Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 14 Sep 2015 15:18:29 +0300
From:      Dmitry Sivachenko <trtrmitya@gmail.com>
To:        Slawa Olhovchenkov <slw@zxy.spb.ru>
Cc:        FreeBSD Stable ML <stable@freebsd.org>
Subject:   Re: process scheduling and cpuset
Message-ID:  <6FCB48B4-02BD-4048-A3A5-3F86A25F5386@gmail.com>
In-Reply-To: <20150913164057.GT21849@zxy.spb.ru>
References:  <623FA99E-04E7-4D29-953A-61EE7B35CBF6@gmail.com> <20150913130920.GR3158@zxy.spb.ru> <C6ADA4B7-1126-44A5-94B3-97FA79C8582A@gmail.com> <20150913164057.GT21849@zxy.spb.ru>

next in thread | previous in thread | raw e-mail | index | archive | help

> On 13 =D3=C5=CE=D4. 2015 =C7., at 19:40, Slawa Olhovchenkov =
<slw@zxy.spb.ru> wrote:
>=20
> On Sun, Sep 13, 2015 at 04:44:40PM +0300, Dmitry Sivachenko wrote:
>=20
>>=20
>>> On 13 =D3=C5=CE=D4. 2015 =C7., at 16:09, Slawa Olhovchenkov =
<slw@zxy.spb.ru> wrote:
>>>=20
>>> On Sun, Sep 13, 2015 at 02:52:08PM +0300, Dmitry Sivachenko wrote:
>>>=20
>>>> Hello,
>>>>=20
>>>> I have 32 processor machine (2x CPU E5-2650) running several =
CPU-bound processes (ULE scheduler).
>>>> 3 processes are 32-threaded, and 8 are single threaded.
>>>>=20
>>>> I bind all 3 32-threaded processes to CPUs 0-24 (cpuset -C -l 0-24 =
-p XXX).
>>>>=20
>>>> I expect that the remaining 8 single-threaded processes will =
(mostly) run on the remaining 25-31 CPU cores and use (almost) 100% cpu =
each.
>>>>=20
>>>> But this is not the case (according to top(1)):  they spend a lot =
of time on 0-24 CPUs and CPU Idle time is about 10%.
>>>>=20
>>>> These are all purely computational programs, in idle system =
single-threaded programs steadily consume 100% of a core, and =
32-threaded programs consume all 32 cores and idle time is zero.
>>>>=20
>>>> Is it an ULE scheduler feature or am I doing something wrong?
>>>>=20
>>>> The goal is to give a single-threaded program a chance to run when =
somebody started several 32-threaded processes.
>>>=20
>>> You don't have 32 processor machine, you have only 16 processor
>>> machine.
>>> SMT/hyperthreading don't give real processor, SMT "CPU" have
>>> unpredicable power and his load depend on load parent CPU.
>>>=20
>>> For example, for my case I see such condition (simpliy) on CPU 0 and =
1
>>> (SMT of one real core) with rise load:
>>>=20
>>> load 0.1  0.1
>>> load 0.2  0.2
>>> load 0.3  0.3
>>> load 0.4  0.4
>>> load 0.45 0.45
>>> load 0.48 0.48
>>> load 1.00 1.00\
>>=20
>>=20
>> Yes I know about HT.  But how does this explain why I have 10% of CPU =
idle?
>>=20
>> If I explicitly bind my single-threaded processes to the remaining =
CPU cores (25-32), they start to receive expected 100% of CPU and =
overall Idle decreases.
>>=20
>> I just expect scheduler to do the same for me.
>>=20
>=20
> Idle is not goal, goal is lessing task executing time.


Thanks for the explanation.

In my example SMT pairs are numbered with sequential numbers, so 0+1 is =
one SMT group, 2+3 is second SMT group, and so on.

So in 25-32 range there are several real CPU cores which remain idle =
while processes are fighting for overloaded 0-24.

When I explicitly pin my single-threaded processes to 25-32 range, they =
start to receive 100% of CPU (and finish faster to be clear).




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?6FCB48B4-02BD-4048-A3A5-3F86A25F5386>