I was wondering what the recommended amount of child portals are before it causes DNN trouble. We are having problems adding new child portals as it seems like that it times out, when doing SQL traces we see SQL loops around the portals and then times out.

I am in the process of deleting the child portals to see whether thats solves the problem. Currently we have roughly 200 child portals on one installation.

Thanks in advance

I know of a number of people running hundreds of portals, and a couple with a few thousand child portals (I met someone at openforce who ran 1600+ child portals) . The "problem" is that when dotnetnuke starts up (on first hit or after the cache is cleared at host level), there is a lot of work to draw the data and build the various objects , with quite a few database queries executed. This basically means your limit is the limit of your hardware i.e. is it large enough to rebuild this set of data within the 30 second timeout.

A number of amendments were made a few versions back to optimise this area, and an overload was added to the datacache to ensure a full refresh is only done when necessary (there were a few use cases where a partial expiration could work), so I don't think theres a huge amount more we can do to optimise this.

Cathal