Implement retry backoff for GetConnector API calls

[dstrates]

Description

We're experiencing flaky GetConnector behavior due to API calls failing when brokers are rebalancing. Although the client library seems to handle rebalances (see: go-kafka-connect issue #3), we could potentially implement safeguards to handle these cases better.

Suggestion: Implement retry backoff or a similar method to mitigate these issues. We could potentially use the retry mechanism outlined here: Terraform SDK v2 retries and timeouts.

Context

Frequently receiving the following error on Terraform apply:

Error: Get connector : {"error_code":500,"message":"Request cannot be completed because a rebalance is expected"}