Re: Informix Performance Issues

Art S. Kagel schrieb: [ ... snip ... ] > Read the manuals. Any single query that's actually performing > sequential reads of 1088 pages in a row will be using light scans which > neither use the RA_ parameters nor the buffer pool! Besides, unless the > disk farm serving this instance is JBOD and those disks and controllers > are very old, the drives are all performing their own readahead to their > own cache as is the SCSI controller (noone's running Informix on an ATA > controller right? <smirk>) itself to it's own cache. Actually - after some testing - we have excellent performance reading from 3ware Escalde Controllers serving 6 SATA disks each on PCI/X Excellent means > 75.000 2KB pages/sec when readingfrom 'disk', that is when diskreads are counting up. (This nice and cheap machine does 125.000 bufreads per second!) Needless to mention, that we do NOT configure the arrays to RAID5. My findings are that motivating mgmt to NOT going the RAID5 way is much easier when 1 GB costs 0,80 Eurtocent as it is with SATA! SCSI disks here are still around 8 Euro/GB, which is faktor of 10. But: What do I know. I do not dare to compare this SATA setup to whatever has disks inside and is expensive and slow at the samer time. Maufakturers prolly would take me to court. But you can test for yourself if you get similar performance out of your 'EMitachi700' Or you go the short way and simply test how much less than this amount of data rushes thru your supafast 2 Gbit zones on your fabric. *this was a bit sarcastic, yes I know* For those who are new here: I am the maniac who used ONARCHIVE for a long period of time on over 100 machines with great success. Especially ONDATARTR is what I miss a lot ... :) [ ... snip ... ] dic_k -- Richard Kofler SOLID STATE EDV Dienstleistungen GmbH Vienna/Austria/Europe