EPICS Controls Argonne National Laboratory

Experimental Physics and
Industrial Control System

1994  1995  <19961997  1998  1999  2000  2001  2002  2003  2004  2005  2006  2007  2008  2009  2010  2011  2012  2013  2014  2015  2016  2017  2018  2019  2020  2021  2022  2023  2024  Index 1994  1995  <19961997  1998  1999  2000  2001  2002  2003  2004  2005  2006  2007  2008  2009  2010  2011  2012  2013  2014  2015  2016  2017  2018  2019  2020  2021  2022  2023  2024 
<== Date ==> <== Thread ==>

Subject: Re: Standard IOC Benchmarks?
From: [email protected] (Noboru Yamamoto)
To: [email protected]
Date: Thu, 19 Dec 1996 08:32:21 +0900
Marty Kraimer wrote:

>
>Peregrine McGehee wrote:
>>
>> Aloha,
>>         Does anyone have a 'standard' set of IOC performance
>> benchmarks? I mean besides the obvious running of dbt against a number
>> of 'soft' records.
>>
>    #3          16-OCT-1996 22:38:34.34                                   
>EPICS
>>         I'd like to generate some meaningful numbers for my Sparc 5s.
>>
>> Thanks,
>>         Peregrine
>
>
>anonymous ftp to epics.aps.anl.gov. Then
>
>cd pub/dbperform
>
>retrieve the files
>
>perform.db
>timeMeter.adl
>
>
>The adl display brings up 100 meters each connected to a different
>record. In addion the left side of screen has 10 menu controls
>that allow you to set scan rate for records in that row.
>
>At bottom are three menu controllers. Two allow you to set scan
>rate for groups of 5 rows. The last sets scan rate for ALL meters.
>This last on is connected to a fanout record that cause everything
>to scan 3 times each time record is executed. Thus setting thi
>to .1 second cause 3000 records/sec to process with monitors
>going to display.
>
The "mrkfunout" record in perform.db had had only two links, which caused 
1000 records/sec record processing with .1 second periodic scan.

To cause 3000 recods/sec processing, you should replace the difinition
of "mrkfount" record in perform.db as:

record(fanout, "mrkfanout")
{
        field(LNK1, "mrkf0.VAL ")
        field(LNK2, "mrkf1.VAL ")
        field(LNK3, "mrkf0.VAL ")
        field(LNK4, "mrkf1.VAL ")
        field(LNK5, "mrkf0.VAL ")
        field(LNK6, "mrkf1.VAL ")
}

It should be also noted that CPU usage of ScanPeriod depends on the 
number of client. Number of timemeter.adl displays, which are 
running at the measurement,  does matter to compare the results of
benchmark from other Labs.

>
>Run this and then start spy. Give results.
>
>
>Marty Kraimer
>

I thank Marty for the discussion on these points.

Regards,

Noboru
Noboru Yamamoto
KEKB Accelerator Control Group
KEK, JAPAN
Fax:+81-298-64-0321
Tel:+81-298-64-5309
e-mail:[email protected]



Navigate by Date:
Prev: Re: Bug in sequencer? Rozelle Wright
Next: Should dfanout be called dLoFanout? Rozelle Wright
Index: 1994  1995  <19961997  1998  1999  2000  2001  2002  2003  2004  2005  2006  2007  2008  2009  2010  2011  2012  2013  2014  2015  2016  2017  2018  2019  2020  2021  2022  2023  2024 
Navigate by Thread:
Prev: Re: Standard IOC Benchmarks? Marty Kraimer
Next: [no subject] Xiaosong GENG
Index: 1994  1995  <19961997  1998  1999  2000  2001  2002  2003  2004  2005  2006  2007  2008  2009  2010  2011  2012  2013  2014  2015  2016  2017  2018  2019  2020  2021  2022  2023  2024 
ANJ, 10 Aug 2010 Valid HTML 4.01! · Home · News · About · Base · Modules · Extensions · Distributions · Download ·
· Search · EPICS V4 · IRMIS · Talk · Bugs · Documents · Links · Licensing ·