roachtest: cluster_creation failed

roachtest.cluster_creation [failed](https://teamcity.cockroachdb.com/buildConfiguration/Cockroach_Nightlies_RoachtestNightlyAwsBazel/21454324?buildTab=log) with [artifacts](https://teamcity.cockroachdb.com/buildConfiguration/Cockroach_Nightlies_RoachtestNightlyAwsBazel/21454324?buildTab=artifacts#/admission-control/disk-bandwidth-limiter) on release-25.2.23-rc @ [e8a36748d0b3bce58abc8a12ccd02a347322b60f](https://github.com/cockroachdb/cockroach/commits/e8a36748d0b3bce58abc8a12ccd02a347322b60f):

```
test admission-control/disk-bandwidth-limiter failed: (test_runner.go:840).func4: in provider: aws: failed to run: aws ec2 run-instances --associate-public-ip-address --count 1 --instance-type m7g.xlarge --image-id ami-05983a09f7dc1c18f --key-name teamcity-runner-hHvbiNEyCms90reRke8FMM_DVIQ= --region us-east-2 --security-group-ids sg-04d72b57e29d671f1 --subnet-id subnet-0258dc9d1b6473d84 --tag-specifications ResourceType=instance,Tags=[{Key=usage,Value=roachtest},{Key=roachtest-branch,Value=release-25-2-23-rc},{Key=cluster,Value=teamcity-21454324-1785996031-10-n2cpu8},{Key=lifetime,Value=12h0m0s},{Key=arch,Value=arm64},{Key=created,Value=2026-08-06T06:11:50Z},{Key=Name,Value=teamcity-21454324-1785996031-10-n2cpu8-0002},{Key=Cluster,Value=teamcity-21454324-1785996031-10-n2cpu8},{Key=spot,Value=true},{Key=roachprod,Value=true}] ResourceType=volume,Tags=[{Key=usage,Value=roachtest},{Key=roachtest-branch,Value=release-25-2-23-rc},{Key=cluster,Value=teamcity-21454324-1785996031-10-n2cpu8},{Key=lifetime,Value=12h0m0s},{Key=arch,Value=arm64},{Key=created,Value=2026-08-06T06:11:50Z},{Key=Name,Value=teamcity-21454324-1785996031-10-n2cpu8-0002},{Key=Cluster,Value=teamcity-21454324-1785996031-10-n2cpu8},{Key=spot,Value=true},{Key=roachprod,Value=true}] --user-data file:///tmp/aws-startup-script3856454410 --iam-instance-profile Name=roachprod-testing --block-device-mapping file:///tmp/aws-block-device-mapping3018437384 --instance-market-options MarketType=spot,SpotOptions={SpotInstanceType=one-time,InstanceInterruptionBehavior=terminate} --output json: stderr:
An error occurred (InsufficientInstanceCapacity) when calling the RunInstances operation (reached max retries: 2): We currently do not have sufficient m7g.xlarge capacity in the Availability Zone you requested (us-east-2a). Our system will be working on provisioning additional capacity. You can currently get m7g.xlarge capacity by not specifying an Availability Zone in your request or choosing us-east-2b, us-east-2c.: exit status 254 [owner=test-eng]
```

Parameters:
- MetamorphicFilesystem=ext4
- MetamorphicVolumeType=io2
- cloud=aws
- coverageBuild=false
- cpu=8
- diskCount=0
- runtimeAssertionsBuild=false
Help

See: [roachtest README](https://github.com/cockroachdb/cockroach/blob/master/pkg/cmd/roachtest/README.md)

See: [How To Investigate \(internal\)](https://cockroachlabs.atlassian.net/l/c/SSSBr8c7)

Same failure on other branches

- #172510 roachtest: cluster_creation failed [O-roachtest O-robot T-testeng X-infra-flake branch-release-26.1.8-rc]
- #172508 roachtest: cluster_creation failed [O-roachtest O-robot T-testeng X-infra-flake branch-release-26.1.8-rc]
- #172507 roachtest: cluster_creation failed [O-roachtest O-robot T-testeng X-infra-flake branch-release-25.4.14-rc]
- #172506 roachtest: cluster_creation failed [O-roachtest O-robot T-testeng X-infra-flake branch-release-25.2.22-rc]
- #172191 roachtest: cluster_creation failed [O-roachtest O-robot T-testeng X-infra-flake branch-release-25.4]
- #172020 roachtest: cluster_creation failed [O-roachtest O-robot T-testeng X-infra-flake branch-release-25.2]

/cc @cockroachlabs/test-eng

[This test on roachdash](https://roachdash.crdb.dev/?filter=status:open%20t:.*cluster_creation.*&sort=title+created&display=lastcommented+project) | [Improve this report!](https://github.com/cockroachdb/cockroach/tree/master/pkg/cmd/bazci/githubpost/issues)

Start with the roachtest README and test_runner.go around line 840, then inspect the cluster_creation report and its AWS RunInstances error. Compare the repeated related failures linked in the issue to determine whether the capacity error has a code-level handling path. Done is not defined by the report, so confirm the intended resolution with test-eng before changing anything.

Written by the indexing model from the issue text.