@rybalkinsd What if the retry-after delay is too large? And how should I couple this feature with old logic?

@Russiancold I think there should a limit, however, it's better to check how it's recommended to handle this delay

I believe that it can varies pretty much, so what about providing maxRetryTime field? If the header value overlimits it then it should be passed further to the chain then.

In our case, maxRetryTime should be somehow aligned with client connectTimeout, readTimeout and writeTimeout

So, it's probably no need to introduce a new variable, but use something based on existing

duplicate of #108