I have a self-hosted TTS V3 Open Source in AWS EC2. Self-generated Certs. With default EC2 domain name (ec2-xxx-xxx-xxx-xxx.ap-southeast-1.compute.amazonaws.com). It seems to work fine. I can create users, create applications, create gateways, create devices. My Basic Station gateway is connected and my devices are landing data into TTS. Now I am trying to install the AWS IoT Default Integration, but the CloudFormation fails as follows:
Received response status [FAILED] from custom resource. Message returned: Dial Application Server at "ttn-v3-as:///ec2-xxx-xxx-xxx-xxx.ap-southeast-1.compute.amazonaws.com/": connection error: desc = "transport: error while dialing: dial tcp: lookup [ec2-xxx-xxx-xxx-xxx.ap-southeast-1.compute.amazonaws.com/](http://ec2-xxx-xxx-xxx-xxx.ap-southeast-1.compute.amazonaws.com/): no such host"
That is from installing the Enterprise edition of the CloudFormation. I have successfully installed the Community edition of the CloudFormation for my Community account and application in the past. So I hacked that template to point to ec2-xxx-xxx-xxx-xxx.ap-southeast-1.compute.amazonaws.com and tried to install that. Then I hit this error:
sending status failed: Dial Application Server at "ttn-v3-as:///ec2-xxx-xxx-xxx-xxx.ap-southeast-1.compute.amazonaws.com": context deadline exceeded
I am uncertain what’s going on. It seems the error is related to service discovery and DNS SRV records: Networking | The Things Stack for LoRaWAN
My stack config file has no elements for gRPC configuration: General Options | The Things Stack for LoRaWAN
I suspect that the configuration documentation possibly neglects to cover this part of the setup for self-hosted solutions. And that even if it did, I perhaps can’t do it unless I own the domain. Can anyone point me in the right direction?