Reading the docs I didn't see any mention of retrying if a call is taking too long. Does the RetryPolicy allow you to cancel a call before any response is received and initiate a retry?
A crucial part of the retry pattern is an exception, otherwise it cannot be called "retry" if no exception.
Your "too long request" easily fits into a request timeout concept which is usual fail in client with a RequestTimeoutException. So, your retry advice for the service call with a timeout should be configured to catch such an exception and retry on it.
This is really not a retry pattern responsibility to cancel request since there are just might not be such a hook or the cancelation can be too involved.
See for example, RestTemplate and its SimpleClientHttpRequestFactory:
/**
* Set the underlying URLConnection's read timeout (in milliseconds).
* A timeout value of 0 specifies an infinite timeout.
* <p>Default is the system's default timeout.
* @see URLConnection#setReadTimeout(int)
*/
public void setReadTimeout(int readTimeout) {
And its delegate:
/**
* Sets the read timeout to a specified timeout, in
* milliseconds. A non-zero value specifies the timeout when
* reading from Input stream when a connection is established to a
* resource. If the timeout expires before there is data available
* for read, a java.net.SocketTimeoutException is raised. A
* timeout of zero is interpreted as an infinite timeout.
*
* <p> Some non-standard implementation of this method ignores the
* specified timeout. To see the read timeout set, please call
* getReadTimeout().
*
* @param timeout an {@code int} that specifies the timeout
* value to be used in milliseconds
* @throws IllegalArgumentException if the timeout parameter is negative
*
* @see #getReadTimeout()
* @see InputStream#read()
* @since 1.5
*/
public void setReadTimeout(int timeout) {
The RestTemplate will wrap that SocketTimeoutException into a ResourceAccessException: Using sping's restTemplate with a timeout, how do I detect a timeout?