W3C home > Mailing lists > Public > ietf-http-wg@w3.org > April to June 2014

Re: Graceful shutdown #458

From: Jeff Pinner <jpinner@twitter.com>
Date: Thu, 8 May 2014 16:09:27 -0700
Message-ID: <CA+pLO_hVVKK11VUi0e+nHLnR=0=GDci8TdNNpF4-0_EHZ5LbCA@mail.gmail.com>
To: Daniel Sommermann <dcsommer@fb.com>
Cc: Roberto Peon <grmocg@gmail.com>, Martin Thomson <martin.thomson@gmail.com>, HTTP Working Group <ietf-http-wg@w3.org>
We have a SPDY proof of concept running that will send two GOAWAY frames on
shutdown:

GOAWAY(max stream id, NO_ERROR)
(send ping and wait for pong or timeout)
GOAWAY(last good stream id, *some code*)
TCP FIN


On Thu, May 8, 2014 at 4:01 PM, Daniel Sommermann <dcsommer@fb.com> wrote:

> I'm not very keen on using ALTSVC to indicate a connection draining state.
> Especially since client implementation of alternative services is rather
> fuzzy (not much normative language in that spec), I'd rather stick to a
> mechanism where the client behavior is well understood and requires minimal
> changes to the spec.
>
> I have a slight preference for GOAWAY over DRAINING. With DRAINING we gain
> added explicitness, but at the cost of verbosity (the state diagram would
> grow at least). The spec already establishes that multiple GOAWAYs can be
> received (NO_ERROR followed by some other error during the graceful
> connection shutdown), so I think the work we need to do here is minimal.
> Either mechanism is fine for me, and I'm happy to resolve this at the
> interim.
>
>
Received on Thursday, 8 May 2014 23:09:55 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 17:14:30 UTC