HP3000-L Archives

January 2000, Week 4

HP3000-L@RAVEN.UTC.EDU

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
"Bailie, Jack" <[log in to unmask]>
Reply To:
Bailie, Jack
Date:
Wed, 26 Jan 2000 13:42:22 -0600
Content-Type:
text/plain
Parts/Attachments:
text/plain (34 lines)
My experience is that there is a lot of overhead in the FCONTROL 6 - I
usually do it in the writer process to make sure it's updated, but don't
usually worry with it in the reader process.

Jack

Jeff wrote:

                We have a system that normally logs to a message file, and a
logging
                collector process copies this information to tape.  However,
during
                bulk loads, we don't bother with copying the info out, it
just needs
                to go away.  The [third-party] system is doing this in the
'dummy'
                collector by enabling extended wait, reading a record,
issuing an
                FCONTROL 6 (write EOF), and go back to start.

                From looking at SOS, it appears to be writing to disk A
WHOLE LOT and
                not reading much of anything, e.g., 8 reads, 632 writes in
30 secs.
                It appears there is A LOT of overhead in trashing out a
message file
                this way.  Would it be more efficient to just sit there and
loop on
                the FREAD?  I believe it is, but thought I'd ask before
second-guessing
                a fairly-well established application.



ATOM RSS1 RSS2