Path: utzoo!utgpu!news-server.csri.toronto.edu!helios.physics.utoronto.ca!ists!yunexus!geac!torsqnt!tmsoft!becker!bdb From: bdb@becker.UUCP (Bruce Becker) Newsgroups: comp.os.misc Subject: Re: (Free) book on the Amoeba distributed system available Message-ID: <6548@becker.UUCP> Date: 24 Mar 90 00:20:48 GMT References: <6067@star.cs.vu.nl> <2604@quiche.cs.mcgill.ca> <6106@star.cs.vu.nl> Reply-To: bdb@becker.UUCP (Bruce Becker) Organization: G. T. S., Toronto, Ontario Lines: 50 In article <6106@star.cs.vu.nl> ast@cs.vu.nl (Andy Tanenbaum) writes: |In article <2604@quiche.cs.mcgill.ca> utility@quiche.cs.mcgill.ca (Ronald BODKIN) writes: |>Sounds like Amoeba is quite close to QNX. QNX also has the virtual machine, |>the same kernal design, the same high-speed features. And QNX has years |>of a good track record. |> Ron | | |I thought QNX was basically a UNIX-clone. You log into one specific machine |and work there, with access to others via rlogin and explicit file motion. |Correct me if I am wrong. You are about to be corrected 8^) - QNX is a fully distributed message-passing system with remote task execution and resource access as an integral part of the system. |In Amoeba you just log in. When you say the equivalent of ls -l, you see |a bunch of files. Some might be in your machine room, and the rest might |be spread over a dozen sites in as many countries. You can't even tell |where they are unless you look hard. When you start make, it runs all |the compilations on as many processors as it can find, in parallel. This |is what I mean by a distributed system--there is no concept of a home machine |where work is done by default. There is just a big pool of resources which |the system allocates automatically. Is QNX like that? More or less. All resources can be allocated to a job, including cpu, memory, files, disk drives, modems, etc. The difference is that you can tell (until the new version this year perhaps) where the resource lives, but you treat it locally (except for diskette drives - inserting a diskette halfway across a continent is a bit inconvenient 8^). One can get an executable from one machine, execute it on another, and have the screen/kybd i/o sent to your own machine - all simple to do, built in to the system. The name space isn't as global as the way you describe, in the current version - I imagine better name space management might come with the POSIX version due out this spring... Cheers, -- ,u, Bruce Becker Toronto, Ontario a /i/ Internet: bdb@becker.UUCP, bruce@gpu.utcs.toronto.edu `\o\-e UUCP: ...!uunet!mnetor!becker!bdb _< /_ "Paranoia is its own reward" - W. Disney