Do not retry when timeout is hit

[Mpdreamz]

this PR adds a safeguard for retrying after the timeout is hit as reported on #1080 .

This prevents:

• a query that took 60seconds on one node to be retried on a second node.
• a query that timed out after 30seconds on 2 nodes to be retried a third time.

By default this listens to SetTimeout() which itself defaults to 60 seconds. The max retry timeout can be configured separately by calling SetMaxRetryTimeout().

The MaxRetryException message itself clearly reflects when its thrown because we exceeded the MaxRetryTimeout.

[gmarz]

LGTM

[ankitkinra]

I want confirm behavior of SetTimeout() from pre 1.3 version to 1.3 version. Say we have the following code in Nest Api 1.0

var timeoutInMillis = 100;
var maxRetries = 3;
var config= new ConnectionSettings(pool);
config.SetTimeout(timeoutInMillis);
config.MaximumRetries(maxRetries);

Now imagine the scenario where request times out on the first node in 100 milliseconds, the request will repeat for 3 more times with each timeout being 100 milliseconds, when the same code is migrated to Nest 1.3.1 with the above fix, the request will only be tried for 100 milliseconds and will not be retried as by default MaxRetryTimeout will be equal to Timeout, hence this becomes a breaking change.
I wonder if the default value of MaxRetryTimeout could have been maxRetries * timeoutInMillis