Column Store Index should be read/write by AndrewTuerk

While useful in warehousing scenarios and read only workloads, the primary capacity in which i would like to use a column store index is in a read/write scenario. I do not have the luxury (in most cases) of being able to have async updates. Thus, a read only index is not useful outside of warehousing scenarios. I do, however, have cases where the data i would index in a column or group of columns is highly duplicated. for example, in one table i have 5 columns, 100,000,000 records. distinct records per column:

column a - 1,000,000
column b - 12,000
column c - 3
column d - 50
column e - 3,000,000

Due to varying access patterns, there are multiple indexes on this table. Read/Write is about 20:1. In my particular case, i believe it would also be beneficial if i could have a writable clustered column store index. Enforcing uniqueness as a row based concept is still important on this table.