hello

i would recommend the following formula:

FarmSize = (X * (C + 1))

where C is the number of Cores of the system.

where X is the threading/processing factor, which depends on factors like Intel 
vs AMD and Hyperthreading.

a conservative safe value for X is 2 -- for AMD you can assume up to 4 as safe.

so for 4 Cores you can safely configure a FarmSize of 10 (20 for AMD).

since Radiator does cooperative blocking on a select queue,
so processes that are not needed will only require a minuscle amount of cpu.

from operation experience, i have 8 and 16 core servers in production,
running multiple (8+) instances of Radiator with FarmSize from 32 to 128 each.

hope that helped

Yours sincerely

Alfred Reibenschuh

# Network System Management Platform Lead

Value Transformation Services GmbH
An IBM / KYNDRYL Company
Obere Donaustrasse 95
1020 Wien

Phone: +43-1-2056320-143
Mobile: +43-664-3523820
mail: 
[email protected]<mailto:[email protected]>
webex: https://kyndryl.webex.com/meet/alfred.reibenschuh

Please consider the environment before printing this e-mail.

This e-mail is confidential and may also contain privileged information. If you 
are not the intended recipient you are not authorized to read, print, save, 
process or disclose this message. If you have received this message by mistake, 
please inform the sender immediately and delete this e-mail, its attachments 
and any copies.
Any use, distribution, reproduction or disclosure by any person other than the 
intended recipient is strictly prohibited and the person responsible may incur 
penalties.

Thank you!

________________________________
From: radiator <[email protected]> on behalf of 
[email protected] <[email protected]>
Sent: Thursday, April 14, 2022 2:00 PM
To: [email protected] <[email protected]>
Subject: [EXTERNAL] radiator Digest, Vol 155, Issue 7

Send radiator mailing list submissions to
        [email protected]

To subscribe or unsubscribe via the World Wide Web, visit
        https://lists.open.com.au/mailman/listinfo/radiator
or, via email, send a message with subject or body 'help' to
        [email protected]

You can reach the person managing the list at
        [email protected]

When replying, please edit your Subject line so it is more specific
than "Re: Contents of radiator digest..."


Today's Topics:

   1. Re: Issue with authentication and farmsize option
      (Stefan Paetow (OpenSource))


----------------------------------------------------------------------

Message: 1
Date: Thu, 14 Apr 2022 01:16:06 +0100
From: "Stefan Paetow (OpenSource)" <[email protected]>
To: "[email protected]" <[email protected]>
Subject: Re: [RADIATOR] Issue with authentication and farmsize option
Message-ID:
        <cagmwa8r9pgdzgkqw4qx0yy8-ajaq5fvjuii04gvgnesxfor...@mail.gmail.com>
Content-Type: text/plain; charset="utf-8"

Hi Dubravko,

In the UK service, because we only have a limited number of processors
(with lscpu) on our new platform, we set our FarmSize to 3 times the lscpu
number, and it's certainly helped with throughput. So, I'd suggest setting
yours to 16, just to see how that improves the throughput?

Our old hardware actually had 12 and 16 processors (with 24 and 32 cores
respectively) and in those instances we literally got away with just
running the number of lscpu instances in the FarmSize.

Hope that helps.

Stefan


On Tue, 12 Apr 2022 at 07:59, Dubravko Penezic <[email protected]> wrote:

> Hi ,
>
> we run latest RADIATOR on VM with 4 processor and 4G RAM, and have set
> FarmSize to 4 (actually that mean 4 spare RADIATOR instance), usually we
> have about 80 request on all 4 instance per second, how ever in some
> pick time we reach 150+ authentication per second. In that period we
> have issue.
>
> In RADIATOR log we see no authentication request for second or two, then
> process continue normaly, and that may repeat few times in period of
> 5-10 minutes ... then everything continue to work normally.
>
> Any idea, suggestion ?
>
> Regards,
> Dubravko Penezic
> SRCE
> _______________________________________________
> radiator mailing list
> [email protected]
> https://lists.open.com.au/mailman/listinfo/radiator
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: 
<https://lists.open.com.au/pipermail/radiator/attachments/20220414/553cab61/attachment-0001.html
 >

------------------------------

Subject: Digest Footer

_______________________________________________
radiator mailing list
[email protected]
https://lists.open.com.au/mailman/listinfo/radiator

------------------------------

End of radiator Digest, Vol 155, Issue 7
****************************************
_______________________________________________
radiator mailing list
[email protected]
https://lists.open.com.au/mailman/listinfo/radiator

Reply via email to