Implements migration of a QUIC connection to a different destination address if specified by the se…

Implements migration of a QUIC connection to a different destination address if specified by the server in an ASAD tag in the Server Hello. Quietly falls back to original socket if creation and connecting a new socket fails.

R=rch@chromium.org
BUG=

Committed: https://crrev.com/217455a1456e1152b9d937073e51cabcb1baca79
Cr-Commit-Position: refs/heads/master@{#405262}

[Jana, 2016-07-06 06:44:14]

Description was changed from

==========
Implements migration of a QUIC connection to a different destination address if
specified by the server in an ASAD tag in the Server Hello. Quietly falls back
to original socket if creation and connecting a new socket fails.

R=rch@chromium.org
BUG=
==========

to

==========
Implements migration of a QUIC connection to a different destination address if
specified by the server in an ASAD tag in the Server Hello. Quietly falls back
to original socket if creation and connecting a new socket fails.

R=rch@chromium.org
BUG=
==========

[Ryan Hamilton, 2016-07-06 19:29:38]

That looked pretty straightforward!

https://codereview.chromium.org/2124753005/diff/1/net/quic/quic_stream_factor...
File net/quic/quic_stream_factory.cc (right):

https://codereview.chromium.org/2124753005/diff/1/net/quic/quic_stream_factor...
net/quic/quic_stream_factory.cc:1483: void
QuicStreamFactory::MigrateSessionToNewSocket(
(Alternative name "MigrateSession"?)

https://codereview.chromium.org/2124753005/diff/1/net/quic/quic_stream_factory.h
File net/quic/quic_stream_factory.h (right):

https://codereview.chromium.org/2124753005/diff/1/net/quic/quic_stream_factor...
net/quic/quic_stream_factory.h:304: NetworkChangeNotifier::NetworkHandle
network,
Can you document what happens when |network| is kInvalid?

Alternatively, consider making this method private and have two wrappers:
  MigrateSessionToNewNetwork
  MigrateSessionToNewPeerAddress

*shrug*

https://codereview.chromium.org/2124753005/diff/1/net/quic/quic_stream_factor...
File net/quic/quic_stream_factory_test.cc (right):

https://codereview.chromium.org/2124753005/diff/1/net/quic/quic_stream_factor...
net/quic/quic_stream_factory_test.cc:2677: // migrate to using a new socket to
the server's alternate address.
I don't think this comment is correct since there is no second socket data
provider yet?

https://codereview.chromium.org/2124753005/diff/1/net/quic/quic_stream_factor...
net/quic/quic_stream_factory_test.cc:2750: }
This seems like a test which might be a better fit with
QuicNetworkTransactionTests, but perhaps that's hard to arrange?

[Jana, 2016-07-12 22:34:10]

Thanks, Ryan -- comments addressed, PTAL.

https://codereview.chromium.org/2124753005/diff/1/net/quic/quic_stream_factor...
File net/quic/quic_stream_factory.cc (right):

https://codereview.chromium.org/2124753005/diff/1/net/quic/quic_stream_factor...
net/quic/quic_stream_factory.cc:1483: void
QuicStreamFactory::MigrateSessionToNewSocket(
On 2016/07/06 19:29:37, Ryan Hamilton wrote:
> (Alternative name "MigrateSession"?)

Done.

https://codereview.chromium.org/2124753005/diff/1/net/quic/quic_stream_factory.h
File net/quic/quic_stream_factory.h (right):

https://codereview.chromium.org/2124753005/diff/1/net/quic/quic_stream_factor...
net/quic/quic_stream_factory.h:304: NetworkChangeNotifier::NetworkHandle
network,
On 2016/07/06 19:29:37, Ryan Hamilton wrote:
> Can you document what happens when |network| is kInvalid?
> 
> Alternatively, consider making this method private and have two wrappers:
>   MigrateSessionToNewNetwork
>   MigrateSessionToNewPeerAddress
> 
> *shrug*

Good suggestion. Done.

https://codereview.chromium.org/2124753005/diff/1/net/quic/quic_stream_factor...
File net/quic/quic_stream_factory_test.cc (right):

https://codereview.chromium.org/2124753005/diff/1/net/quic/quic_stream_factor...
net/quic/quic_stream_factory_test.cc:2677: // migrate to using a new socket to
the server's alternate address.
On 2016/07/06 19:29:37, Ryan Hamilton wrote:
> I don't think this comment is correct since there is no second socket data
> provider yet?

Yes, that's right -- leftover text from previous incarnation. Removed second
sentence.

https://codereview.chromium.org/2124753005/diff/1/net/quic/quic_stream_factor...
net/quic/quic_stream_factory_test.cc:2750: }
On 2016/07/06 19:29:37, Ryan Hamilton wrote:
> This seems like a test which might be a better fit with
> QuicNetworkTransactionTests, but perhaps that's hard to arrange?

I think the test would basically end up testing the same code path, so unless
you feel strongly, I'm inclined to leaving it here with the other connection
migration tests.

[Ryan Hamilton, 2016-07-12 22:39:26]

lgtm, modulo one minor tweak

https://codereview.chromium.org/2124753005/diff/20001/net/quic/quic_stream_fa...
File net/quic/quic_stream_factory.h (right):

https://codereview.chromium.org/2124753005/diff/20001/net/quic/quic_stream_fa...
net/quic/quic_stream_factory.h:306: scoped_refptr<StringIOBuffer> packet);
Looks like the only caller of this method passes in nullptr here, so let's drop
this argument (and the impl of this method will pass nullptr to
MigrateSession()).

[Jana, 2016-07-12 23:21:53]

Added finch control -- PTAL

https://codereview.chromium.org/2124753005/diff/20001/net/quic/quic_stream_fa...
File net/quic/quic_stream_factory.h (right):

https://codereview.chromium.org/2124753005/diff/20001/net/quic/quic_stream_fa...
net/quic/quic_stream_factory.h:306: scoped_refptr<StringIOBuffer> packet);
On 2016/07/12 22:39:26, Ryan Hamilton wrote:
> Looks like the only caller of this method passes in nullptr here, so let's
drop
> this argument (and the impl of this method will pass nullptr to
> MigrateSession()).

Done.

[Ryan Hamilton, 2016-07-12 23:47:25]

lgtm

Sweet

[Jana, 2016-07-13 06:45:03]

The CQ bit was checked by jri@chromium.org

[commit-bot: I haz the power, 2016-07-13 06:45:20]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-07-13 07:06:55]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-07-13 07:06:56]

Try jobs failed on following builders:
  linux_chromium_compile_dbg_ng on master.tryserver.chromium.linux (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_...)

[Jana, 2016-07-13 18:47:54]

The CQ bit was checked by jri@chromium.org

[Jana, 2016-07-13 18:47:55]

The patchset sent to the CQ was uploaded after l-g-t-m from rch@chromium.org
Link to the patchset: https://codereview.chromium.org/2124753005/#ps80001
(title: "fixing linker error.")

[commit-bot: I haz the power, 2016-07-13 18:49:15]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-07-13 20:15:21]

Description was changed from

==========
Implements migration of a QUIC connection to a different destination address if
specified by the server in an ASAD tag in the Server Hello. Quietly falls back
to original socket if creation and connecting a new socket fails.

R=rch@chromium.org
BUG=
==========

to

==========
Implements migration of a QUIC connection to a different destination address if
specified by the server in an ASAD tag in the Server Hello. Quietly falls back
to original socket if creation and connecting a new socket fails.

R=rch@chromium.org
BUG=
==========

[commit-bot: I haz the power, 2016-07-13 20:15:23]

Committed patchset #5 (id:80001)

[commit-bot: I haz the power, 2016-07-13 20:16:44]

Description was changed from

==========
Implements migration of a QUIC connection to a different destination address if
specified by the server in an ASAD tag in the Server Hello. Quietly falls back
to original socket if creation and connecting a new socket fails.

R=rch@chromium.org
BUG=
==========

to

==========
Implements migration of a QUIC connection to a different destination address if
specified by the server in an ASAD tag in the Server Hello. Quietly falls back
to original socket if creation and connecting a new socket fails.

R=rch@chromium.org
BUG=

Committed: https://crrev.com/217455a1456e1152b9d937073e51cabcb1baca79
Cr-Commit-Position: refs/heads/master@{#405262}
==========

[commit-bot: I haz the power, 2016-07-13 20:16:45]

Patchset 5 (id:??) landed as
https://crrev.com/217455a1456e1152b9d937073e51cabcb1baca79
Cr-Commit-Position: refs/heads/master@{#405262}

[Ryan Hamilton, 2016-12-23 00:52:46]

On 2016/07/13 20:16:45, commit-bot: I haz the power wrote:
> Patchset 5 (id:??) landed as
> https://crrev.com/217455a1456e1152b9d937073e51cabcb1baca79
> Cr-Commit-Position: refs/heads/master@{#405262}

It looks like this change was not merged back to the internal repository. I'll
do so now, but please be mindful of back-merges.

[Jana, 2016-12-23 04:40:46]

On 2016/12/23 00:52:46, Ryan Hamilton wrote:
> On 2016/07/13 20:16:45, commit-bot: I haz the power wrote:
> > Patchset 5 (id:??) landed as
> > https://crrev.com/217455a1456e1152b9d937073e51cabcb1baca79
> > Cr-Commit-Position: refs/heads/master@{#405262}
> 
> It looks like this change was not merged back to the internal repository. I'll
> do so now, but please be mindful of back-merges.

Apologies!