# Troubleshoot Amazon Lambda deployment issues
<a name="troubleshooting-deployments-lambda"></a>

**Topics**
+ [Amazon Lambda deployments fail after manually stopping a Lambda deployment that does not have configured rollbacks](#troubleshooting-manually-stopped-lambda-deployment)

## Amazon Lambda deployments fail after manually stopping a Lambda deployment that does not have configured rollbacks
<a name="troubleshooting-manually-stopped-lambda-deployment"></a>

In some cases, the alias of a Lambda function specified in a deployment might reference two different versions of the function. The result is that subsequent attempts to deploy the Lambda function fail. A Lambda deployment can get in this state when it does not have rollbacks configured and is manually stopped. To proceed, use the Amazon Lambda console to make sure the function is not configured to shift traffic between two versions:

1. Sign in to the Amazon Web Services Management Console and open the Amazon Lambda console at [https://console.amazonaws.cn/lambda/](https://console.amazonaws.cn/lambda/).

1. From the left pane, choose **Functions**.

1. Select the name of the Lambda function that is in your CodeDeploy deployment.

1. From **Aliases**, choose the alias used in your CodeDeploy deployment, and then choose **Edit**.

1. From **Weighted alias**, choose **none**. This ensures the alias is not configured to shift a percentage, or weight, of traffic to more than one version. Make a note of the version selected in **Version**.

1. Choose **Save**.

1. Open the CodeDeploy console and attempt a deployment of the version displayed in the drop-down menu in step 5.