No Proxy support for DynamoDB streams

[ThinkBriK]

DynamoDB streams do not honor HTTPS_PROXY env variable.
What happened:
Tried to set HTTP_PROXY and HTTPS_PROXY to use dynamoDB backend in a private environment, and I've got the error :

ERRO [DYNAMODB]  "Poll streams returned with error: RequestError: send request failed\ncaused by: Post https://streams.dynamodb.eu-west-3.amazonaws.com/: dial tcp 52.46.68.12:443: connect: connection timed out." dynamo/shards.go:62

When using aws cli from the same VM :

[~]$ export NO_PROXY=169.254.169.254
[~]$ export HTTP_PROXY=http://proxy.my.cloud:8080
[~]$ export HTTPS_PROXY=http://proxy.my.cloud:8080
[~]$ aws dynamodbstreams list-streams
{
    "Streams": [
        REDACTED
}

What you expected to happen:
Working HTTP Proxy support for DynamoDB streams

How to reproduce it (as minimally and precisely as possible):
Add HTTPS_PROXY and HTTP_PROXY variable to the SystemD unit and NO_PROXY=169.254.169.254 for metadata access in AWS.
Check logs

Environment:

• Teleport version (use teleport version): Teleport v4.0.4 git:v4.0.4-0-g1a2ed507 go1.12.1
• Tsh version (use tsh version): Teleport v4.0.4 git:v4.0.4-0-g1a2ed507 go1.12.1
• OS (e.g. from /etc/os-release): Centos 7

Relevant Debug Logs If Applicable

ERRO [DYNAMODB]  "Poll streams returned with error: RequestError: send request failed\ncaused by: Post https://streams.dynamodb.eu-west-3.amazonaws.com/: dial tcp 52.46.68.12:443: connect: connection timed out." dynamo/shards.go:62

[ThinkBriK]

Basically this issue prevent any private deployment with a proxified internet proxy : As DynamoDB Streams are not available via VPC endpoints and Teleport code doesn't support using a proxy, cluster members never reload their conf.

[zmb3]

Looks like this was fixed in #4271