 22:51 UTC, describing elevated launch failures in US-WEST-2. By 23:19 UTC it had traced the cause to the EBS service, warning that anything depending on EBS or EC2 would see failure rates of its own. At 00:08 UTC AWS confirmed EC2 instance launch failures across all four availability zones in the region, usw2-az1 through usw2-az4, and pinned the issue to a subsystem that supports EBS volumes required during instance launch. Services that create new instances as part of their normal operation were caught too, including cross-region data copy jobs in DataSync that used US-WEST-2 as either source or destination. AWS was specific about what was not affected: already-running EC2 instances and already-created EBS volumes kept operating normally. It also advised against mutating running instances in ways that would trigger replacement or creation of new ones in the region, since those would hit the same failure path. The incident touched a wide set of named components, including VM Import/Export, Elastic Load Balancing, Elastic MapReduce, EMR Serverless and AWS Batch, all flagged separately for the Oregon region alongside their global listings. AWS logged early signs of recovery in its last update and the incident closed out at 2 hours and 23 minutes. Only three updates were published for the whole incident, and no post-mortem has been filed beyond them. No user reports were filed during the window, which fits an incident that hit instance launches rather than requests to already-running infrastructure.

Learning

UptimeRobot's own probes against DynamoDB, S3 and SQS in regions including us-east-1 and eu-west-1 held 100% uptime over 24 hours, 7 days and 30 days, and that is consistent with what AWS described: the fault was narrow, sitting in the subsystem that attaches EBS volumes during a new EC2 launch, not in the data plane that already-running services and already-created volumes use. A check that polls a live endpoint will stay green through this kind of fault because the endpoint it is hitting was never the thing that broke. The failure only shows up if you are trying to launch, replace or scale an instance in the affected region, which is exactly the action AWS told customers to avoid during the window. This is a useful pattern to recognise: your own monitoring tells you whether your product is reachable right now, but it cannot tell you whether your next deploy, autoscale event or instance replacement will succeed, because that depends on a layer you don't query directly. AWS's history here isn't rare either, with 7 incidents in the past 90 days and a median duration of 2 hours 24 minutes, almost identical to this one. Your monitoring covers your half, UptimeRobot now watches the provider's half too, so next time you know which side broke without guessing.

4.7
stars out of 5
284+ reviews on

Start monitoring in 30 seconds.

There's nothing to install. No credit card required. 50 monitors for free.