Hi all,
I'm in the process of evaluating MySQL/InnoDB for a big project.
The data is composed from few tens millions of records with foreign keys
used between tables.
In the main table the search will be based on a string key (varchar(32)).
The scenarios in the systems are:
1. Bulk update (initialization, once an hour).
2. ~10 updates per second.
3. ~10000 queries per second.
We'll also need replication (master-slave).
From your experience can we achieve this on Netra computers running SunOS
5.8? The other natural selection is Oracle.
How much memory will we need (cache buffers ?)
Should we prefetch and search in the memory (in memory db?) or can we really
work in front of the MySQL server? How many queries can we expect to be
served simultaneously? What is the expected time for a single query (from
the main table?)
Thanks for your inputs,
Zbenmo.
_________________________________________________________________
Send and receive Hotmail on your mobile device: http://mobile.msn.com
| Thread |
|---|
| • newbie question (can I build on MySQL/InnoDB for my project) | Oren Zeev-Ben-Mordehai | 7 May |