home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!ukma!darwin.sura.net!jhunix.hcf.jhu.edu!jhuvms.hcf.jhu.edu!ecf_stbo
- From: ecf_stbo@jhuvms.hcf.jhu.edu (Have unix ever /usr/local/bin experienced?)
- Newsgroups: comp.os.vms
- Subject: Re: Image activations, page faults and disk I/O
- Date: 28 Jan 1993 10:07 EST
- Organization: The Johns Hopkins University - HCF
- Lines: 20
- Distribution: world
- Message-ID: <28JAN199310073497@jhuvms.hcf.jhu.edu>
- References: <1993Jan26.155208.131143@mswe.dnet.ms.philips.nl> <1993Jan27.163409.1298@cmkrnl.com>
- NNTP-Posting-Host: jhuvms.hcf.jhu.edu
- News-Software: VAX/VMS VNEWS 1.41
-
- In article <1993Jan27.163409.1298@cmkrnl.com>, jeh@cmkrnl.com writes...
- >> because even on an empty VMS system the hard page faults keep the disk busy.
- >
- >Copy-on-reference pages being faulted in.
-
- What vms should do is make memory resident copies of the CRF sections so
- processes can soft fault out of the copy for even non shareable pages. The
- amount of hard faults generated on a busy general purpose system just by people
- firing up the commonly used utilities is HORRENDOUS, even though everything
- has been installed. 'Wasting' the memory to do this is worth it considering the
- savings in system disk I/O and probably cpu time. Resolving a hard fault even
- to a ram disk is CPU expensive. This could be implemented as a special
- qualifier to INSTALL, so sites without the extra memory (not that many, in this
- day and age) could override this behavior if they desired. Please do something
- like this, DEC! This is one of the biggest performance problems in VMS.
-
- Tom O'Toole - ecf_stbo@jhuvms.hcf.jhu.edu - JHUVMS system programmer
- Homewood Computing Facilities, Johns Hopkins University, Balto. Md. 21218
- >Here comes a jet ski.
- >weuh weeuhh weeuhh weeuhh WEEUHH WEEUHH WEEUHH WEEUHH weeuhh weeuhh weeuhh weuh
-