How do I troubleshoot QueueDoesNotExist errors when making API calls to my Amazon SQS queue?

Resolution

QueueDoesNotExist errors can result from various Amazon SQS API calls, including GetQueueAttributes, SendMessage, and DeleteMessage. To troubleshoot this error, review the following possible causes and resolution steps:

Incorrect queue URL

Verify that the queue URL provided in the request is correct and contains no typos.

Important: If the destination queue type is FIFO, then you must append the suffix .fifo to the queue URL.

Incorrect Region

A QueueDoesNotExist exception returns when the request is made to the incorrect AWS Region. The software development kit (SDK) and AWS Command Line Interface (AWS CLI) don’t pick up the destination Region from the queue URL. Instead, the Region is set by the client configuration.

You must correctly configure the Region on the Amazon SQS client before making an API call. Review the Amazon SQS client configuration to confirm that the correct Region is configured on the client. If a Region isn’t configured on the client, then the SDK or AWS CLI chooses the Region from the configuration file or the environment variable. If a Region isn’t found in the configuration file, then the SDK sets the Region to us-east-1 by default.

If the failed API call is supported by AWS CloudTrail, then verify all Regions in the customer account for the failed Amazon SQS operation. Verifying all Regions can help determine if the Region is the cause of the issue.

You can also verify the Region of the request by activating the debug log on the SDK or AWS CLI. Debug logs show the destination host for the request.

Example: Host: sqs.us-east-1.amazonaws.com

Queue recently purged or deleted

You might temporarily receive a QueueDoesNotExist error if the queue was recently purged. Identify the timestamp of the failed API call, and then verify CloudTrail for any PurgeQueue operations around the time of the error.

The error can also occur when the queue is deleted if the queue is part of an AWS CloudFormation or similar deployment stack. Updates to or deletion of a stack can cause the queue to be deleted and recreated. If the API call is made to the queue during deletion, the request can fail with a QueueDoesNotExist error. Verify CloudTrail for any DeleteQueue operations during the time of the error.

Cross account GetQueueUrl

For API calls, the SDK or AWS CLI usually take the destination queue account number from the queue URL. However, the GetQueueUrl API call doesn’t provide a queue's AWS account in the request. This means that by default the request is made against the caller account. If the request is intended for a cross account queue, then you must specify the destination queue account number as the API call QueueOwnerAWSAccountId parameter.

Message moved to a dead-letter queue

For standard SQS queues, when a dead-letter queue (DLQ) is configured, messages are moved to the DLQ after retries. QueueDoesNotExist errors can occur if you perform a DeleteMessage operation using an old ReceiptHandle from the main queue after the message is moved to the DLQ. You must delete messages within the configured VisibilityTimeout window.

Insufficient permissions

You can receive the QueueDoesNotExist when sending the GetQueueUrl API if your IAM user or role doesn't have sufficient permissions. To run the query, the IAM user or role must have GetQueueUrl permissions for the SQS queue that you run the query against.

For more information, see GetQueueURL.

Additional troubleshooting

If you contact AWS Support for additional troubleshooting, be sure to identify the RequestId and timestamp of the failed API calls.

---