Subject: | |
From: | |
Reply To: | |
Date: | Wed, 23 Jun 1999 00:04:33 -0500 |
Content-Type: | text/plain |
Parts/Attachments: |
|
|
> -----Original Message-----
> From: HP-3000 Systems Discussion [mailto:[log in to unmask]]On
> Behalf Of Bill Lancaster
> Sent: Tuesday, June 22, 1999 8:20 PM
> To: [log in to unmask]
> Subject: Re: System Log File Decoding
>
>
> Bob,
>
> If you have SOS from Lund, it has a little-known
> process-tracking-and-evaluation capability. If you want more information,
> call Lund (541-926-3800) or me (541-926-1542).
>
> Bill
I like Bill's suggestion to use the Lund product, but...
> > At 10:41 AM 6/22/99 -0400, Schlosser, Robert (Contractor) wrote:
[stuff deleted...]
> From what I can see, it appears that the newly created PIN number is
> left justified in the PID, with no indication as to how
> many bits are used for the PIN. I have looked in my limited
> documentation, but it is limited. Can anyone point me to the
> proper construction / extraction of the information.
AFAIK it is not documente, and we (HP) should provide a mapping routine...
The type used for a PID is as follows:
PROCID_INTERNAL_TYPE =
PACKED RECORD
MACHINE : SHORTINT;
PIN : BIT16;
FIELD2 : BIT32;
END
This means that the PIN number is the "second" 16-bits of the 64-bit PID.
Caveat: I did not look at the logfile code to be sure it is writting a
standard "PID" to the logfile. It should be pretty obvious from the
raw data. All valid PIN numbers fall in the range from 1..8190. Internally
PIN numbers are of type BIT16 (or at least are supposed to be). A "BIT16"
is an unsigned 16-bit integer.
|
|
|