[Orca-dev] Orcallator.se changes and additions

[Dmitry Berezin]

Well, this means that the problem is not local to just _my_ servers and is
indeed a bug in SE, which causes data to be occasionally incorrect.
Just wondering - do you run Oracle database(s) on this server?

> -----Original Message-----
> From: Jon Tankersley [mailto:jon.tankersley at eds.com]
> Sent: Tuesday, May 18, 2004 3:35 PM
> To: 'Dmitry Berezin'
> Subject: RE: [Orca-dev] Orcallator.se changes and additions
> 
> So far, I've had only a couple of mismatches and one bad hit since you
> sent
> me the file....
> 
> 14:13:00    103    103    103
> 14:13:11    103    103    103
> 14:13:21    103    103    103
> 14:13:31    103    103    103
> 14:13:41    103    103    103
> 14:13:52    106    106     45 **
> 14:14:02    110    110    110
> 14:14:12    107    107    107
> 14:14:23    107    107    107
> 14:14:33    107    107    107
> 14:14:43    107    107    107
> 
> -----Original Message-----
> From: Dmitry Berezin [mailto:dberezin at surfside.rutgers.edu]
> Sent: Tuesday, May 18, 2004 1:16 PM
> To: 'Jon Tankersley'
> Cc: orca-dev at orcaware.com
> Subject: RE: [Orca-dev] Orcallator.se changes and additions
> 
> 
> Jon,
> 
> > I can test the proc code if you want.
> 
> Here you go (test program is attached).
> 
> By default, it will use 10 seconds as the interval, but you can specify
> any
> other interval on the command line (se proctest.se <seconds>).
> 
> The program will print out the results of counting processes and will
> print
> one or more '*', signifying the magnitude of the difference, if the
> results
> are different. Since it runs two counting loops sequentially, you might
> see
> some single digit differences, which is normal, but eventually, you should
> get the _huge_ difference that demonstrates the problem I've described
> earlier.
> 
> Here is the sample output (I just ran it on my system). Notice the results
> at the end.
> 
> ****************************************
> 
> NP - pp.nproc
> C1 - pp.index$ process count
> C2 - first_proc/next_proc process count
> 
> Interval is 10 second(s)
> 
>     Time     NP     C1     C2
> 13:50:08    411    411    411
> 13:50:19    412    412    411 *
> 13:50:29    412    412    412
> 13:50:40    412    412    412
> 13:50:51    412    412    412
> 13:51:01    424    424    424
> 13:51:12    413    413    413
> 13:51:23    415    415    415
> 13:51:33    418    418    418
> 13:51:44    418    418    418
> 13:51:55    417    417    416 *
> 13:52:06    412    412    412
> 13:52:16    412    412    412
> 13:52:27    412    412    412
> 13:52:38    412    412    412
> 13:52:48    412    412    413 *
> 13:52:59    412    412    412
> 13:53:10    412    412    412
> 13:53:20    412    412    413 *
> 13:53:31    412    412    412
> 13:53:42    414    414    414
> 13:53:52    416    416    415 *
> 13:54:03    413    413    415 *
> 13:54:14    414    414    414
> 13:54:24    414    414    414
> 13:54:35    415    415    415
> 13:54:46    415    415    415
> 13:54:57    415    415    415
> 13:55:07    417    417    417
> 13:55:18    419    419    419
> 13:55:29    415    415    415
> 13:55:39    415    415    415
> 13:55:50    415    415    415
> 13:56:01    426    426    426
> 13:56:11    413    413    413
> 13:56:22    413    413    413
> 13:56:33    411    411    411
> 13:56:43    411    411    411
> 13:56:54    411    411    411
> 13:57:04    416    416    100 ***
> 13:57:15    414    414    415 *
> 13:57:26    414    414    412 *
> 
> ****************************************
> 
>   -Dmitry.
>