[Web10g-user] Instrumentation frequency

[rapier]

So I took a few minutes... It doesn't have any error checking in it in 
case usleep is out of bounds but for now this should do what you need:

-i is the interval in milliseconds.

--- ../../../web10g-userland-read-only/util/C/web10g-watchvars.c 
2013-08-21 11:16:33.916865931 -0400
+++ web10g-watchvars.c	2013-09-12 13:04:50.215190026 -0400
@@ -45,6 +45,7 @@
  	estats_val_data* data = NULL;
  	int cid, i, j;
  	int opt, option;
+	int interval = 1000;

  	char *strmask = NULL;
          const char delim = ',';
@@ -67,7 +68,7 @@
                  exit(EXIT_FAILURE);
          }	

-        while ((opt = getopt(argc, argv, "hm:")) != -1) {
+        while ((opt = getopt(argc, argv, "hm:i:")) != -1) {
                  switch (opt) {
  		case 'h':
  			usage();
@@ -91,6 +92,9 @@
                          option = opt;

                          break;
+		case 'i':
+			interval = atoi(optarg);
+			break;
                  default:
                          exit(EXIT_FAILURE);
                          break;
@@ -152,7 +156,7 @@
                  }
  	}

-	sleep(1);
+	usleep(interval * 1000);
  	printf("\n\n");
  	}


On 9/12/13 12:51 PM, rapier wrote:
> Arvind,
>
> Sorry to take so long to reply. The frequency is controlled by a simple
> sleep statement in the main loop. I'll look at changing that to a user
> defined parameter in the near future. However, if it's critical you
> could change the sleep(1) to usleep(500000) for half second intervals,
> etc. and recompile.
>
> I'm out all of next week and I'm working on an annual report right now.
> If I can get the time to get a patch file for a user defined sleep
> interval I'll post it here. Otherwise I'll get something out the
> following week.
>
>
>
> On 9/12/13 12:09 AM, Arvind Mahla wrote:
>> Hi Rapier
>>
>> Is there any breakthrough in increasing frequency of reporting of
>> variables from 1 second to a lower value
>>
>> Arvind
>>
>>
>>
>> On Tue, Sep 3, 2013 at 4:37 AM, Chris Rapier <rapier at psc.edu
>> <mailto:rapier at psc.edu>> wrote:
>>
>>      I'll take a look at that tomorrow. One thing to keep in mind it's
>>      that there will likely be some tradeoffs in terms of performance
>>      when the granularity hits some point.
>>
>>
>>      Sent with AquaMail for Android
>>      http://www.aqua-mail.com
>>
>>
>>      On September 2, 2013 1:51:38 PM Arvind Mahla __ wrote:
>>
>>>      Hi,
>>>
>>>      Can we increase the frequency of watchvars from once every second
>>>      to once in some milliseconds.
>>>
>>>      Arvind
>>
>>
> _______________________________________________
> Web10g-user mailing list
> Web10g-user at web10g.org
> https://lists.psc.edu/mailman/listinfo/web10g-user
>
> To UNSUBSCRIBE visit https://lists.psc.edu/mailman/unsubscribe/web10g-user
>