Date: Sun, 20 Jun 1999 14:34:47 +0200 (CEST) From: "O. Hartmann" <ohartman@ipamzlx.physik.uni-mainz.de> To: freebsd-smp@freebsd.org Subject: SMP, 4GB RAM, 4x CPU Message-ID: <Pine.BSF.4.10.9906201414380.28322-100000@ipamzlx.physik.uni-mainz.de>
next in thread | raw e-mail | index | archive | help
Dear Sirs. It is really hard to figure out how "good" the SMP implementation of FreeBSD is. I have had a lot of discussions with other FBSD users, developers, scientists using FBSD boxes and I got a summary of information which is confusing me. The situation in our lab is this: I built up a server for our NT Workstations to serve lots of drivespaces, building up a database and a webserver for our meteorological datas. Some guys from our theoretical meteorological group now want to use the SMP sytem (2x 350MHz Intel PII, 256 MB RAM). They use lots of fortran programs. The first tests offered that FreeBSD is not dispatching fortran jobs so jobs are running either on CPU0 or on CPU1. Well, it doesn't matter, we can start up two similar jobs to get the result in the time as we have used only the 350MHz system with UP ( :-( ). Now the discussion is going on to buy a new system for the whole institute (we are operating as small workgroups with a limited budget ...). For such enormous numbercrunching purposes we considered to buy (money rules, sorry, no Alpha!) a 4x SMP machine with 2 or 4GB RAM to server the necessary performance and to keep costs low -> so we have to rely on a free UN*X and I like FBSD because of its stability ... and I like BSD style systems. But that is not the question. We heard about tests, tests and tests again, made with the new Linux kernel (2.2x) and many provider offering so called "number crunching" Linux systems with 2 CPUs ( P III/550MHz). I read a lot about problems with DRAM growing up to 4GB and problems with 4 CPUs. I have problems with two - how big are then problems and "performance losses" with four CPUs ... Again, and again, I see so many unreflecting "performance tests" made by simply compiling the system. No, no, no. Well, listen to this: some guy wrote me, that he use an AMD K6-2/400MHz with 128MB SDRAM/100. It's system is ready within 45 minutes. That seems really fast! I have two 350MHz CPUs, 256 MBytes of RAM (upgrading to 512 MBytes next time) and my /usr/obj and /usr/src tree are on two different SCSI devices (UW, Adaptec 2940). When I tested the "make buildworld speed" of our "fast" system I always get a maketime of 90 to 100 minutes. That's funny, isn't it? Well, I tried make -j8, make -j12, make -j16 and lowered it to make -j5, but always the same result - and be aware of the fact, that the system is not used in the time of making world!!!! My question is, hopefuly, simple: I need objective and true informations about how "ggod" the SMP implementation of FreeBSD 3.2 is, how "stable" and usable the system is for usage with 4x CPU (Xeon) and 4GB RAM. We have some offers of Fortran vendors, and I don't want me spending a lot of money for a Linux- emulation to get not the power of the native system running on a Linux box. Where is the FreeBSD-SMP Roadmap? What has changed in FBSD 4.0? I don't want to start a polemic or philosophic discissuion, I need serious facts ... Thank you a lot in advance and thanks for all the guys who have written to me in the past. O. Hartmann Gruss O. Hartmann ------------------------------------------------------------------- ohartman@ipamzlx.physik.uni-mainz.de Klimadatenserver des IPA, Universitaet Mainz To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.4.10.9906201414380.28322-100000>