I am wondering if someone with experience using various DNN versions can comment on performance differences.  I have a site built on DNN 5.6.2 with an 80 millisecond wait from server to get page html on a basic page with a few cached modules.  I have another using 6.1.3 that runs around 300 milliseconds on a similar basic page (no modules).  I have controlled for controls that are registered etc. in fact I stripped the page on the latter to have literally no controls except a menu and it still ran 230 ms.  We are doing further testing now but I don't feel comfortable with the difference and based on testing thus far I suspect it could be DNN version.  I see some posts about DNN 6 having better performance but no proof and certainly no metrics.  Also, the performance of this site, dotnetnuke.com, is significantly poorer than any DNN site we run and in fact so slow I would cringe at running a production site that performs similarly.  This makes me apprehensive about scalability.  Specifically, I wonder if there are some inefficiencies when using the DNN web farm, I can only assume this website uses it and could be behind the slowness.  Both servers our sites run on are plenty powerful enough (36 GB RAM between db and IIS server on the slower site, multiple quad core processors, RAID 10, etc.)  All of the typical performance best practices we already do and DNN and IIS settings are the same between both sites.  Ah, one other mentionable, the DNN 6 site uses Pro version and DNN 5 site uses CE.

We are going to narrow this down but I would really appreciate comments from anyone who has been down this road.

 Thanks :)

HI Cathal,

Thanks a lot for that, I had seen that article but didn't notice the page load speed graph with the various DNN versions, but there it is plain as day showing DNN 5.6.2 outperforming 6.1 so yes I think I was right to suspect that and our tests are pointing there as opposed to db as well.  Just by chance, would you or someone else happen to know what changes are resulting in performance loss, I would be interested in going into the core code and making some changes to increase performance but a list of culprits would be a really good start?

Thanks,

Jason