John Hicks wrote: >I don't see a question here. > >But that won't stop me from giving a little advice :) > >It is generally more important to keep things simple (by not >denormalizing) than to try to optimize performance by complicating >things significantly. > >Moreover, I can't see how combining several columns into one will >improve performance. I would think it will slow things down whenever you >have to retrieve data, particular if you query against anything in column 3. > >And now you say you want to save disk space by compressing the field >separators in your combined column? > >Forget it all! Go back to a fully normalized design. If you have >problems, post them here > >
I kind of disagree on what you said regarding denormalization, but believe me when I say that I have experienced a 90% improvement on performance with that. As I said before, my table has +20 million entries; if it was normalized this number would be around 20 billion, since it would be a 1 N relation. Off course I dont make any selections based on column 3, but only by the tables keys. Forget that!!! Runing for normalization would not be viable for me. I need a response time lower than 0.01 sec. (and Ive been achieving less than that) However I would like to make a better use of this columns space, once I use two only characters for separators. Heres my question: Is there anyway I could minimize that? Is there any specific character that would occupy less space? Once again thank you very much ============== Atenciosamente, Jan Gomes - [EMAIL PROTECTED] >Jan Gomes wrote: >> Hy Guys, >> >> I needed denormalized my table to obtain high performance, but i want best >> appropriate the >space. >> >> I joint two column (of the JOIN) intro one column with two separadores (# >> and ;) >> >> Example: >> ID | column_1 | column_denormalized >> 1 | Test | 1#20202;5#1000101; >> >> It has some method to minimize the space(disk space) required for this >> separadores ? Like >some character that i >> can use for minimize the table size? >> >> PS: The table has 20.000.000 of rows with 2 GB data length. -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
