Hello,
We are planning to reprogramm our search machine and website. Since we
have call agents who query the same database as the customers do, we
had the idea of splitting the database to 2 servers.
The call agents can update customer data with our intern webinterface,
which not only concerns tables that are for the call agents, but also
concerns tables of the search machine, holding company data such as
address, search words, etc.
O.k., now my questions. Since the customer himself can administer his
own data, as well as the call agent (or media agent), I was thinking
of doing following:
* I have 2 servers. Server 1 is responsible for the search machine and
just serves customers searching for a company or for products. The
MySQL database is there just for querying.
* Server 2 serves call agents, as well customers who want to change
their data, reading and writing.
* Since the search machine should of course show updated data the call
agent or the customer have eventually changed, the data from server 2
must be synchronized with the MySQL databese of server 1. (e.g. at
midnight)
Questions:
-------------
0. Is the above described the right way?
1. What do I have to consider when planning the database structure?
2. How do I synchronize? Can I copy specific tables from one server to
another using cronjobs (crontables)? Since this is absolute newland to
me, how?
Any expert knowledge or experience with MySQL replication or MySQL
clusters? I think replication is just thought for mirroring, or am I
mistaken?
Thanks a lot.