RFR: 8289291: HttpServer sets incorrect value for "max" parameter in Keep-Alive header value [v2]
Daniel Fuchs
dfuchs at openjdk.org
Wed Jun 29 15:15:40 UTC 2022
On Wed, 29 Jun 2022 12:05:42 GMT, Jaikiran Pai <jpai at openjdk.org> wrote:
>> Can I please get a review for this change which addresses https://bugs.openjdk.org/browse/JDK-8289291?
>>
>> As noted in that issue, right now, the Http(s)Server sets an incorrect value for the `max` parameter of the `Keep-Alive` header. The `max` value is supposed to be the number of subsequent requests that the server is willing to serve over that specific connection. The current value it sets is instead the number of idle connections that are configured for the server.
>>
>> The commit in this PR removes that `max` parameter altogether, since it isn't mandated by the spec, nor does the Http(s)Server have any specific construct to come up with a right value. Furthermore, on the client side the HttpURLConnection based client doesn't mandate the presence of this parameter in the `Keep-Alive` header. So this change won't cause any regressions in that area.
>>
>> tier1, tier2 and tier3 testing passed without any related issues after this change.
>
> Jaikiran Pai has updated the pull request incrementally with one additional commit since the last revision:
>
> Michael's suggestion - add a test to verify "max" parameter isn't present in Keep-Alive response header
Marked as reviewed by dfuchs (Reviewer).
Just wondering about using a simple Socket as the client (which is fine). I'd expect it should be possible to see the value of the header with the new HttpClient (unless we filter it?).
-------------
PR: https://git.openjdk.org/jdk/pull/9326
More information about the net-dev
mailing list