Get rid of assertions on socket pools in HttpStreamFactoryBidirectionalQuicTest

Get rid of assertions on socket pools in HttpStreamFactoryBidirectionalQuicTest

This CL get rid of the assertions on socket pools in
RequestBidirectionalStreamImplQuicAlternative and 
RequestBidirectionalStreamImplHttpJobFailsQuicJobSucceeds 
unit test. The assertions are
flaky on Dr Memory bots. This is probably because even though 
we add hanging mock socket data for H2, TCP socket pools 
sometimes do get established.

BUG=611533

[xunjieli, 2016-05-12 21:12:03]

Description was changed from

==========
Get rid of assertions on socket pools in
HttpStreamFactoryBidirectionalQuicTest.RequestBidirectionalStreamImplQuicAlternative

This CL get rid of the assertions on socket pools in
RequestBidirectionalStreamImplQuicAlternative unit test. The assertions are
flaky on Dr Memory bots. This is probably because even though we add hanging
mock socket data for H2, TCP socket pools sometimes does get established.

BUG=611533
==========

to

==========
Get rid of assertions on socket pools in
HttpStreamFactoryBidirectionalQuicTest.RequestBidirectionalStreamImplQuicAlternative

This CL get rid of the assertions on socket pools in
RequestBidirectionalStreamImplQuicAlternative unit test. The assertions are
flaky on Dr Memory bots. This is probably because even though we add hanging
mock socket data for H2, TCP socket pools sometimes do get established.

BUG=611533
==========

[xunjieli, 2016-05-12 21:12:03]

xunjieli@chromium.org changed reviewers:
+ rch@chromium.org

[xunjieli, 2016-05-12 21:12:18]

Ryan, PTAL. Thanks!

[xunjieli, 2016-05-18 13:48:21]

Description was changed from

==========
Get rid of assertions on socket pools in
HttpStreamFactoryBidirectionalQuicTest.RequestBidirectionalStreamImplQuicAlternative

This CL get rid of the assertions on socket pools in
RequestBidirectionalStreamImplQuicAlternative unit test. The assertions are
flaky on Dr Memory bots. This is probably because even though we add hanging
mock socket data for H2, TCP socket pools sometimes do get established.

BUG=611533
==========

to

==========
Get rid of assertions on socket pools in HttpStreamFactoryBidirectionalQuicTest

This CL get rid of the assertions on socket pools in
RequestBidirectionalStreamImplQuicAlternative and 
RequestBidirectionalStreamImplHttpJobFailsQuicJobSucceeds 
unit test. The assertions are
flaky on Dr Memory bots. This is probably because even though 
we add hanging mock socket data for H2, TCP socket pools 
sometimes do get established.

BUG=611533
==========

[xunjieli, 2016-05-18 13:48:50]

On 2016/05/12 21:12:18, xunjieli wrote:
> Ryan, PTAL. Thanks!

Ping!

[Ryan Hamilton, 2016-05-19 19:38:32]

I don't think I understand what makes these assertions flaky on the bots. Why do
the unexpected sockets get created? Would it be possible to disable these
expectations only for those bots? Is there any sort of #if SOMETHING you could
wrap them in?

[xunjieli, 2016-05-20 17:07:58]

On 2016/05/19 19:38:32, Ryan Hamilton wrote:
> I don't think I understand what makes these assertions flaky on the bots. Why
do
> the unexpected sockets get created? Would it be possible to disable these
> expectations only for those bots? 

I was hoping you might know why the TCP sockets get created. I do not know
either :) These assertions are not crucial to the tests, so I thought that I
would just get rid of them. But since these two tests are already disabled on
these DrMemory bots, I am happy to abandon this CL instead. WDYT?

> Is there any sort of #if SOMETHING you could
> wrap them in?

[Ryan Hamilton, 2016-05-20 20:47:47]

On 2016/05/20 17:07:58, xunjieli wrote:
> On 2016/05/19 19:38:32, Ryan Hamilton wrote:
> > I don't think I understand what makes these assertions flaky on the bots.
Why
> do
> > the unexpected sockets get created? Would it be possible to disable these
> > expectations only for those bots? 
> 
> I was hoping you might know why the TCP sockets get created. I do not know
> either :) These assertions are not crucial to the tests, so I thought that I
> would just get rid of them. But since these two tests are already disabled on
> these DrMemory bots, I am happy to abandon this CL instead. WDYT?

Heh, works for me.

[xunjieli, 2016-05-20 20:53:15]

On 2016/05/20 20:47:47, Ryan Hamilton wrote:
> On 2016/05/20 17:07:58, xunjieli wrote:
> > On 2016/05/19 19:38:32, Ryan Hamilton wrote:
> > > I don't think I understand what makes these assertions flaky on the bots.
> Why
> > do
> > > the unexpected sockets get created? Would it be possible to disable these
> > > expectations only for those bots? 
> > 
> > I was hoping you might know why the TCP sockets get created. I do not know
> > either :) These assertions are not crucial to the tests, so I thought that I
> > would just get rid of them. But since these two tests are already disabled
on
> > these DrMemory bots, I am happy to abandon this CL instead. WDYT?
> 
> Heh, works for me.

Thanks. I will WontFix on the crbug then.