Slow pipeline causes prompts to become corrupted

Platform API CX as Code
1

Sometimes our deployment is slow due to our CI/CD doing many security checks. This may cause an issue where the prompts end up corrupted. A corrupted prompt returns the following error:

prompt_error
prompt_error1276×111 6.57 KB

We can also see the following error in the pipeline.

2024-05-13T18:26:21.666Z [INFO]  provider.terraform-provider-genesyscloud_v1.29.0: 2024/05/13 18:26:21 Content of upload: <html>
<head><title>502 Bad Gateway</title></head>
<body>
<center><h1>502 Bad Gateway</h1></center>
</body>
</html>
: timestamp=2024-05-13T18:26:21.666Z

It is an issue for us since our pipeline will often be slow and we believe that Genesys is timing out.

2

Hi Walin,

A couple of things:

  1. You are running a really old version of CX as Code. I would recommend you upgrade to the latest CX as Code provider.

  2. Please turn on the SDK debugging in the provider and set the TF_LOG=json. If you are able to see in the logs the correlation id of the call for the prompt, this could help. This is more then likely not a problem with CX as Code itself but rather the underlying API that is being called.

Thanks,
John

3

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.