What to buy for cheap Zope host?
Hello everyone! I'm building a computer to train & test Zope. I hope to use PostgreSQL or MySQL for storing a good-sized database. Zope would be used for allowing users to enter data & for output of formated querries. I wish to spend no more than $700 on this new computer I'm building (not including sound & video card + monitor). What should I spend the most money on, CPU, memory, HD, or a balance? I'm leaning towards a PPGA Celeron 366 (possibly overclocked), buying a 128MB stick of P100 RAM, and the fastest ATA drive $120 can buy. Is this the right direction? Ram's rather expencive now, or else I'd get 256MB right away. thanks for your time. --== Sent via Deja.com http://www.deja.com/ ==-- Share what you know. Learn what you don't.
I wish to spend no more than $700 on this new computer I'm building (not including > sound & video card + monitor). What should I spend the most money on, CPU,
memory, HD, or a balance? My school just bought a Pentium 450 with128MB for a Zope server (hehe, I forced them to use Zope :) ). I use PostGres on it as well. Runs like a rocket for test and development purposes. regards, stephan -- Stephan Richter iXL - Software Designer and Engineer CBU - Physics, Computer Science and Chemistry Student
On 14 Nov 1999 19:40:00 -0600, tommy_b@my-deja.com wrote:
I'm building a computer to train & test Zope. I hope to use PostgreSQL or MySQL for storing a good-sized database. Zope would be used for allowing users to enter data & for output of formated querries.
I wish to spend no more than $700 on this new computer I'm building (not including sound & video card + monitor). What should I spend the most money on, CPU, memory, HD, or a balance?
We recently ran some tests on three machines to give us an idea of what we needed. The ApacheBench queries were done with my Zope PedSys (genetic database) Publisher product, which monitors the filesystem for changes to wrapped tabular objects. I.e. it isn't as FS-intensive as a relational database engine, but it is more so than most Zope applications. With the results I got (I didn't publish the extended test results as they are harder to evaluate), under light-to-moderate loads such as what you would see under development purposes, I'd shoot for CPU speed first (running unix with at least 128MB ram), OS second, disk I/O third (especially where database applications are involved), and memory over 128MB fourth. If your database tables are large, I'd probably bump up disk I/O and then ram in the list a notch. -------------------------------------------- 140 MHz Ultra 170 running Solaris 5.7 512MB ram: Python compiled with SUNWspro/SC5.0 1 Pystone - 3 separate times : This machine benchmarks at 1776.2 pystones/second This machine benchmarks at 1782.53 pystones/second This machine benchmarks at 1779.36 pystones/second 4 Pystones simultaneously: This machine benchmarks at 1801.8 pystones/second This machine benchmarks at 1769.91 pystones/second This machine benchmarks at 1785.71 pystones/second This machine benchmarks at 1763.67 pystones/second 4-2G striped Barracudas (bonnie): -------Sequential Output-------- ---Sequential Input-- --Random-- -Per Char- --Block--- -Rewrite-- -Per Char- --Block--- --Seeks--- MB K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU /sec %CPU 100 2170 96.8 11282 30.7 10998 35.0 1543 99.5 102978 99.6 5601.0 70.0 ApacheBench, Version 1.2 : Complete requests: 10 Total transferred: 121600 bytes Files on local drives: Requests per second: 0.70 Transfer rate: 8.56 kb/s received Files via NFS : Requests per second: 0.71 Transfer rate: 8.63 kb/s received -------------------------------------------------- 400MHz K6-2 running NetBSD with 128MB ram: Python compiled with gcc 1 Pystone - 3 separate times : This machine benchmarks at 5434.78 pystones/second This machine benchmarks at 5524.86 pystones/second This machine benchmarks at 5464.48 pystones/second 4 Pystones simultaneously: This machine benchmarks at 5494.51 pystones/second This machine benchmarks at 5464.48 pystones/second This machine benchmarks at 5524.86 pystones/second This machine benchmarks at 5434.78 pystones/second ccd0a (2 striped) ibm 4.5G 7200rpm scsi-2 UW (bonnie): -------Sequential Output-------- ---Sequential Input-- --Random-- -Per Char- --Block--- -Rewrite-- -Per Char- --Block--- --Seeks--- MB K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU /sec %CPU 100 15688 85.3 15252 27.4 4266 8.2 10796 51.3 11186 11.1 110.4 1.9 100 15696 86.3 15293 26.8 4667 8.7 11485 55.0 12397 12.0 114.7 1.9 Files on local drives: --------------------- This is ApacheBench, Version 1.3 Complete requests: 10 Total transferred: 120240 bytes Requests per second: 2.23 Transfer rate: 26.83 kb/s received Files via NFS: --------------------- ApacheBench, Version 1.2 : Complete requests: 10 Total transferred: 120150 bytes Requests per second: 2.36 Transfer rate: 28.35 kb/s received -------------------------------------------------- 333MHz Ultra 5 running Solaris 5.7 with 256MB ram: Python compiled with SUNWspro/SC5.0 1 Pystone - 3 separate times : This machine benchmarks at 3831.42 pystones/second This machine benchmarks at 3875.97 pystones/second This machine benchmarks at 3861 pystones/second 4 Pystones simultaneously: This machine benchmarks at 3861 pystones/second This machine benchmarks at 3861 pystones/second This machine benchmarks at 3816.79 pystones/second This machine benchmarks at 3816.79 pystones/second Ultra5 8G EIDE (bonnie): -------Sequential Output-------- ---Sequential Input-- --Random-- -Per Char- --Block--- -Rewrite-- -Per Char- --Block--- --Seeks--- MB K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU /sec %CPU 100 4925 96.5 12174 16.3 4355 12.0 3445 98.9 12153 15.0 116.7 2.3 Data mounted via local fs: --------------------- This is ApacheBench, Version 1.3 Complete requests: 10 Total transferred: 121620 bytes Requests per second: 1.55 Transfer rate: 18.87 kb/s received
participants (3)
-
kent＠tiamat.goathill.org -
Stephan Richter -
tommy_b＠my-deja.com