I don't think the RAM footprint of a properly configured Linux OS is much of a concern on a compute node these days.  And by properly configured I mean unnecessary packages are not installed, and unnecessary services are not started in the first place. When I configure an OS install for a cluster, I go through every package one-at-a-time and determine whether or not I need it on a cluster node.

A bigger concern would be being able to disable as many unnecessary resources as possible to reduce OS overhead and therefore reduce 'OS jitter' in job performance. It amazes me with each new version of a distro how many more services seem to need to be running that you can't easily remove.

Prentice

On 2/1/21 9:46 AM, Jonathan Aquilina via Beowulf wrote:

Hi Tim,

I am asking from the non container side. If one wants to go bare metal wouldn’t that lower ram footprint that alpine provides be advantageous so you can maximize resource usage for what you want to crunch?

Regards,

Jonathan

*From:*Tim Cutts <t...@sanger.ac.uk>
*Sent:* 01 February 2021 15:43
*To:* Jonathan Aquilina <jaquil...@eagleeyet.net>
*Cc:* Beowulf <beowulf@beowulf.org>
*Subject:* Re: [Beowulf] apline linux [EXT]

The only place I regularly encounter is in containers, so people who are running singularity might find their users are using alpine in a way they can’t easily see.  Presumably the libm point Janne made applies in the container context as well, since the application will presumably be using the container’s libm, not the host node’s library.

Tim



    On 30 Jan 2021, at 05:56, Jonathan Aquilina via Beowulf
    <beowulf@beowulf.org <mailto:beowulf@beowulf.org>> wrote:

    Hey guys,

    Recently been giving alpine linux a try. I am super impressed on
    virtualized version of it what a low ram footprint it has 60mb
    starting. What do you guys think when it comes to alpine linux in
    the HPC space?

    Regards,

    Jonathan

    _______________________________________________
    Beowulf mailing list,Beowulf@beowulf.org
    <mailto:Beowulf@beowulf.org>sponsored by Penguin Computing
    To change your subscription (digest mode or unsubscribe)
    
visithttps://urldefense.proofpoint.com/v2/url?u=https-3A__beowulf.org_cgi-2Dbin_mailman_listinfo_beowulf&d=DwICAg&c=D7ByGjS34AllFgecYw0iC6Zq7qlm8uclZFI0SqQnqBo&r=gSesY1AbeTURZwExR_OGFZlp9YUzrLWyYpGmwAw4Q50&m=k_QXQ5VMTqsc46RivJoRKsv3-qKbVEVhs9N9jZy-Myc&s=98F3h1exjDbPRlv7rG0Aqry9txEvhut_Ujq46FeA-5s&e=
    
<https://urldefense.proofpoint.com/v2/url?u=https-3A__beowulf.org_cgi-2Dbin_mailman_listinfo_beowulf&d=DwICAg&c=D7ByGjS34AllFgecYw0iC6Zq7qlm8uclZFI0SqQnqBo&r=gSesY1AbeTURZwExR_OGFZlp9YUzrLWyYpGmwAw4Q50&m=k_QXQ5VMTqsc46RivJoRKsv3-qKbVEVhs9N9jZy-Myc&s=98F3h1exjDbPRlv7rG0Aqry9txEvhut_Ujq46FeA-5s&e=>

-- The Wellcome Sanger Institute is operated by Genome Research Limited, a charity registered in England with number 1021457 and a company registered in England with number 2742969, whose registered office is 215 Euston Road, London, NW1 2BE.


_______________________________________________
Beowulf mailing list, Beowulf@beowulf.org sponsored by Penguin Computing
To change your subscription (digest mode or unsubscribe) visit 
https://beowulf.org/cgi-bin/mailman/listinfo/beowulf

--
Prentice Bisbal
Lead Software Engineer
Research Computing
Princeton Plasma Physics Laboratory
http://www.pppl.gov

_______________________________________________
Beowulf mailing list, Beowulf@beowulf.org sponsored by Penguin Computing
To change your subscription (digest mode or unsubscribe) visit 
https://beowulf.org/cgi-bin/mailman/listinfo/beowulf

Reply via email to