EPICS Home EPICS Controls Argonne National Laboratory

Experimental Physics and
Industrial Control System

Format page
for printing


Search Tech-talk
1994  1995  1996  1997  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  1996  1997  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: cron-type scheduling in database or sequencer?
From: Andrew Johnson <[email protected]>
To: Kay-Uwe Kasemir <[email protected]>
Cc: tech talk <[email protected]>
Date: Thu, 14 Sep 2006 09:43:08 -0500
Hello Kay,

Kay-Uwe Kasemir wrote: 

Both the database and the sequencer have good support for
periodic scheduling (scan field resp. delay() routine).
But I now have a need for scheduling things at certain times,
for example "at 00:15 and then every 6 hours",
meaning: run at 00:15, 06:15, 12:15 and 18:15 each day.
... 
Has anybody done something like this?
PVs for configuring the time, plus sequencer code
for determining the next runtime etc.?

We have a 'console chime' subroutine record that we run on all our vxWorks IOCs here; it prints out the time on the IOC console at a selected interval but could be used as the basis of what you're after.

The default is to chime hourly on the hour, but it can be set to any minute period by changing the value in A - set it to 15 and you'll get chimes at 15, 30 and 45 minutes past the hour (not just at 15 minute intervals after the IOC started up - it's cleverer than that).

For your purposes you'd want to replace the 'printf' with a call to post_event(psub->b) and you will get the soft event number in B posted at the minute interval in A. The subroutines used are:


/* consoleChime.c,v 1.1.2.1 2004/12/23 16:53:44 anj Exp */

/* consoleChime field usage:
 *
 *  A = chime period in minutes, defaults to 60 if not set
 *  L = timestamp (secPastEpoc) when we last chimed
 */

#include <stdio.h>
#include <sysLib.h>
#include <shellLib.h>

#include "tsDefs.h"
#include "subRecord.h"

static char prompt[20];

long consoleChimeInit(subRecord *psub)
{
    TS_STAMP now;

if (psub->a < 1.0) psub->a = 60.0; /* Default to chiming hourly */

    tsLocalTime(&now);
    psub->l = now.secPastEpoch;

    sprintf(prompt, "%.13s> ", sysBootParams.targetName);
    shellPromptSet(prompt);

    return 0;
}

long consoleChime(subRecord *psub)
{
    epicsUInt32 period, next;
    TS_STAMP now;

    if (psub->a < 1.0) psub->a = 1.0;  /* Chime at most once a minute */
    period = 60 * (epicsUInt32) psub->a;
    next = period * (1 + (epicsUInt32) psub->l / period);

    tsLocalTime(&now);
    if (now.secPastEpoch >= next) {    /* Bong bong bong bong */
        TS_DETAIL chime;
        tsStampToLocal(now, &chime);
        printf("\n%4.4d-%2.2d-%2.2d %2.2d:%2.2d   %s",
                chime.year, chime.monthNum+1, chime.dayMonth+1,
                chime.hours, chime.minutes, prompt);
        psub->l = now.secPastEpoch;
    }
    return 0;
}

The record looks like this:

record(sub, "$(IOC):consoleChime") {
        field(DESC, "Print time on IOC console")
        field(INAM, "consoleChimeInit")
        field(SNAM, "consoleChime")
        field(SCAN, "10 second")
        field(PRIO, "LOW")
}


For those curious, the console chime allows us to see when any error messages that appear on an IOC's console actually occurred, since it may be days or even longer between someone looking at the IOC console and many errors don't get sent to the error logger. If course if it's weeks you would probably want to set the chime period to something like 1440=24 hours unless you have a *really* long history buffer on your screens sessions.

- Andrew
--
There is considerable overlap between the intelligence of the smartest
bears and the dumbest tourists -- Yosemite National Park Ranger
Replies:
Re: cron-type scheduling in database or sequencer? Andrew Johnson
References:
cron-type scheduling in database or sequencer? Kay-Uwe Kasemir
Navigate by Date:
Prev: cron-type scheduling in database or sequencer? Kay-Uwe Kasemir
Next: Re: cron-type scheduling in database or sequencer? Andrew Johnson
Index: 1994  1995  1996  1997  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: cron-type scheduling in database or sequencer? Kay-Uwe Kasemir
Next: Re: cron-type scheduling in database or sequencer? Andrew Johnson
Index: 1994  1995  1996  1997  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, 02 Sep 2010 Valid HTML 4.01! · Home · News · About · Base · Modules · Extensions · Distributions · Download ·
· Search · EPICS V4 · IRMIS · Talk · Bugs · Documents · Links · Licensing ·