EBS Volume Detach throws 500

[AnirudhNarsipur]

Trying to detach an EBS volume throws a 500. LocalStack logs show:

2026-06-25T01:59:18.485 ERROR --- [et.reactor-2] l.aws.handlers.logging     : exception during call chain: object of type 'NoneType' has no len()
2026-06-25T01:59:18.486  INFO --- [et.reactor-2] localstack.request.aws     : AWS ec2.DetachVolume => 500 (InternalError)

Full repro script attached. Output from running repro script:

> bash repro_detach_volume_bug.sh
LocalStack CLI: LocalStack CLI 2026.6.0
AWS CLI: aws-cli/2.33.2 Python/3.13.11 Darwin/25.4.0 exe/arm64
endpoint: http://localhost:4566
region: us-west-2
availability zone: us-west-2a
starting LocalStack

     __                     _______ __             __
    / /   ____  _________ _/ / ___// /_____ ______/ /__
   / /   / __ \/ ___/ __ `/ /\__ \/ __/ __ `/ ___/ //_/
  / /___/ /_/ / /__/ /_/ / /___/ / /_/ /_/ / /__/ ,<
 /_____/\____/\___/\__,_/_//____/\__/\__,_/\___/_/|_|

- LocalStack CLI: 2026.6.0
- Profile: default
- App: https://app.localstack.cloud

[19:18:36] starting LocalStack in Docker mode 🐳                                                                localstack.py:172
2026-06-24T19:18:36.934  INFO --- [  MainThread] l.p.c.b.licensingv2        : Successfully activated cached license <redacted_license_id>:trial from /Users/latch/Library/Caches/localstack-cli/license.json 🔑✅
           preparing environment                                                                                 bootstrap.py:437
[19:18:37] configuring container                                                                                 bootstrap.py:440
           starting container                                                                                    bootstrap.py:440
           detaching                                                                                             bootstrap.py:440
waiting for LocalStack
curl: (52) Empty reply from server
curl: (52) Empty reply from server
curl: (52) Empty reply from server
curl: (52) Empty reply from server
curl: (52) Empty reply from server
resetting all LocalStack state at http://localhost:4566
creating EC2 resources
waiting for instance to run
waiting for volume to become available
{
    "VolumeId": "vol-512248300002efbe4",
    "InstanceId": "i-2d95601b9ff975a1c",
    "Device": "/dev/sdf",
    "State": "attaching",
    "AttachTime": "2026-06-25T02:18:49+00:00"
}
waiting for volume to become in-use
created:
  vpc: vpc-f16ed8c37fd2a1c63
  subnet: subnet-22dd93d6fd7a454f4
  image: ami-a8808e3b
  instance: i-2d95601b9ff975a1c
  volume: vol-512248300002efbe4
  device: /dev/sdf
volume before detach:
{
    "Volumes": [
        {
            "Iops": 3000,
            "VolumeType": "gp3",
            "VolumeId": "vol-512248300002efbe4",
            "Size": 8,
            "SnapshotId": "",
            "State": "in-use",
            "CreateTime": "2026-06-25T02:18:48+00:00",
            "Attachments": [
                {
                    "DeleteOnTermination": false,
                    "VolumeId": "vol-512248300002efbe4",
                    "InstanceId": "i-2d95601b9ff975a1c",
                    "Device": "/dev/sdf",
                    "State": "attached",
                    "AttachTime": "2026-06-25T02:18:49+00:00"
                }
            ],
            "Encrypted": false
        }
    ]
}
calling DetachVolume with VolumeId only

An error occurred (InternalError) when calling the DetachVolume operation (reached max retries: 2): exception while calling ec2.DetachVolume: object of type 'NoneType' has no len()
reproduced LocalStack DetachVolume error

Script:

#!/usr/bin/env bash
set -euo pipefail

endpoint="http://localhost:4566"
region="us-west-2"
az="us-west-2a"

aws_local() {
  aws --endpoint-url "${endpoint}" --region "${region}" --no-sign-request "$@"
}

wait_for_localstack() {
  for _ in $(seq 1 90); do
    if curl -fsS "${endpoint}/_localstack/health" >/dev/null; then
      return 0
    fi

    sleep 1
  done

  echo "LocalStack did not become healthy at ${endpoint}" >&2
  return 1
}

echo "LocalStack CLI: $(localstack --version)"
echo "AWS CLI: $(aws --version)"
echo "endpoint: ${endpoint}"
echo "region: ${region}"
echo "availability zone: ${az}"

echo "starting LocalStack"
localstack start -d

echo "waiting for LocalStack"
wait_for_localstack

echo "resetting all LocalStack state at ${endpoint}"
curl -fsS --request POST "${endpoint}/_localstack/state/reset" >/dev/null

echo "creating EC2 resources"
vpc_id="$(
  aws_local ec2 create-vpc \
    --cidr-block 10.0.0.0/16 \
    --query 'Vpc.VpcId' \
    --output text
)"

subnet_id="$(
  aws_local ec2 create-subnet \
    --vpc-id "${vpc_id}" \
    --cidr-block 10.0.1.0/24 \
    --availability-zone "${az}" \
    --query 'Subnet.SubnetId' \
    --output text
)"

image_id="$(
  aws_local ec2 register-image \
    --name "detach-volume-repro-$(date +%s)" \
    --description "Minimal DetachVolume repro" \
    --architecture x86_64 \
    --root-device-name /dev/xvda \
    --virtualization-type hvm \
    --block-device-mappings '[{"DeviceName":"/dev/xvda","Ebs":{"VolumeSize":8,"DeleteOnTermination":true,"VolumeType":"gp3"}}]' \
    --query 'ImageId' \
    --output text
)"

instance_id="$(
  aws_local ec2 run-instances \
    --image-id "${image_id}" \
    --instance-type t3.small \
    --subnet-id "${subnet_id}" \
    --count 1 \
    --query 'Instances[0].InstanceId' \
    --output text
)"

echo "waiting for instance to run"
aws_local ec2 wait instance-running --instance-ids "${instance_id}"

volume_id="$(
  aws_local ec2 create-volume \
    --availability-zone "${az}" \
    --size 8 \
    --volume-type gp3 \
    --query 'VolumeId' \
    --output text
)"

echo "waiting for volume to become available"
aws_local ec2 wait volume-available --volume-ids "${volume_id}"

device="/dev/sdf"
aws_local ec2 attach-volume \
  --volume-id "${volume_id}" \
  --instance-id "${instance_id}" \
  --device "${device}" \
  --output json

echo "waiting for volume to become in-use"
aws_local ec2 wait volume-in-use --volume-ids "${volume_id}"

echo "created:"
echo "  vpc: ${vpc_id}"
echo "  subnet: ${subnet_id}"
echo "  image: ${image_id}"
echo "  instance: ${instance_id}"
echo "  volume: ${volume_id}"
echo "  device: ${device}"

echo "volume before detach:"
aws_local ec2 describe-volumes --volume-ids "${volume_id}" --output json

# DetachVolume only requires VolumeId for a normal single-attach EBS volume.
echo "calling DetachVolume with VolumeId only"
set +e
detach_output="$(aws_local ec2 detach-volume --volume-id "${volume_id}" --output json 2>&1)"
detach_status="$?"
set -e

printf '%s\n' "${detach_output}"

if [[ "${detach_status}" -eq 0 ]]; then
  echo "DetachVolume succeeded; bug did not reproduce"
  exit 1
fi

if grep -Fq "object of type 'NoneType' has no len()" <<<"${detach_output}"; then
  echo "reproduced LocalStack DetachVolume error"
  exit 0
fi

echo "DetachVolume failed with a different error" >&2
exit "${detach_status}"

note: Repro script was authored by Codex.

[EddEcho]

Hi @AnirudhNarsipur,

Thanks for the detailed report the repro script made it really easy to confirm. We reproduced it on 2026.6.0 and on our latest dev build. Already passed it to engineering for a fix.

As a workaround, you can pass InstanceId and Device along with VolumeId — that returns 200:

awslocal ec2 detach-volume --volume-id <VOL_ID> --instance-id <INSTANCE_ID> --device /dev/sdf

We'll let you know here once the fix ships. 🚀