Tremendous number of reads

Hi guys, any ideas how to investigate or determine the casue for the following ? Platform Solaris 9 on sparc, Informix 9.4FC5 New application causes a very high cpu-load, and we see about 2BILLION reads on a specific index in 1 day (more then 4 million in a minute) for 2 other tables we have more then 200 million reads a day. These figures seem not to be normal, but how to determine what's happening ? The first thing I recommended was to perform update statistics (not sure it's done already), furtheron I noticed that the customer changed the config and set NUMCPUVP to 1 and added 2 cpuvps dynamically afterwards. He also removed the AIOVPS (now about 20 aio-vps are running) Originally it was configured with VP class.... (1 cpu, max 3 cpu, and 2 aiovps because kio is doing the work) Can this be related with the parameters ? Example : Table lkrqs lkwts dlks touts isrd iswrt isrwt isdel BFRD bfwrt seqsc hit f/petlib_idx1 0x40018f 4545536 0 0 0 1995152617 0 0 0 2275444058 0 0 100 F/epi 0x4000be 25870167 2 1 0 213467 839 306 0 334726097 1482 2162 100 i/labres2 0x900094 131465200 0 0 0 136682358 0 0 0 212853488 35497 0 100 I/systables 0x100106 97396570 0 0 0 35555087 0 0 0 177179723 0 18 100 I/labautoj 0x900083 193754783 0 0 0 7746851 39965 44695 8429 162642226 100470 13 02 100 i/art1 0x50005f 53482232 1 0 0 156163337 0 0 0 162586579 56 0 100 i/rvdeta2 0x800039 151015709 0 0 0 150968665 0 0 0 152915304 3984 0 100 i/labdem6 0x90007c 54736331 6 0 0 68242561 0 0 0 144385016 949 0 100 i/preausr2 0x80001a 78961586 0 0 0 113647076 0 0 0 119989504 163 0 100 i/artfour2 0x500063 19928680 0 0 0 112186262 0 0 0 116731050 41 0 100 I appreciate every feedback ! Thanks, Jacques Lapeire