How do you cope with huge MySQL databases?

With huge I mean millions and millions of rows. Already red here and there some recommendations. Increasing key cache size and such stuff is not helping since the key exceeds the memory size. Using no PI at all: quick inserts, slow selects. Vice versa for e.g. a autoincrement PI.

What I have: only simple selects and inserts. No joins, no other stuff. Only one primary index as autoincrement int.

Solutions? Oracle? PostgreSQL? SSD-Disk?

A simple solution is to split the table on the costs of higher maintenance? Anything new from mysql to do that "automatically"?

MySQL has no problem handling millions or rows. I currently have ~100 million rows for one table.

• Is your application read or write heavy?
• If your application is read heavy, are you caching results?
• Can you prune old records from the database?
• Are your queries optimized? Sometimes really stupid things like OFFSET can crush a db.
• Is your table properly indexed.
• The list goes on an on...

If you need any help you may contact me privately.

I have the same problem slow searching, the only thing I could do is to increase the file execution time, until I can get an i7

caching files to lower the amount of database calls

Paste your my.cnf here