I have a table with about 200,000 records (Invoice line items) and I have an index tag on the Invoice number and Deleted(). Sys(3054,1) indicates full optimization. What I observe is when you include a 'group by' Invoice (multiple items per invoice) clause, having an index tag on Deleted() has no effect on performance. The query was very simple:

select invoice from lineitms group by invoice

Anyone else observer this as well. Any other optimization tips? All I wanted was a unique list of invoice numbers in this table.

Thanks