home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky comp.protocols.nfs:1988 comp.protocols.tcp-ip.ibmpc:4543
- Path: sparky!uunet!sun-barr!news2me.ebay.sun.com!seven-up.East.Sun.COM!tyger.Eng.Sun.COM!geoff
- From: geoff@tyger.Eng.Sun.COM (Geoff Arnold @ Sun BOS - R.H. coast near the top)
- Newsgroups: comp.protocols.nfs,comp.protocols.tcp-ip.ibmpc
- Subject: Re: PC-NFS 4.0, RTM, Heap size
- Date: 31 Jul 1992 17:48:25 GMT
- Organization: SunSelect
- Lines: 24
- Message-ID: <15bud9INNel@seven-up.East.Sun.COM>
- References: <1992Jul30.143320.5227@doc.ic.ac.uk>
- NNTP-Posting-Host: tyger.east.sun.com
-
- Quoth kjw@doc.ic.ac.uk (Kevin J Walters) (in <1992Jul30.143320.5227@doc.ic.ac.uk>):
- #
- #There are quite a few computers i've come across which are loading rtm
- #with the parameter '/heap 64'. I had a look in the manuals and found
- #the value to be 64k, as i suspected, so basically rtm is using up 64k
- #plus the resident portion of that oh-so-precious ram.
- #
- #The manual gives a defualt of 22k if the heap option isn't specified - is
- #there any reason for a heap as large as 64k?
- #
- #I suspect it all comes down to efficiency, but there must be a more
- #reasonable balance between memory use and nfs performance.
- ^^^
- Minor clarification: the heap size does not affect NFS performance: only
- the performance of Toolkit apps. (Remember that in PC-NFS the NFS
- client is embedded, along with some RPC/XDR and a subset of UDP) in
- the PCNFS.SYS driver.
-
- Geoff
- --
- Geoff Arnold, PC-NFS architect | In the years after WW2 it was said that Oxford
- (geoff.arnold@East.Sun.COM) | philosophers had found a cure for atheism.
- SunSelect, | Young people coming up as atheists soon found
- a Sun Microsystems Business | that they no longer knew what to disbelieve...
-