[quicwg/base-drafts] Separate HTTP/3 stream errors from connection errors. (#2911)

Daan De Meyer <notifications@github.com> Sat, 20 July 2019 14:46 UTC

Return-Path: <noreply@github.com>
X-Original-To: quic-issues@ietfa.amsl.com
Delivered-To: quic-issues@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 7EE42120103 for <quic-issues@ietfa.amsl.com>; Sat, 20 Jul 2019 07:46:02 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -8
X-Spam-Level:
X-Spam-Status: No, score=-8 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.001, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, HTML_MESSAGE=0.001, MAILING_LIST_MULTI=-1, RCVD_IN_DNSWL_HI=-5, SPF_HELO_NONE=0.001, SPF_PASS=-0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (1024-bit key) header.d=github.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id hAToKSmya4CY for <quic-issues@ietfa.amsl.com>; Sat, 20 Jul 2019 07:46:00 -0700 (PDT)
Received: from out-24.smtp.github.com (out-24.smtp.github.com [192.30.252.207]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 28F221201B3 for <quic-issues@ietf.org>; Sat, 20 Jul 2019 07:46:00 -0700 (PDT)
Date: Sat, 20 Jul 2019 07:45:59 -0700
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=github.com; s=pf2014; t=1563633959; bh=9/J7r6DjdJFDy+4hlIB7A1mUgYP9jKAtBsMqDj6+pMQ=; h=Date:From:Reply-To:To:Cc:Subject:List-ID:List-Archive:List-Post: List-Unsubscribe:From; b=ph/UfRCCSpQPKl6QpKwyaJyeiDN/zRNyrEdLjarhVlvbMBOAFdy0iH96LbB0c0eG5 GXKBRdrFtNeDN2M6VREj/vVunInFXZ/MixoVv9Nlv+lQ5gs35qelHknMPmel7ZOSOJ W1EsI+vIWvlUm7RaqktF5ZqxCteSH+JxbU7nNqhQ=
From: Daan De Meyer <notifications@github.com>
Reply-To: quicwg/base-drafts <reply+AFTOJK4V4RQKSHRNKHMBBQN3IBN2PEVBNHHBYDQKAQ@reply.github.com>
To: quicwg/base-drafts <base-drafts@noreply.github.com>
Cc: Subscribed <subscribed@noreply.github.com>
Message-ID: <quicwg/base-drafts/issues/2911@github.com>
Subject: [quicwg/base-drafts] Separate HTTP/3 stream errors from connection errors. (#2911)
Mime-Version: 1.0
Content-Type: multipart/alternative; boundary="--==_mimepart_5d33292746d84_1e8b3f85920cd96814093d7"; charset="UTF-8"
Content-Transfer-Encoding: 7bit
Precedence: list
X-GitHub-Sender: DaanDeMeyer
X-GitHub-Recipient: quic-issues
X-GitHub-Reason: subscribed
X-Auto-Response-Suppress: All
X-GitHub-Recipient-Address: quic-issues@ietf.org
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic-issues/SyjaBmFdk2Iriicp0jldLeWXiPw>
X-BeenThere: quic-issues@ietf.org
X-Mailman-Version: 2.1.29
List-Id: Notification list for GitHub issues related to the QUIC WG <quic-issues.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic-issues>, <mailto:quic-issues-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic-issues/>
List-Post: <mailto:quic-issues@ietf.org>
List-Help: <mailto:quic-issues-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic-issues>, <mailto:quic-issues-request@ietf.org?subject=subscribe>
X-List-Received-Date: Sat, 20 Jul 2019 14:46:03 -0000

With the recent changes in HTTP/3 errors, it becomes possible to completely divide errors into stream errors and connection errors, with each error being either a stream error or connection error. As a result, we can also remove the possibility to treat a stream error as a connection error which allows implementations to determine whether an error is a stream error or a connection error based solely on its type.

As far as I can determine from the draft, only HTTP_STREAM_CREATION_ERROR is used as both a stream error and a connection error. It is only used as a stream error when a peer opens a unidirectional stream of an unknown type. If a new error HTTP_UKNOWN_STREAM_TYPE is split of as a stream error from HTTP_STREAM_CREATION_ERROR, HTTP_STREAM_CREATION_ERROR can be classed as solely a connection error.

[This analysis](https://github.com/quicwg/base-drafts/pull/2662#issuecomment-507529717) from @kazuho further lists HTTP_NO_ERROR and HTTP_GENERAL_PROTOCOL_ERROR as errors that can be both connection and stream errors. I think these can be safely classed as solely connection errors. 

HTTP_NO_ERROR is only defined in the draft for usage in CONNECTION_CLOSE, I'm not sure when it would be used as a stream error (we have more specific error codes to indicate intent when closing a stream).

HTTP_GENERAL_PROTOCOL_ERROR is only used as a stream error in the section discussing malformed requests/responses. I'd suggest using a more descriptive error for this case (HTTP_MALFORMED_MESSAGE  for example) and treating HTTP_GENERAL_PROTOCOL_ERROR as solely a connection error.

An error being either a connection error or a stream error simplifies implementation as they don't have to check on every error if it was communicated as a stream error or a connection error. Upgrading a stream error to a connection error is not a sensible approach for the existing stream errors (for example, closing the connection because of `REQUEST_CANCELLED` does not sound like a reasonable approach). If an implementation chooses to treat a stream error as a connection error, perhaps the INTERNAL_ERROR code should be used instead as not being able to handle a stream error without closing the connection is an implementation issue and should be treated as a bug.

There's probably some things I'm overlooking. Any feedback is welcome.

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/quicwg/base-drafts/issues/2911