My cube is taking 13 hours to process and I want to bring this time down to a couple of hours. I have 13 dimensions and four measure groups, I need all of them in one cube since the report is complicated and need all of them at once. I have 350 million records in one of the measure groups and growing rapidly, this is taking most of the processing time. I am doing a full process and cannot set a date to look only at changes since this is a relatively a new project (5 months) and records are still changing. I have eight partitions on each measure group. Most of my queries are simple select * from a table.
What can I do to improce processing time? It does not look like MSAS is made for huge data, it does not scale well, processing jumped from an hour of processing to 8 hours then to 13 hours. Sometimes I think about changing the technology to one of the new big data solutions or to do the queries in SQL Server and dump the cube!
Thanks,
Tarek