Return-Path: Delivered-To: apmail-jakarta-ant-user-archive@apache.org Received: (qmail 5656 invoked from network); 27 Feb 2002 15:15:32 -0000 Received: from unknown (HELO nagoya.betaversion.org) (192.18.49.131) by daedalus.apache.org with SMTP; 27 Feb 2002 15:15:32 -0000 Received: (qmail 7889 invoked by uid 97); 27 Feb 2002 15:15:24 -0000 Delivered-To: qmlist-jakarta-archive-ant-user@jakarta.apache.org Received: (qmail 7873 invoked by uid 97); 27 Feb 2002 15:15:24 -0000 Mailing-List: contact ant-user-help@jakarta.apache.org; run by ezmlm Precedence: bulk List-Unsubscribe: List-Subscribe: List-Help: List-Post: List-Id: "Ant Users List" Reply-To: "Ant Users List" Delivered-To: mailing list ant-user@jakarta.apache.org Received: (qmail 7858 invoked from network); 27 Feb 2002 15:15:24 -0000 Message-ID: <6F539391E018D3119B570008C75D91E806F7998E@SV018442> From: "Koegel, Michael" To: Ant Users List Subject: AW: Win2K vs Sun Date: Wed, 27 Feb 2002 16:15:20 +0100 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2653.19) Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable X-Spam-Rating: daedalus.apache.org 1.6.2 0/1000/N X-Spam-Rating: daedalus.apache.org 1.6.2 0/1000/N Hi Kyle, have you tried jikes? I would recommend it as build compiler! Is Linux an option? My own experience is that a Pentium 4 makes a difference. Are you = talking about PIII or PIV with your W2K-Machine. Are you using Sun or IBM JDK? I don't know for the compiler, but = normally IBM is faster. Regards, Michael -----Urspr=FCngliche Nachricht----- Von: Kyle Adams [mailto:kadams@gfs.com] Gesendet am: Mittwoch, 27. Februar 2002 16:07 An: ant-user@jakarta.apache.org Betreff: Win2K vs Sun Some background: my company is migrating all of our mainframe=20 applications to Java. Our integration builds (think continuous = integration) currently produce about 200 jars (with more being added every week), = run=20 twice a day, and take ~50 min on the WinNT 4 workstation they're = running=20 on. We're moving the integration build to a server so we can build faster,=20 build more, and build more frequently. We're looking at running the = builds=20 on a server running either Solaris 2.8 or Win2K, but we don't have much = (any) experience running large Java builds on a high-end machine. My question to ant users out there - what are your experiences running = large builds on either of these two platforms? Does either one seem = inherently=20 better performance-wise for Java builds? Why? Here are the results = from=20 our preliminary testing: Baseline (current) workstation WinNT 4, 512 MB RAM, 750 MHz processor Here are the specs on platforms under consideration: Compaq Proliant 380DL, Generation 2 Win2K, 2GB RAM, Dual 1.2 Ghz processors Sun Netra T1 Solaris 2.8, 1GB RAM, 440 MHz processor Sun E450 Solaris 2.8, 2GB RAM, Quad 480 MHz processors Results of building a test suite on each platform, using JDK 1.3 and = javac: Baseline: 14 min Compaq: 7.5 min Netra T1: 23 min E450: 21 min The dual processors on the Compaq both took about a 58% load during the = build, while the processors on both the baseline and Netra T1 took a = 100%=20 load. The E450 showed a single processor taking a 100% load, with no=20 load on the other processors. It was kind of surprising to see both Sun boxes run slower than the = baseline WinNT 4 workstation. That's why I'm looking for advice from others who = have experience running large Java builds. Thanks, Kyle Adams Java Developer Gordon Food Service -- To unsubscribe, e-mail: = For additional commands, e-mail: = -- To unsubscribe, e-mail: For additional commands, e-mail: