Power Platform PostgreSQL Connector Error: 'Tried to write more bytes than the message length specified'

Question

Hi there,

I have encountered an issue using the Microsoft Power Platform. This is my first time posting a question - I did not see a perfect tag so I selected what I believe to be the most relevant.

I have a couple power apps and several power automate flows that rely on a postgrsql connection that I've made in ower platform. It points at a PostgreSQL DB hosted in AWS RDS - PostgreSQL v16.8. As of 3 days ago, everything was functioning properly but this morning, the connector fails to connect and provides the following error message:

An error happened while reading data from the provider: 'Tried to write more bytes than the message length specified' inner exception: An error happened while reading data from the provider: 'Tried to write more bytes than the message length specified' clientRequestId: e5579687-f527-41b1-a47f-1be35337ed07

As far as I can tell, there haven't been any updates on my DB server side. I have tried rebooting it to no avail. Have there been any updates to the connector drivers on microsoft's side? Any suggestions?

Answer 1

Hi ,

Thanks for reaching out to Microsoft Q&A.

A transport layer mismatch between the power platform PostgreSQL connector and your RDS instance is the most likely cause. Recreating the connection and checking SSL protocol settings usually resolves it.

imo, this error almost never comes from PostgreSQL itself. It usually indicates a protocol-level mismatch or corruption between the client (power platform PostgreSQL conn) and the server. In practice, three things commonly cause it:

1. A Power Platform connector update has rolled out and is no longer compatible with your server’s TLS or network configuration. These silent connector updates do happen.

A network device (AWS NLB/ALB, VPN, proxy, firewall, inspection layer) is modifying or truncating packets.

PostgreSQL RDS recently enforced or changed SSL/TLS settings, especially if “require SSL” is enabled.

What you can try next?

Recreate the connection in Power Platform. This forces the connector to renegotiate protocol and TLS.

Check that your RDS instance still allows TLS 1.2 and verify rds.force_ssl and ssl_min_protocol_version parameters.

Test the same connection from a local machine using the latest Npgsql client. If that works, the issue is definitely in the Power Platform connector path.

If you are using an AWS load balancer, temporarily bypass it and connect directly.

Raise a support ticket with Microsoft. These connector regressions are not uncommon and are not visible in public change logs.

Please 'Upvote'(Thumbs-up) and 'Accept' as answer if the reply was helpful. This will be benefitting other community members who face the same issue.