Hi guys first time poster. We have recently upgraded our application (medical application) from a pervasive SQL backend to a MS SQL backend. Since we (really the software vendor) did the db conversion we have noticed a major loss in performance. Now we did of course upgrade our hardware along with this change over but i'm not 100% sure what the problem is. i when configuring our environment followed MS/HP/Our software vendors best practices as best i could being some times all 3 did not recommend the same thing. EG: i configured our san and servers hardware as per HP the cluster as per MS/Vendor and SQL as per Vendor. Now i'm not the most MS SQL savy person but have been doing and will be doing more MS SQL courses along with the other staff which has helped. i've also been doing alot of reading on google and this site (this site was recommended by one of my course trainers). Now down to it. Our setup is 2 x HP DL580G5 4 x 6 core 2.67ghz with 32gb of ram DUAL Path 8GB FC Server 2008 R2 cluster. a HP EVA4400 with 32 x 15k FC disks with 2 x Brocade 8GB FC switches. i have spent the last few days logging the average disk latencys across the tempdb/mdf/ldf drives via windows as per Software vendor i have the MDF and LDF on different drives and TempDB also has it's own drive. alothugh the recommendation from HP is to have all these share the same spindles a hot topic i know. from 6am - 6pm (our main business hours) i am seeing an average of 4ms on the tempdb drive 4ms on the mdf drive and 1ms on the ldf drive. i am seing an average disk queue length of less then 1. CPU unt is averaging 3% total across an average business day. next i though the memory might not be sufficent but the buffer cache hit ratio is constantly on 100. Hopefully someone can offer some advice on what the problem maybe our DB is only 23gb in size the tempdb was sized according to vendor. the only thing i haven't done which the vendor recommend is split tempdb into multiple files per cpu. does this make a large amount of difference most of my research suggest this is more an sql 2000 thing? just looking for some advice to investigate further possible hardware issues before i go to our vendor to look at the DB. not sure if it helps but users report that at 6am when theres about 5 users on the application is just as slow as 11:30am when there is around 120 users connected. all this said reports that use to take 45minutes to run on our old pervasive sql db now run in about 5minutes on the new db.