I've seen some of the discussions already out there, but I've been told my case is somewhat different than the norm.
I'm working with a dataset of approximately 75 million records, all in one table. I use aggr and rank functions in calculated dimensions quite a bit. I know that these are very processor intensive. On my hard drive, with just me using it, it works, but somewhat slowly - a chart can take 4-5 minutes to load. I don't want to think what will happen when 20 users are all on it at the same time.
Any idea on what kind of server I'd need to make this work? My data file is a QVD file about 4 GB right now, although it may get larger.
Please don't just snd back the QlikTech recommendation, because that doesn't take into account all the aggregating and ranking I'm doing.
