It's important to read the "How much RAM" part if you are running any version lower than 5.1. In 5.0 and lower clusters store all
information in memory, which can be a very limiting factor.
My experience with replication is that it is fairly quick, in seconds at most
rather than minutes.
One type of setup that always intrigued me was using the blackhole storage engine to capture the data, then replicating it to store
it. This sounds like an excellent setup if you get high bursts of traffic, but don't need to store it in real time.
http://dev.mysql.com/doc/refman/5.0/en/blackhole-storage-engine.html
Not sure if that would be helpful to you.
----- Original Message -----
From: "Jimmy Guerrero" <[EMAIL PROTECTED]>
To: "'Jesse'" <[EMAIL PROTECTED]>; "'MySQL List'" <mysql@lists.mysql.com>
Sent: Thursday, July 27, 2006 8:47 AM
Subject: RE: Replication vs. Clustering
Hello,
Your description is fairly accurate and we can boil it down even further...
Replication is Asynchronous, Cluster is Synchronous, in regards to how data
is replicated.
Keep in mind that in Cluster, the MySQL Servers really only act as SQL
interfaces for the data in the Cluster, it is the NDB storage engine (Data
Nodes) that deal with ensuring that data is replicated and available.
I regards to the the setup, there is no special hardware, networking or
software requirements for either. Although Replication would require a
minimum of two machines, Cluster likely four.
Tough to say what the "cost" would be depending on which option you go with
and whether you'd need support as well. Needless to say, it would be many
times more economical to go with MySQL then going with something like Oracle
or SQL Server. Also note, MySQL and MySQL Replication are fairly ubiquitous
for websites, so there are a lot of resources on basic and advanced
topologies.
For more info on replication see:
http://dev.mysql.com/doc/refman/5.0/en/replication.html
http://dev.mysql.com/doc/refman/5.0/en/replication-faq.html
For Cluster, start with the FAQ if you haven't already:
http://dev.mysql.com/doc/refman/5.0/en/mysql-cluster-faq.html
Thanks,
Jimmy Guerrero
Sr Product Manager
MySQL, Inc
-----Original Message-----
From: Jesse [mailto:[EMAIL PROTECTED]
Sent: Thursday, July 27, 2006 7:23 AM
To: MySQL List
Subject: Replication vs. Clustering
I think I know the difference, but wanted to make sure. What
is the difference between Replication and Clustering? My
understanding is that in Replication, changes made on the
Master Server are downloaded to a Slave server periodically,
and thus, the slave server is up-to-date within a few minutes
of the Master server.
Am I correct in assuming that in a cluster situation, there
are actually multiple servers, all updated at the same time,
and if any one goes down, then the others can pick up the slack?
Generally (don't need details, but a very general idea), what
is involved in setting up each? What would be the cost of
doing such for a business who has a "mission critical" web
application accessed from all over the U.S.?
Thanks,
Jesse
--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:
http://lists.mysql.com/[EMAIL PROTECTED]
--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
