Designing a database schema

[Rashid Sarwar]

I am having problem designing a schema because of large data storage.I have to calculate aggregates on this schema.
Below is schema

CREATE TABLE group_1sec_1 (tagid mediumint, value varchar(256), Currtime datetime, quality smallint, PRIMARY KEY(tagid, time), INDEX I_TagID_Time(tagid, time))

Columns attribute

Time: Sequentially incrementing by 1 sec.
Tag ID: Row identification.
Quality: Fixed range .(like 1-40)
Value: It can be Boolean, Integer, Float, String having 256 charaters.(temporarily storing 4 bytes for test purpose)

Requirments: Log 100,000 records per second for 6 months.
(I record row size 21 bytes excluding index size).

Disk Space required = 1,55,52000 * 100,000 * 21 = 30,416 GB

Please suggest schema or technique that can reduce my storage requirement with out effecting performance efficiency.

can i use partitioning along with compression to reduce storage and query performance.

I would like to have a quick reply as deadline to solve the problem is drawing close.