Hi,
Am 30.06.2014 um 08:55 schrieb Derrick Lin:
> A typical node on our cluster has 64 cores and 512GB memory. So it's about
> 8GB/core. Occasionally, we have some jobs that utilizes only 1 core but
> 400-500GB of memory, that annoys lots of users.
This is a general question, how you judge a n
On Mon, 30 Jun 2014 14:19:23 -0700
Skylar Thompson wrote:
> h_vmem actually is scaled by slot, unless it is set to be per-"j"ob
> in its consumable attribute.
In our experiecen if you define the resource per job, the limit is not
respected. We had to redefine it as slot base and do some maths in
h_vmem actually is scaled by slot, unless it is set to be per-"j"ob in its
consumable attribute. UGE also adds a per-host option, not sure if the
other GE derivatives also have that. Your local complex(5) man page should
have the specifics.
On Mon, Jun 30, 2014 at 05:01:57PM -0400, Michael Stauffe
quot; of 40GB, while the "RES" is only 100MB.
>>
>> On Mon, Jun 30, 2014 at 3:01 PM, Michael Stauffer
>> wrote:
>> >> Message: 4
>> >> Date: Mon, 30 Jun 2014 11:53:12 +0200
>> >> From: Txema Heredia
>> >> To: Der
are also "in use" even though 49 are
>>> idle), and other jobs either run somewhere else, or queue up.
>>>
>>> Ian
>>>
>>> On Mon, Jun 30, 2014 at 12:01 PM, Michael Stauffer
>>> wrote:
>>>>> Message: 4
&
> shows "VET" of 40GB, while the "RES" is only 100MB.
>
> On Mon, Jun 30, 2014 at 3:01 PM, Michael Stauffer
> wrote:
> >> Message: 4
> >> Date: Mon, 30 Jun 2014 11:53:12 +0200
> >> From: Txema Heredia
> >> To: Derrick Lin , SGE Mai
ngle threaded, would have the same end result -
>> 400GB is in use (and 50 cores are also "in use" even though 49 are
>> idle), and other jobs either run somewhere else, or queue up.
>>
>> Ian
>>
>> On Mon, Jun 30, 2014 at 12:01 PM, Michael Stauffer
>&g
3:12 +0200
>>> From: Txema Heredia
>>> To: Derrick Lin , SGE Mailing List
>>>
>>> Subject: Re: [gridengine users] Enforce users to use specific amount
>>>of memory/slot
>>> Message-ID: <53b13388.5060...@gmail.com>
" of 40GB, while the "RES" is only 100MB.
>
> On Mon, Jun 30, 2014 at 3:01 PM, Michael Stauffer wrote:
>>> Message: 4
>>> Date: Mon, 30 Jun 2014 11:53:12 +0200
>>> From: Txema Heredia
>>> To: Derrick Lin , SGE Mailing List
>>>
t;> Date: Mon, 30 Jun 2014 11:53:12 +0200
>> From: Txema Heredia
>> To: Derrick Lin , SGE Mailing List
>>
>> Subject: Re: [gridengine users] Enforce users to use specific amount
>> of memory/slot
>> Message-ID: <53b13388.5060...@gmail
edia
>> To: Derrick Lin , SGE Mailing List
>>
>> Subject: Re: [gridengine users] Enforce users to use specific amount
>> of memory/slot
>> Message-ID: <53b13388.5060...@gmail.com>
>> Content-Type: text/plain; charset="iso-8859-1"; Format=
>
> Message: 4
> Date: Mon, 30 Jun 2014 11:53:12 +0200
> From: Txema Heredia
> To: Derrick Lin , SGE Mailing List
>
> Subject: Re: [gridengine users] Enforce users to use specific amount
> of memory/slot
> Message-ID: <53b13388.5060...@gmail.c
Hi Derrick,
You could either set h_vmem as a consumable (consumable=yes) attribute
and set a default value of 8GB for it. This way, whenever a job doesn't
request any amount of h_vmem, it will automatically request 8GB per
slot. This will affect all types of jobs.
You could also define a JSV
There is unfortunately no way in SGE to limit main memory.
h_rss / s_rss does not work with the rlimit call in Linux kernel version above
2.4.
Hence in Univa Grid Engine we introduced multiple ways for doing main memory
limitations. If you have cgroups support turned on then the cgroup takes car
Hi guys,
A typical node on our cluster has 64 cores and 512GB memory. So it's about
8GB/core. Occasionally, we have some jobs that utilizes only 1 core but
400-500GB of memory, that annoys lots of users. So I am seeking a way that
can force jobs to run strictly below 8GB/core ration or it should b
15 matches
Mail list logo
