From eacg at cin.ufpe.br  Wed Jul 25 19:12:29 2012
From: eacg at cin.ufpe.br (Erico Augusto Cavalcanti Guedes)
Date: Wed, 25 Jul 2012 16:12:29 -0300
Subject: Change log rate output
Message-ID: <CAKPcEv2Zh3aHmVrp45FUocfLcYL_0Z86mKo_dObuZjZxpJHE9Q@mail.gmail.com>

Hello,

during execution of experiments, log output is generated at every 7-9 seconds:
000.84| i-FirstRound  19740 293.03    337  54.68   0  100
000.92| i-FirstRound  21374 326.76    278  56.79   0  100
001.00| i-FirstRound  23061 337.44    310  58.68   0  100
001.09| i-FirstRound  25018 391.39    278  54.88   0  100
001.17| i-FirstRound  26994 395.21    267  57.95   0  100
001.25| i-FirstRound  28766 354.39    292  55.98   0  100
001.34| i-FirstRound  30734 393.61    261  57.77   0  100
001.42| i-FirstRound  32572 367.56    267  54.95   0  100
001.50| i-FirstRound  34631 411.80    236  55.80   0  100
001.59| i-FirstRound  36768 427.44    242  54.28   0  100

Is there a way to customize log rate output ?
I double checked man for pgserver and pgclient, without success.

Thanks in advance,

Erico.


From dmitry.kurochkin at measurement-factory.com  Wed Jul 25 19:44:00 2012
From: dmitry.kurochkin at measurement-factory.com (Dmitry Kurochkin)
Date: Wed, 25 Jul 2012 23:44:00 +0400
Subject: Change log rate output
In-Reply-To: <CAKPcEv2Zh3aHmVrp45FUocfLcYL_0Z86mKo_dObuZjZxpJHE9Q@mail.gmail.com>
References: <CAKPcEv2Zh3aHmVrp45FUocfLcYL_0Z86mKo_dObuZjZxpJHE9Q@mail.gmail.com>
Message-ID: <87vchbpqwv.fsf@gmail.com>

Hi Erico.

Erico Augusto Cavalcanti Guedes <eacg at cin.ufpe.br> writes:

> Hello,
>
> during execution of experiments, log output is generated at every 7-9 seconds:
> 000.84| i-FirstRound  19740 293.03    337  54.68   0  100
> 000.92| i-FirstRound  21374 326.76    278  56.79   0  100
> 001.00| i-FirstRound  23061 337.44    310  58.68   0  100
> 001.09| i-FirstRound  25018 391.39    278  54.88   0  100
> 001.17| i-FirstRound  26994 395.21    267  57.95   0  100
> 001.25| i-FirstRound  28766 354.39    292  55.98   0  100
> 001.34| i-FirstRound  30734 393.61    261  57.77   0  100
> 001.42| i-FirstRound  32572 367.56    267  54.95   0  100
> 001.50| i-FirstRound  34631 411.80    236  55.80   0  100
> 001.59| i-FirstRound  36768 427.44    242  54.28   0  100
>
> Is there a way to customize log rate output ?
> I double checked man for pgserver and pgclient, without success.
>

You can stat cycle period using --stats_cycle option for
polygraph-client(1) and polygraph-server(1), e.g. --stats_cycle 10sec.
Default is 5sec.

Note that this option does not just control the console output.  It sets
interval stats period.  The less this period is, the more frequently
interval stats are collected.  Hence, more precise interval stats at
expense of larger binary log size.

Regards,
  Dmitry

> Thanks in advance,
>
> Erico.
> _______________________________________________
> Users mailing list
> Users at web-polygraph.org
> http://www.web-polygraph.org/mailman/listinfo/users


From eacg at cin.ufpe.br  Wed Jul 25 20:10:34 2012
From: eacg at cin.ufpe.br (Erico Augusto Cavalcanti Guedes)
Date: Wed, 25 Jul 2012 17:10:34 -0300
Subject: Change log rate output
In-Reply-To: <87vchbpqwv.fsf@gmail.com>
References: <CAKPcEv2Zh3aHmVrp45FUocfLcYL_0Z86mKo_dObuZjZxpJHE9Q@mail.gmail.com>
	<87vchbpqwv.fsf@gmail.com>
Message-ID: <CAKPcEv0iiRva0_ACbGaX+g76aR7d3SEhGLkzKfUd-9pV5U+jAQ@mail.gmail.com>

Hi Dmitry,

I had already tried this before(without success):

$ polygraph-client --config WebWorkload.pg --stats_cycle 5sec
--verb_lvl 10 --log TestStatistcsClient.bin >> TestStatistcsClient.log

snippet of TestStatistcsClient.log:
...
001.67| i-FirstRound  38988 444.00    221  53.38   0  100
001.75| i-FirstRound  41065 415.38    251  55.22   0  100
001.84| i-FirstRound  43235 434.03    222  58.71   0  100
001.92| i-FirstRound  45410 434.99    220  56.92   0  100
002.00| i-FirstRound  47449 407.79    232  56.40   0  100
002.09| i-FirstRound  49464 403.00    250  56.08   0  100
002.17| i-FirstRound  51573 421.78    218  57.85   0  100
002.25| i-FirstRound  53249 335.23    276  52.68   0  100
002.34| i-FirstRound  55097 369.60    256  56.17   0  100
002.42| i-FirstRound  56671 314.80    289  58.64   0  100
002.50| i-FirstRound  58444 354.60    309  58.54   0  100
002.59| i-FirstRound  60111 333.41    273  54.89   0  100
...

Capture rate suffers no change ... :-/


Thanks again,

Erico.


From dmitry.kurochkin at measurement-factory.com  Thu Jul 26 22:23:09 2012
From: dmitry.kurochkin at measurement-factory.com (Dmitry Kurochkin)
Date: Fri, 27 Jul 2012 02:23:09 +0400
Subject: Change log rate output
In-Reply-To: <CAKPcEv0iiRva0_ACbGaX+g76aR7d3SEhGLkzKfUd-9pV5U+jAQ@mail.gmail.com>
References: <CAKPcEv2Zh3aHmVrp45FUocfLcYL_0Z86mKo_dObuZjZxpJHE9Q@mail.gmail.com>
	<87vchbpqwv.fsf@gmail.com>
	<CAKPcEv0iiRva0_ACbGaX+g76aR7d3SEhGLkzKfUd-9pV5U+jAQ@mail.gmail.com>
Message-ID: <87ehnyp3g2.fsf@gmail.com>

Hi Erico.

Erico Augusto Cavalcanti Guedes <eacg at cin.ufpe.br> writes:

> Hi Dmitry,
>
> I had already tried this before(without success):
>
> $ polygraph-client --config WebWorkload.pg --stats_cycle 5sec
> --verb_lvl 10 --log TestStatistcsClient.bin >> TestStatistcsClient.log
>
> snippet of TestStatistcsClient.log:
> ...
> 001.67| i-FirstRound  38988 444.00    221  53.38   0  100
> 001.75| i-FirstRound  41065 415.38    251  55.22   0  100
> 001.84| i-FirstRound  43235 434.03    222  58.71   0  100
> 001.92| i-FirstRound  45410 434.99    220  56.92   0  100
> 002.00| i-FirstRound  47449 407.79    232  56.40   0  100
> 002.09| i-FirstRound  49464 403.00    250  56.08   0  100
> 002.17| i-FirstRound  51573 421.78    218  57.85   0  100
> 002.25| i-FirstRound  53249 335.23    276  52.68   0  100
> 002.34| i-FirstRound  55097 369.60    256  56.17   0  100
> 002.42| i-FirstRound  56671 314.80    289  58.64   0  100
> 002.50| i-FirstRound  58444 354.60    309  58.54   0  100
> 002.59| i-FirstRound  60111 333.41    273  54.89   0  100
> ...
>
> Capture rate suffers no change ... :-/
>

It works as expected.  Please note that time is printed on console in
minutes, it is not min.sec format.  I.e. "002.09" is not "2 minutes 9
seconds" but "2 and 9 hundredths minutes".

Regards,
  Dmitry

>
> Thanks again,
>
> Erico.