Am 28.05.2014 22:29, schrieb Rajeev Prasad: > I am using phpMyAdmin, i looked closely and found "index" is this index you > are talking about? (earlier i used "Primary").
surely primary is a uniqe key, honestly consider to read some manuals > My further question is: the index key here is going to be epoch system time. > I currently > have it as integer 10. The table will grow very fast (about 5 million rows > within 12 months) > and will be updated atleast twice in 24 hours. So I have to consider the > price i am going to > pay to rebuild the index on each INSERT, (about ~42000 records every 12 > hours). you need to try it out in doubt > what would you/all suggest? > shoud I have an index key? all searches will be based on index key mostly, > and sometimes another 'unindexed' column. > should i change my index key to character type? > should i index the other 'unindexed column' also? > _or_ > should i simply go grabbing on an unindexed table? that are basic questions http://dev.mysql.com/doc/refman/5.5/en/optimization-indexes.html > On Wednesday, May 28, 2014 2:50 PM, Reindl Harald <h.rei...@thelounge.net> > wrote: > > Am 28.05.2014 21:43, schrieb Rajeev Prasad: >> I am going to have a big table with lot of records, to expedite searching i >> wanted to index on a key field (which is numeric value). BUT, there will be >> records which will have same value for the key field (other columns will be >> different). >> >> so how can i do this? right now, i am getting error, about duplicate entries >> and they are being discarded. All entries are important and I have to find a >> way to locate records based on this key field. > > who said that a key needs to be unique? > > just get phpMyAdmin to learn such things > there you see two different types of keys and after assign you get the sql > command
signature.asc
Description: OpenPGP digital signature
