Notebook possbile to use mTLS as Auth to API Service

[floody]

Hi everyone,

I am currently looking into options to ingest data from a external service via their API.

The API uses mTLS client certificates and private keys for Auth and was wondering what is the possibilities of using this auth method within Fabric Notebook using Python.

For testing purposes tried to upload the pem files to OneLake storage and reference them via "abfss" path and then use them in the request, but I get errors (OSError: Could not find the TLS certificate file, invalid path) so I assume something needs to be setup/configured differently.

cert = ('path/to/client/cert', 'path/to/client/key')
response = requests.get(url, cert=cert)

Hope someone can advise.

Cheers

[v-shex-msft]

HI @floody,

I'd like to suggest you try to use azure key vault to manage these credentials and you can use correspond libraries to invoke them from notebook python code:

Solved: Access Key-vault in notebooks - Microsoft Fabric Community

Accessing Azure Key Vault Secrets from Fabric Notebooks

Regards,

Xiaoxin Sheng

[floody]

Thanks! and yes, once we have Key Vault setup this will be the way to go for sure.

[richbenmintz]

Have you tried with the File API Path and or the or the realtive path for pyspark

[floody · Accepted solution]

Thanks Richard, I tried both, but no joy.

But what did work was using notebookutils e.g.

path = notebookutils.fs.getMountPath("/test")

https://learn.microsoft.com/en-us/fabric/data-engineering/notebook-utilities#access-files-under-the-...