Hi Team, I’m experiencing issues with my data source connection. The recent changes are not reflecting, and I’m also encountering repeated connection failures. When I test the connection in the Data Sources tab, it initially shows as successful but then if I hit it again it fails with the error: “Could not reach URL.” For context, my database is hosted on AWS Redshift, and I’m connecting via a VPC endpoint service. Both the Redshift cluster and VPC appear healthy. Could you please help investigate this issue?

Thank you for providing detailed context. Based on the knowledge sources, here are some key points and troubleshooting steps for your scenario: 1. Intermittent Connection Issues via VPC Endpoint - Even if the initial connection test is successful, repeated failures with "Could not reach URL" can indicate networking or configuration issues with the VPC endpoint or the Redshift security group settings. - Double-check that the VPC endpoint DNS is correct and does not contain any trailing spaces or formatting errors, as this has caused similar issues for other users ([Slack thread: Hostname not valid](https://gooddataconnect.slack.com/archives/C04S1MSLEAW/p1748679803727319?thread_ts=1748679803.727319&cid=C04S1MSLEAW)). - Ensure that the Redshift security group allows inbound connections from the correct GoodData IP ranges for your region. If you are using PrivateLink, confirm that the endpoint and security group rules are set up as described in the [AWS PrivateLink documentation](https://www.gooddata.com/docs/cloud/connect-data/use-aws-privatelink/). 2. VPC Endpoint and Routing - If you have multiple environments (e.g., Prod and Non-Prod), each may require its own VPC endpoint or endpoint service, depending on your network architecture. Simply having VPC peering may not be sufficient for PrivateLink-based connectivity ([Slack thread: VPC endpoint for each environment](https://gooddataconnect.slack.com/archives/C04S1MSLEAW/p1748679803727319?thread_ts=1748679803.727319&cid=C04S1MSLEAW)). - Confirm that the endpoint service is correctly configured and that GoodData has been granted access via the correct AWS account ID. 3. Redshift and GoodData Known Issues - There are no known issues on the GoodData side that would cause intermittent failures if the network and endpoint are configured correctly ([Slack thread: No known issues](https://gooddataconnect.slack.com/archives/C04S1MSLEAW/p1748278507284779?thread_ts=1748278507.284779&cid=C04S1MSLEAW)). - If you continue to see "Could not reach URL," it is often due to network misconfiguration, security group rules, or endpoint DNS issues. 4. Next Steps - Double-check the VPC endpoint DNS and remove any trailing spaces. - Verify security group rules for both Redshift and the VPC endpoint. - If using PrivateLink, ensure the endpoint is set up for the correct environment and region. - If the issue persists, collect the traceId from the error message and share it with GoodData support for further investigation ([Slack thread: TraceId for troubleshooting](https://gooddataconnect.slack.com/archives/C04S1MSLEAW/p1748679803727319?thread_ts=1748679803.727319&cid=C04S1MSLEAW)). If you need more specific guidance, please provide the exact error message (with traceId) and confirm your VPC endpoint DNS and security group settings. This will help narrow down the root cause.

I have verified the dns name and the endpoint security group config to allow traffic from Gooddata's IP's. I am still getting this error.

Hi Aryan, We recently expanded our us-east cluster (new namespaces, nodes, pods), which required adding a new internal IP range:

10.247.0.0/16

If you are using AWS PrivateLink with IP whitelisting (e.g., restricting source IPs on your side via Security Groups or NLB rules), you’ll need to update your whitelist to include this new range. Otherwise, GoodData Cloud may not be able to reach your data source. We are gonna update our AWS PrivateLink add-on documentation to reflect this change. Apologies for the inconvenience caused.

Hi @Julius Kos. Thanks for the reply I'll add this CIDR range to the Security Groups and let you know. Yeah there wasn't any notification about the updated IP range and as this was affecting production systems, a heads up would have been nice.

Agreed! The same happened for us.

This was indeed a mistake on our side and is by no means standard. We have already discussed this internally and are taking steps to ensure it does not happen again in the future. Please accept our sincere apologies for the inconvenience caused.

Faced the same “Could not reach URL.” PriivateLink connection issue to our AWS RDS hosted DB. Shout out to @Aryan Arora to raise it🙌