ID:	4590
From:	rn...@igor2.repo.hu
Date:	Sun, 29 Nov 2020 10:38:33 +0100 (CET)
Subject:	Re: [pcb-rnd] compile/test farm - poll on potential participation
in-reply-to:	4582 from John Griessen <jo...@cibolo.com>
replies:	4591 from TQ Hirsch <th...@thequux.com> , 4594 from John Griessen <jo...@cibolo.com>

Hello John,

On Sat, 28 Nov 2020, John Griessen wrote:

> On 11/28/20 6:39 AM, rnd2@igor2.repo.hu wrote:
>> Please let me know if you are interested in participating in CEC, ST or
>> MS, so I know which variants I should develop.
>
> I have two servers at atlantic.net USA Central 1 location, with 1 virtual CPU,
>
> 0.5 GB RAM,  20 GB disk each that can run "farmed" compiles and tests.
> Server 1. does web pages only,
> Server 1. does emails, DNS, email lists and very little web pages.

Thank you, this sounds excellent!

> 1.  CEC on a rented server slice with nice set so the server can still serve
> web pages somewhat

please show me the output of 

cat /proc/cpuinfo

on server 1!

With server slices like that, the virtual processor you get from your 
provider may not have capabilities to further virtualize. Without being 
able to virtualize, we shouldn't do a CEC: it would eat 100% CPU 
constantly, regardless of the emulated system being idle for most of the 
time. So if cpuinfo shows no virtualization capabilities, we should 
probably do an ST there too.

> 2.  ST on a rented server slice during 9PM - 8AM US Mountain time  (UTC is 7
> hours ahead of MST.)
> with nice set so the server can still serve web pages and postfix email
> processing somewhat.
>

Thanks! It seems the bottleneck would be your ram so I am going to select 
some old systems that can run on very low memory. Will take some time 
because I need to do a bit more on these systems to be able to hook up 
with the dispatcher.

Best regards,

Igor2

Mailing list archives : pcb-rnd

Navigation