Brent Baisley wrote:
> It all depends on how complicated your data and searches are. I've got
> tables that add 2-3 million per day and I don't have performance
> problems. Although we only retain at most 500 millions records, not a
> full years worth.
> That said, you can get horrible performance out of mysql with tables as
> small as 100,000 records if you don't structure your queries correctly
> or use a good table structure. If I know the tables are going to grow
> quickly and I don't need the entire dataset all the time, I'll use merge
> tables. This makes it easy to remove "old" data easily from the default
> table set.
Hi! Have you tried out the new partitioning features of MySQL 5.1 to do
this? Would be cool if you had some performance numbers comparing the
older MERGE table method with the newer partitioning...
> ----- Original Message ----- From: "He, Ming Xin PSE NKG"
> To: <mysql@stripped>
> Sent: Tuesday, April 17, 2007 11:03 PM
> Subject: Problem on millions of records in one table?
> The number of the records in one table increase constantly. As
> evaluated, the amount would increase to at least 30 millions within one
> year. So we worry about whether mysql could handle such a big amount of
> records with good performance. Or need we some other solutions to avoid
> this problem ,such as using Partition, dividing a big table and etc. Any
> help or idea would be greatly appreciated.
> Best Regards