MissingAction error for everything

[meza]

Hey, just trying to use the absolute basics and this is what I'm getting for all command line attempts:

An error occurred (400) when calling the CreateQueue operation: <ErrorResponse xmlns="http://queue.amazonaws.com/doc/2012-11-05/">
      <Error>
        <Type>Sender</Type>
        <Code>MissingAction</Code>
        <Message>MissingAction; see the SQS docs.</Message>
        <Detail/>
      </Error>
      <RequestId>00000000-0000-0000-0000-000000000000</RequestId>
    </ErrorResponse>

Attempting with the commands in the readme aws --endpoint-url http://localhost:9324 sqs create-queue --queue-name newqueue

What am I doing wrong?

[apynes]

I am getting the exact same issue. We recently upgraded to aws api v2 and this is all I can get out of the queue as well.

[devvaneio]

Same here

[alifsabrani]

I also experience the same issue

[devvaneio]

I recommend using localstack instead

[anumbskull]

I'm also getting this error. I've tried reordering the params, adding region, and reformatting my attributes. Here is what my last try looked like:
aws sqs create-queue --queue-name dev-content-import.fifo --attributes "{\"FifoQueue\":\"true\"}" --endpoint-url http://localhost:9324 --region us-east-1

[ztaf]

Had similar issue and finally just reverted to below mock/workaround:

docker run -p 9324:9324 -p 9325:9325 -p softwaremill/elasticmq-native
aws --endpoint-url http://localhost:9324/ sqs create-queue --queue-name software-update-events

[priyankgosalia]

To everyone who's getting a 'MissingAction' while running any aws command against alpine-sqs, it's because of the version of AWS CLI you are using.

Alpine-SQS works only with AWS CLI v1.x and not with v2.x.

Check your aws cli version by typing:
aws --version

if it shows v2.x, then your commands wont work. You need to install AWS CLI v1.x. In my case, I am using v1.22.34 and I am able to create queues on Alpine SQS.

$ aws --version
aws-cli/1.22.34

Hope this helps.

[matt-dies-tenet3]

We're experiencing the same thing, and the aws CLI seems to be unrelated. We're experiencing the same problem using boto3 instead. We've correctly run the container and are able to communicate with it through curl commands successfully:

$ curl -X GET "http://sqs:9324?Action=ListQueues"
<ListQueuesResponse xmlns="http://queue.amazonaws.com/doc/2012-11-05/">
                <ListQueuesResult>
                  <QueueUrl>http://sqs:9324/queue/default</QueueUrl><QueueUrl>http://sqs:9324/queue/service-queue</QueueUrl>
                </ListQueuesResult>
                <ResponseMetadata>
                  <RequestId>00000000-0000-0000-0000-000000000000</RequestId>
                </ResponseMetadata>
              </ListQueuesResponse>

However, the equivalent call in boto3 fails:

"""Purely exists to test if the `alpine-sqs` integration is working as expected."""

import boto3
from botocore.exceptions import NoCredentialsError, PartialCredentialsError

sqs_client = boto3.client(
    "sqs",
    region_name="us-east-1",
    endpoint_url="http://sqs:9324",
    use_ssl=False,
)
print(sqs_client.list_queues())  # failure will occur in this line

# in the off chance it doesn't fail (it will), this will add a message to the queue
try:
    _ = sqs_client.send_message(QueueUrl="http://sqs:9324/queue/default", MessageBody="Hello, this is a test message!")
except NoCredentialsError:
    print("Error: No AWS credentials found.")
except PartialCredentialsError:
    print("Error: Incomplete AWS credentials.")
except Exception as e:
    print(f"An error occurred: {e}")

And the exact error raised from the above:

botocore.exceptions.ClientError: An error occurred (400) when calling the ListQueues operation: <ErrorResponse xmlns="http://queue.amazonaws.com/doc/2012-11-05/">
      <Error>
        <Type>Sender</Type>
        <Code>MissingAction</Code>
        <Message>MissingAction; see the SQS docs.</Message>
        <Detail/>
      </Error>
      <RequestId>00000000-0000-0000-0000-000000000000</RequestId>
    </ErrorResponse>

Changing the endpoint_url argument will cause an EndpointConnectionError, implying it's connecting just fine. Similarly, it's the same endpoint used in the curl command.

If anyone has any resolution to this or this is determined to be a bug, it would be greatly helpful to point it out. Thank you!

[alifsabrani]

For anyone whose still facing this issue, i recommended you guys to just use https://github.com/softwaremill/elasticmq. The configuration is pretty much the same, but it is still being actively developed.

[matt-dies-tenet3]

For anyone whose still facing this issue, i recommended you guys to just use https://github.com/softwaremill/elasticmq. The configuration is pretty much the same, but it is still being actively developed.

I can happily confirm that these issues do not occur on elasticmq!

[sparrowt]

We hit this when upgrading boto3 & botocore from:

boto3==1.28.15
botocore==1.31.15

to

boto3==1.35.68
botocore==1.35.68

so my guess is somewhere in between those versions, there was a change behind the scenes in one or both, which had an equivalent effect (in breaking this repo) as the v1-to-v2 changes in AWS CLI.

Looking at both changelogs:
https://github.com/boto/boto3/blob/develop/CHANGELOG.rst
https://github.com/boto/botocore/blob/develop/CHANGELOG.rst
there are several mentions of This release enables customers to call SQS using AWS JSON-1.0 protocol although these are interspersed with several "Revert"/"Rolled back" entries for SQS - it was first added in 1.26.127 but then removed swiftly again - it finally seemed to settle down most recently in 1.34.90 where it was re-added and not reverted since.