[Info-vax] show sys command output

[Jan-Erik Soderholm]

Den 2017-07-14 kl. 14:10, skrev VAXman- at SendSpamHere.ORG:
> In article <oka796$gku$2 at news.albasani.net>, Jan-Erik Soderholm <jan-erik.soderholm at telia.com> writes:
>> Den 2017-07-14 kl. 12:19, skrev VAXman- at SendSpamHere.ORG:
>>> In article <oka2lh$gku$1 at news.albasani.net>, Jan-Erik Soderholm <jan-erik.soderholm at telia.com> writes:
>>>> Den 2017-07-14 kl. 11:07, skrev Sergejus Zabinskis:
>>>>> Hi all,
>>>>>
>>>>>
>>>>> Maybe somebody knows exactly the meaning of last column (Pages) for this command:
>>>>>
>>>>> $ sh sys/node=abcd/own=dml_2016
>>>>>
>>>>>      Pid    Process Name    State  Pri      I/O       CPU       Page flts  Pages
>>>>> 2608C82C DML_SERVER_2016 HIB      6  4033997   0 00:17:18.93     16214  16662 M
>>>>>
>>>>> I get here value 16662 M for Pages column.
>>>>>
>>>>> And is it possible to get this value using DCL f$getjpi( pid, item ) call(s) ?
>>>>>
>>>>>
>>>>> Kind regards
>>>>> Sergejus
>>>>>
>>>>
>>>> Not sure for the "pages" (I had guessed WSSIZE, but I do not get the
>>>> same values even after scaling). But the "M" is item "MULTITHREAD"...
>>>
>>> $ WRITE SYS$OUTPUT (F$getjpi(PID,"GPGCNT")+F$getjpi(PID,"PPGCNT"))/(F$getsyi("PAGE_SIZE")/512)
>>>
>>
>> OK, thanks! Even if it wasn't me who asked... :-) I think I have noticed
>> before that the "pages" value didn't matched any single getjpi item.
> 
> 
> I know that you didn't ask it but I was continuing your reply which provided
> for the Multiprocessing bit.
> 

Of, course. And as I said, "thanks". :-)

Now, what about the other thing. Does the "Pages" output in any way
reflect the actual *physical* memory by that process? Or is the PPGCNT
alone a better value for that? GPGCNT does sound as some common memeory
shared by multiple processes...