Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!yale!mintaka!bloom-beacon!eru!hagbard!sunic!mcsun!cernvax!chx400!ethz!karrer From: karrer@ethz.UUCP (Andreas Karrer) Newsgroups: comp.unix.ultrix Subject: Re: how do you write a program to determine free swap (like pstat -s) Message-ID: <6517@ethz.UUCP> Date: 7 Nov 90 12:00:04 GMT References: <1990Nov6.033516.18772@NCoast.ORG> <1990Nov7.015529.15623@wrl.dec.com> <1927@shodha.enet.dec.com> Organization: ETH Zuerich Lines: 41 alan@shodha.enet.dec.com ( Alan's Home for Wayward Notes File.) writes: >In article <1990Nov7.015529.15623@wrl.dec.com>, mogul@wrl.dec.com (Jeffrey Mogul) writes: >> In article <1990Nov6.033516.18772@NCoast.ORG> simpsong@NCoast.ORG (Gregory R. Simpson @ The North Coast) writes: >> >I'd like to write a simple vm_check program that returns >> >nothing except the amount of swap remaining... (which is >> >all I'm interested in... >> >> As you might have guessed from the output of /etc/pstat -s, >> there isn't simple value stored in the kernel that says "this >> much swap space is in use"; rather, pstat grovels through >> kernel data structures counting things up, and computes >> a total through a fairly hairy computation. I counted over >> 160 lines in pstat just to do what you're asking for. > Actually it depends on what version of ULTRIX you're > running. In V3.1 and earlier. It's the really hairy > stuff that Jeff mentioned. In V4.0 there is a data > structure called "swapu" that has how much is being > used for various things. The amount free is just the > total - the amount used. You can get total from adding > amount indicated in the "swdevt" table. [... source for simplified pstat -s omitted ...] Thanks for this piece of code. (on Ultrix4.0) it returns the amount of used and wasted swap space using less than 0.1 sec cpu time (in good ol' 512 byte blocks). Please compare: "/bin/time /etc/pstat -s": Machine Ultrix V swap space cpu sec VAX 11/785 3.1 66 M 2.4 - well, a 785... DECstation 5000 4.0 65 2.3 - hmm... DECsystem 5400 4.0 122 3.8 - qualifies as "fairly hairy" DECsystem 5820 4.0 127 3.8 - ditto DECsystem 5840 4.0 156 6.1 - !!! outrageous !!! it seems the faster the machine, the slower is pstat -s :-( +----------- Andi Karrer, Communication Systems, ETH Zuerich, Switzerland karrer@ks.id.ethz.ch karrer@czheth5a.bitnet