Hi The oft quoted figure for the effect of BES servers on Exchange is that BES will increase the amount of IOPS by almost four times. RIM's article here seems to confirm this: http://docs.blackberry.com/en/admin/deliverables/3532/Performance_Benchmarking_Guide.pdf "From the baseline, the largest increase occurred with all BlackBerry features (that were benchmarked) enabled. Disk transfers /sec grew from 77.52 to 368.97. Even though there was a 3.76-time increase in disk transfers /sec, this number is well below the maximum capacity that the disk subsystem can handle. Out of the individual features that were benchmarked, RIM observed that wireless email reconciliation, along with the core BlackBerry features enabled (email, calendar, sent item redirection) caused the largest increase in disk use. Disk transfers /sec grew from 77.52 to 335.82, which was a 3.33-time increase in disk transfers /sec. " That said I have heard people say that with each revision of BES, this figure lowers and lowers until it is practically the same as Outlook. Can anyone confirm this? Is this in reference to Exchange 2003 only, or more specifically to 2007 and 2010 where IO is handheld much more efficiently by Exchange anyway?

On Sat, 5 Jun 2010 18:29:04 +0000, Yoshi66 wrote: > > >Hi > >The oft quoted figure for the effect of BES servers on Exchange is that BES will increase the amount of IOPS by almost four times. > >RIM's article here seems to confirm this: > >http://docs.blackberry.com/en/admin/deliverables/3532/Performance_Benchmarking_Guide.pdf > >"From the baseline, the largest increase occurred with all BlackBerry features (that were benchmarked) enabled. Disk transfers /sec grew from 77.52 to 368.97. Even though there was a 3.76-time increase in disk transfers /sec, this number is well below the maximum capacity that the disk subsystem can handle. > >Out of the individual features that were benchmarked, RIM observed that wireless email reconciliation, along with the core BlackBerry features enabled (email, calendar, sent item redirection) caused the largest increase in disk use. Disk transfers /sec grew from 77.52 to 335.82, which was a 3.33-time increase in disk transfers /sec. " > > > >That said I have heard people say that with each revision of BES, this figure lowers and lowers until it is practically the same as Outlook. > >Can anyone confirm this? Is this in reference to Exchange 2003 only, or more specifically to 2007 and 2010 where IO is handheld much more efficiently by Exchange anyway? BES still generates the same amount of workload. What's changed it Exchange's ability to handle it. Grabbing numbers out of the air: If your typical Outlook user generates 0.4 IOPS in E2K3 then it'll probably do 0.2 IOPS in Exchange 2007. If BES generated 1.6 IOPS in E2K3 it would cause 0.8 in E2K7. The ratio's the same. What RIM downplays is that with the reduced IOPS/user a lot of servers now handle a lot more users. If you had, say, 3700 mailboxes on a 4GB E2K3 machine, you may have more than 5,000 on a 32GB E2K7 server. While the IOPS/user dropped, the number of users/server increased. The I/O load on the server may still be the same. --- Rich Matheisen MCSE+I, Exchange MVP --- Rich Matheisen MCSE+I, Exchange MVP