On Sun, 2019-10-27 at 12:30 +0100, Michael Osipov wrote: > Am 2019-10-23 um 09:35 schrieb Oleg Kalnichevski: > > On Tue, 2019-10-22 at 19:47 +0200, Michael Osipov wrote: > > > Am 2019-10-22 um 15:09 schrieb Michael Osipov: > > > > Friends, > > > > > > > > while working on the PR [1] for WAGON-567, it came to my mind > > > > whether it > > > > would be beneficial if > > > > org.apache.http.impl.client.DefaultServiceUnavailableRetryStrat > > > > egy > > > > would > > > > accept a Collection<Integer> of status codes which can be > > > > retried. > > > > > > > > Internally, that woudd be copied into a HashSet. If no values > > > > are > > > > provided, we stick to SC_SERVICE_UNAVAILABLE. > > > > > > To add some more fuel to the fire. Is > > > ServiceUnavailableRetryStrategy > > > intended for server-side errors only? I wonder why 429 is not > > > included here. > > > > > > > Yes, it was initially intended for recovery from temporary > > unavailability of the target server. > > So from your POV, adding 4xx error would be an abuse of the > interface? > If so, what is the alternative for 408 or 429? >
No, it would not. ServiceUnavailableRetryStrategy was unfortunately not the best choice of the name for the interface. It should have been renamed in 5.0 but I feel it is a bit too late for that. Oleg > Michael > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [email protected] > For additional commands, e-mail: [email protected] > --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
