Newsgroups: php.internals,php.qa
Path: news.php.net
Xref: news.php.net php.internals:63542 php.qa:66764
Mailing-List: contact internals-help@lists.php.net; run by ezmlm
Received-SPF: pass (pb1.pair.com: domain zerocue.com designates 74.115.204.53 as permitted sender)
To: Pierre Joye <pierre.php@gmail.com>
CC: zoe slattery <aparachic@gmail.com>, Nuno Lopes <nlopess@php.net>,
	"internals@lists.php.net" <internals@lists.php.net>, PHP QA
	<php-qa@lists.php.net>
Thread-Topic: [PHP-DEV] Re: [PHP-QA] Re: [PHP-DEV] Re: [PHP-QA] Parallel
 run-tests
Thread-Index: AQHNNxUY5wSrDrtsc0qE3a55AoIzN5bUGu8AgAFz8ICA1eXCAIAO992AgAGd6gCAAsZUAIABXDiAgAFD33A=
Date: Fri, 19 Oct 2012 11:47:19 +0000
Message-ID: <4D09E7E7-9499-4836-AF64-C6B607700DC7@zerocue.com>
References: <4FB4E844.2070305@gmail.com>
	<9AF025709ED6452F9E5A91834F82F788@PC07655>	<4FB9EE1E.3050701@gmail.com>
	<BCC3B4980516429D9BF61D910B09548D@PC07655>	<506E9442.6050501@gmail.com>
	<038CBF95DEC5479DA771E23528A33FB3@pc07654>	<507C7DE0.5030508@gmail.com>
	<507ED1BE.8040903@gmail.com>,<CAEZPtU4YLExeTcqRzyz3_rkXFsGLRy2Cmv5iKuyC=n-CULMyHA@mail.gmail.com>
In-Reply-To: <CAEZPtU4YLExeTcqRzyz3_rkXFsGLRy2Cmv5iKuyC=n-CULMyHA@mail.gmail.com>
Accept-Language: en-US
Content-Language: en-US
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0
Subject: Re: [PHP-DEV] Re: [PHP-QA] Re: [PHP-DEV] Re: [PHP-QA] Parallel
 run-tests
From: cpriest@zerocue.com (Clint Priest)

Could just as easily launch N sub-processes from run-tests.php as well.

I compile php in a VM with -j 40 on a dual core VM powered by quad core chi=
p...

The number of simultaneous tests doesn't need to line up with CPU counts...

-Clint

On Oct 18, 2012, at 7:28 AM, "Pierre Joye" <pierre.php@gmail.com> wrote:

> hi!
>=20
> On Wed, Oct 17, 2012 at 5:41 PM, zoe slattery <aparachic@gmail.com> wrote=
:
>> Nuno - just a PS to the last note. It is (mainly) the task allocation ac=
ross
>> processors which means that running tests in parallel on a 4-way machine=
 is
>> not 4 times as fast as running them in sequence.
>>=20
>> Here are some results from a run on my 2-way Mac -
>> http://static.inky.ws/image/3257/image.jpg. The blocks of colour are jus=
t
>> representations of the time it takes a group to run - and Open office is
>> allocating the colors randomly so they don't have any significance. I've
>> annotated the chart to show which groups are taking a long time....
>>=20
>> The net is that P0 runs its half of the tasks and then just hangs about
>> waiting for P1 to finish :-/.
>>=20
>> I have added a 'debug' flag to the code which will print information abo=
ut
>> how tasks are allocated if anyone wants to try on a 4 or 8 way. Given th=
at
>> there are not many groups that take a long time to run the simplest thin=
g
>> seems to be to map these to specific processors - that's easy enough and
>> requires no difficult code. There are more elegant solutions of course.
>=20
>=20
> Have you considered doing it using a daemon and php as client (to
> launch tasks and the tests)? Gearman or even a webserver could do a
> better job without the hassle to deal with what you are experiencing
> now. We do that using webservers in our labs to test all phpt and
> various apps.
>=20
> That being said, to multiply cores never bring a linear speed
> increase, even less in this case.
>=20
> Cheers,
> --=20
> Pierre
>=20
> @pierrejoye | http://blog.thepimp.net | http://www.libgd.org
>=20
> --=20
> PHP Internals - PHP Runtime Development Mailing List
> To unsubscribe, visit: http://www.php.net/unsub.php
>=20