From: Junio C Hamano <gitster@pobox•com>
To: Duy Nguyen <pclouds@gmail•com>
Cc: Stefan Beller <sbeller@google•com>,
Git Mailing List <git@vger•kernel.org>
Subject: Re: [RFC/PATCH 0/3] protocol v2
Date: Fri, 27 Feb 2015 15:05:58 -0800 [thread overview]
Message-ID: <xmqqioenhs4p.fsf@gitster.dls.corp.google.com> (raw)
In-Reply-To: <xmqqzj80l9c7.fsf@gitster.dls.corp.google.com> (Junio C. Hamano's message of "Thu, 26 Feb 2015 12:13:44 -0800")
Junio C Hamano <gitster@pobox•com> writes:
> I do not think v1 can be fixed by "send one ref with capability,
> newer client may respond immediately so we can stop enumerating
> remaining refs and older one will get stuck so we can have a timeout
> to see if the connection is from the newer one, and send the rest
> for the older client", because anything that involves such a timeout
> would not reliably work over WAN.
Just for fun, I was trying to see if there is a hole in the current
protocol that allows a new client to talk a valid v1 protocol
exchange with existing, deployed servers without breaking, while
letting it to know a new server that it is a new client and it does
not want to get blasted by megabytes of ref advertisement.
The idea is to find a request that can be sent as the first
utterance by the client to an old server that is interpreted as a
no-op and can be recognised by a new server as such a "no-op probe".
If there is such a request, then the exchange can go like this with
(new client, old server) pair:
- new client connects and sends that no-op.
- old server starts blasting the ref advertisement
- new client monitors and notices that the other side
started speaking, and the ref advertisement lacks the
capability bit for new protocol.
- new client accepts the ref advertisement and does the v1
protocol thing as a follow-up to what it already sent.
As long as the first one turns out to be no-op for old server, we
would be OK. On the other hand, (new client, new server) pair
would go like this:
- new client connects and sends that no-op.
- new server notices that there is already a data from the
client, and recognises the "no-op probe".
- new server gives the first v2 protocol message with
capability.
- new client notices thqat the other side started speaking, and
it is the first v2 protocol message.
- both sides happily speak v2.
and (old client, new server) pair would go like this:
- old client connects and waits.
- new server notices that there is *no* data sent from the
client and decides that the other side is a v1 client. It
starts blasting the ref advertisement.
- both sides happily speak v1 from here on.
A misdetected case between (new client, new server) pair might go
like this:
- new client connects and sends that no-op.
- new server accepts the connection, but that no-op probe has
not arrived yet.". It misdetects the other side as a v1
client and it starts blasting the ref advertisement.
- new client notices that the ref advertisement has the
capability bit and the server is capable of v2 protocol. it
waits until the server sends "sorry, I misdetected" message.
- new server eventually notices the "no-op probe" while blasting
the ref advertisement and it can stop in the middle.
hopefully this can happen after only sending a few kilobytes
among megabytes of ref advertisement data ;-). The server
sends "sorry, I misdetected" message to synchronise.
- both sides happily speak v2 from here on.
So the topic of this exercise ("just for fun") is to see if there is
such a no-op request the client side can send as the first thing for
probing.
On the fetch side, the first response upload-pack expects are one
of:
- "want " followed by an object name.
- "shallow " followed by an object name.
- "deepen " followed by a positive integer.
And there _is_ a hole ;-). The parsing of "shallow " object name is
done in such a way that an object name that passes get_sha1_hex()
that results in a NULL return from parse_object() is _ignored_. So
a new client can use "shallow 0{40}" as a no-op probe.
It appears that on the push side, there is a similar hole that can
be used. receive-pack expects either "shallow ", "push-cert" or the
refname updates (i.e. two "[0-9a-f]{40}" followed by a refname); the
parsing of "shallow " is not as loose as the fetch side in that
using a "shallow 0{40}" as a no-op probe will end up causing
prepare_shallow_info() sift the "0{40}" object name into "theirs",
but I think it will be ignored at the end as "unreachable cruft"
without causing harm.
I am _not_ proposing that we should go this route, at least not yet.
I am merely pointing out that an in-place sidegrade from v1 to a
protocol that avoids the megabyte-advertisement-at-the-beginning
seems to be possible, as a food for thought.
next prev parent reply other threads:[~2015-02-27 23:06 UTC|newest]
Thread overview: 80+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-02-24 3:12 [RFC/PATCH 0/3] protocol v2 Stefan Beller
2015-02-24 3:12 ` [PATCH 1/3] Document protocol capabilities extension Stefan Beller
2015-02-24 3:12 ` [PATCH 2/3] receive-pack: add advertisement of different protocol options Stefan Beller
2015-02-24 3:12 ` [PATCH 3/3] receive-pack: enable protocol v2 Stefan Beller
2015-02-24 4:02 ` [RFC/PATCH 0/3] " Duy Nguyen
2015-02-24 5:40 ` Stefan Beller
2015-02-24 6:15 ` Junio C Hamano
2015-02-24 23:37 ` Stefan Beller
2015-02-25 12:44 ` Duy Nguyen
2015-02-25 18:04 ` Junio C Hamano
2015-02-26 7:31 ` Stefan Beller
2015-02-26 10:15 ` Duy Nguyen
2015-02-26 20:08 ` Stefan Beller
[not found] ` <CACsJy8DOS_999ZgW7TqsH-dkrUFpjZf0TFQeFUt9s0bNhHY0Bw@mail.gmail.com>
2015-02-27 22:20 ` Stefan Beller
2015-02-26 20:13 ` Junio C Hamano
2015-02-27 1:26 ` Stefan Beller
2015-02-27 2:15 ` Stefan Beller
2015-02-27 23:05 ` Junio C Hamano [this message]
2015-02-27 23:44 ` Stefan Beller
2015-02-28 0:33 ` Junio C Hamano
2015-02-28 0:46 ` Stefan Beller
2015-02-28 1:01 ` [RFC/PATCH 0/5] protocol v2 for upload-pack Stefan Beller
2015-02-28 1:01 ` [RFC/PATCH 1/5] upload-pack: only accept capabilities on the first "want" line Stefan Beller
2015-02-28 1:01 ` [RFC/PATCH 2/5] upload-pack: support out of band client capability requests Stefan Beller
2015-02-28 7:47 ` Kyle J. McKay
2015-02-28 11:22 ` Duy Nguyen
2015-02-28 22:36 ` Kyle J. McKay
2015-03-01 0:11 ` Duy Nguyen
2015-02-28 11:36 ` Duy Nguyen
2015-02-28 1:01 ` [RFC/PATCH 3/5] connect.c: connect to a remote service with some flags Stefan Beller
2015-02-28 11:11 ` Torsten Bögershausen
2015-03-01 3:22 ` Junio C Hamano
2015-02-28 1:01 ` [RFC/PATCH 4/5] daemon.c: accept extra service arguments Stefan Beller
2015-03-01 3:39 ` Junio C Hamano
2015-02-28 1:01 ` [RFC/PATCH 5/5] WIP/Document the http protocol change Stefan Beller
2015-02-28 12:26 ` Duy Nguyen
2015-03-01 9:11 ` [RFC/PATCH 0/5] protocol v2 for upload-pack Johannes Sixt
2015-03-02 2:58 ` Junio C Hamano
2015-03-02 3:47 ` [RFC/PATCH 0/3] protocol v2 Junio C Hamano
2015-03-02 9:21 ` Duy Nguyen
2015-03-02 9:24 ` Duy Nguyen
2015-03-03 10:33 ` Duy Nguyen
2015-03-03 17:13 ` Junio C Hamano
2015-03-03 19:47 ` Junio C Hamano
2015-03-04 1:54 ` Duy Nguyen
2015-03-04 4:27 ` Shawn Pearce
2015-03-04 12:05 ` Duy Nguyen
2015-03-04 19:10 ` Shawn Pearce
2015-03-05 1:03 ` Stefan Beller
2015-03-05 16:03 ` Stefan Beller
2015-03-24 17:42 ` Stefan Beller
2015-03-24 18:00 ` Junio C Hamano
2015-03-06 23:38 ` [PATCH] protocol upload-pack-v2 Stefan Beller
2015-03-06 23:40 ` Stefan Beller
2015-03-06 23:55 ` Duy Nguyen
2015-03-07 0:00 ` Duy Nguyen
2015-03-07 0:28 ` Junio C Hamano
2015-03-07 4:28 ` Stefan Beller
2015-03-07 5:21 ` Duy Nguyen
2015-03-08 20:36 ` Junio C Hamano
2015-03-31 19:58 ` Junio C Hamano
2015-04-02 12:37 ` Duy Nguyen
2015-04-02 14:45 ` Junio C Hamano
2015-04-02 22:18 ` Martin Fick
2015-04-02 22:58 ` Junio C Hamano
2015-04-02 23:00 ` Stefan Beller
2015-04-02 23:14 ` Stefan Beller
2015-03-10 1:38 ` Duy Nguyen
2015-03-10 19:36 ` Kyle J. McKay
2015-02-28 0:07 ` [RFC/PATCH 0/3] protocol v2 Duy Nguyen
2015-02-28 0:26 ` Junio C Hamano
2015-03-01 8:41 ` Junio C Hamano
2015-03-01 11:32 ` Duy Nguyen
2015-03-01 19:56 ` Stefan Beller
2015-03-02 1:05 ` David Lang
2015-03-01 23:06 ` Junio C Hamano
2015-03-02 1:09 ` David Lang
2015-03-02 3:10 ` Junio C Hamano
2015-03-01 23:06 ` Philip Oakley
2015-03-02 9:32 ` Duy Nguyen
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=xmqqioenhs4p.fsf@gitster.dls.corp.google.com \
--to=gitster@pobox$(echo .)com \
--cc=git@vger$(echo .)kernel.org \
--cc=pclouds@gmail$(echo .)com \
--cc=sbeller@google$(echo .)com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox