RE: Optimizing Informix IO

Some info you might find helpful... I have 125 users (peak load of 300) and run with no performance issues (at least no database performance issues) Sar reports idle time in the 20's Our system: INFORMIX-OnLine Version 7.22.UC2 AIX 4.1.5 IBM RS/6000 R40 8 Processors 2GB memory IBM SSA Disk Array To answer your Q's: NUMAIOVPS: on my system it is 15 RAID: My SSA disks are configured (LVM) for striping. I use Informix mirroring. Memory: (I presume you mean buffers) I have 110000 (4K pages) 12 LRU queues 10 page flushers Checkpoints avg. 2 seconds, max 4 HTH... -----Original Message----- From: Narkinsky, Pat [mailto:Pat.Narkinsky@rivhs.com] Sent: Tuesday, April 07, 1998 9:27 AM To: informix-list@iiug.org Subject: Optimizing Informix IO System Manifest: INFORMIX-OnLine Version 7.14.UD1 AIX 4.3.0.0 IBM RS/6000 S70, 12 Processors, 8GB of memory IBM SSA Disk Array We are running an Informix Based financial application with approximately 150 simultaneous users, and have experienced persistent performance problems. Currently, we have thrown a ridiculous (IMHO) amount of processor at the problem and are getting adequate performance. However, we will soon be adding users, and I am concerned at the fact that we: 1) Have no upgrade path for hardware; we are at the top of the line now. 2) Are still experiencing 0% Idle Time (according to sar). 3) Have to add approximately 100 users in the next six months. Our problems are clearly IO related: examination of the output of sar indicate we are spending approximately 80% of our time waiting for IO. First question: we are running with NUMAIOVPS equal to 1. Could this be aggravating our IO problems? How can I tell if we are running KAIO? Second question: I've heard mention on this list to the effect "Informix doesn't like RAID". Could someone clue me in on the considerations here? Third: I've got 8GB of memory, however I am only using approximately 1.5GB. Would it be possible/advisable to try to get the engine to use more of that memory? Will this make my checkpoints last too long? Fourth: One of the busiest dbspaces on the system is "tempdb1" (according to 'onperf'). I have the following line in my onconfig file: DBSPACETEMP tempdb1:tempdb1:tempdb2 What is the significance of the double listing for tempdb1? Is this probably intentional, or is it the result of a typo on the part of my vendor? Would this account for the fact that tempdb1 is MUCH busier than tempdb2? Should I change it? Thanks! Patrick