Would it be beneficial, performance wise, to delete unused fields in invoice and PO’s and other commonly run reports? Especially in larger batch reports like invoices.
Typically I leave everything alone, but as we grow and batches are getting larger, 50-75 invoices a night, I wonder if not making the report query for data that will never be there in the first place is slowing them down. I don’t have issues, just looking for some performance tuning ideas. We are a factory that only sells B2B and never on payment plans, so there is a lot of tax and payment schedule stuff that never gets used, yet the sub-reports are always being run. I’ve even thought of changing labels to text fields rather than having to look to a data field for it. If the performance gains are hardly noticeable then i probably wont even bother with it.
I’d guess that there few (if any) fields in a standard RDD, that aren’t in the SSRS RDL (or at least used in the RDL for sorting, calculating, etc…)
If there are features that you never use, you could copy the RDD and exclude those fields, then tweak a copy of the RDL accordingly.
For example, if you never use misc charges, you could exclude those fields in the RDD, and remove them from the RDL’s query expression, and any report objects that reference them.
But the savings would be minimal, as the datasets for the tax would be practically empty. And the risk of eventually needing those fields, would outway the savings.
In my experience, most of the degradation is usually caused by unnecessary excessive amount of groups and using SSRS to calculate things that could easily be accomplished by SQL. On the other hand, very very large dataset (hundreds, “yes I used the plural version” of columns) which occurs from clicking that Sync Dataset button will most definitely slow down rendering. It’s really a fine balance in the end, but if you remove something I would test the report out with every possible option to make sure that what you removed is not really being utilized in a suppressed section.