Hello,
I've been reading some howto's about MySQL Cluster, but I didn't have
much information about how it really works, or I'm not sure I understand
it right.
I have 4 PCs:
- 1 web server
- 3 mysql server (Dual Xeon + SCSI RAID5)
For now, nothing is running. I'm looking for a solution to what I want
to do.
I would like to have the 3 mysql servers seen as 1 logical server. When
the web server makes a request to mysql, it would transparently touch 1
IP address (the API node), but the work would be dispatched over
different servers (storage nodes).
I want to make sure how storage nodes are seen inside the MySQL Cluster.
I'm reading about "replication" (i'm not a mysql expert). Would this
mean that each mysql server contains the same data (kind of mirroring) ?
I'm not sure that what I'm looking for... My problem is disk space and
long requests. A lot of data is daily inserted in the database. And when
a client (web) makes a request, the process takes a long time because
there's a lot of data parsing and computing.
So what I wanted to do is to be able to have the database "shared" on
the storage nodes. I thought MySQL Cluster could work like a "RAID 5"
over the network. Meaning that if I run out of disk space, I can add an
extra server in the node and data will shared/stripped and written to
that new server making more disk space available. Each storage node
would contain a piece of the database, not everything. And when a
request is done from a client, all four nodes would work together to
make up the result. And if a server fails, it would keep running like
"RAID 5" does.
What I need to know is the following: does MySQL Cluster only makes the
service available by mirroring the data over all storage nodes or can I
use it as I exposed my project ? If not, maybe you can point me to a
solution. Or maybe an alternative setup.
Thanks for your help.
Regards,
--
Alexandre CONRAD - TLV
Research & Development
tel : +33 1 30 80 55 05
fax : +33 1 30 80 55 06
6, rue de la plaine
78860 - SAINT NOM LA BRETECHE
FRANCE
