ankush,
fault tolerance/redundancy is not just an issue of multiple mysql databases. the issue os redundancy has to be approached from a top/system level. you essentially need to have the ability to essentially turn off a machine, and have the backup take over...
most situations require a sort of slave/master configuration with whatever machine is the slave being tied to the master, in terms of data/process replication... if the master is working correctly, the slave simply acts as a backup... if the master goes down, the slave takes over, performing as the new master.. when the master comes back on line.. it gets 'up to speed' and may or may not resume the role of the master, or the slave...
this kind of configuration requires an understanding of the required processes from the apache/mysql/application points of view.
if you've never created this kind of system, i would strongly advise that you talk to someone who's created these systems. if your app is going to be in a critical area of the hospital, then you don't want to screw this up.
Bottom post, people, bottom post please!
Fault tolerance, as Bruce said, is not a trivial thing. We do it all
the time, but you have to throw iron at it, either by HA hardware and
software or by fronting the cluster with a hardware load balancer (Alteon AD3, F5, Radware, etc.).
The real trick is redundant database servers. You typically have to set up a master/slave or active/passive failover system. The problem is in keeping the database on the inactive server synchronized with the active one. This can be done with transaction logs, physical synching and several other methods. There are benefits and drawbacks to each one and they're not foolproof. Microsoft's database clustering stuff is a freaking nightmare. Oracle's is much better. PostgreSQL has some HA stuff, but I've never used it. AFAIK, MySQL has no such mechanism (I could be wrong on that).
You have to REALLY analyze what you're trying to accomplish before you go headlong into this. It ain't easy and it ain't for the faint-of- heart.
-----Original Message----- From: fedora-list-bounces@xxxxxxxxxx [mailto:fedora-list-bounces@xxxxxxxxxx]On Behalf Of Ankush Grover Sent: Thursday, June 16, 2005 2:15 AM To: For users of Fedora Core releases Subject: fc3 & mysql clustering
Hey friends,
We recently got an order from a Hospital to install our healthcare product in their premises.They want fault tolerance means if one server goes down the other server takes over it.
The packages which will be running on FC3 will be
a) FC3 itself( 32bit intel) b) mysql as database(4.1) c) apache as web server d) our software.
Can anyone tell me any good links or tutorials for such scenario.It there is some other OS as compared to FC3 which also supports clustering but which is more less works like FC3,as i don't want to spend too much time in learning that OS,I would like to spend the time on learning and configuring clustering.
Thanks & Regards
Ankush Grover
-- fedora-list mailing list fedora-list@xxxxxxxxxx To unsubscribe: http://www.redhat.com/mailman/listinfo/fedora-list
-- ---------------------------------------------------------------------- - Rick Stevens, Senior Systems Engineer rstevens@xxxxxxxxxxxxxxx - - VitalStream, Inc. http://www.vitalstream.com - - - - A day for firm decisions!!! Well, then again, maybe not! - ----------------------------------------------------------------------
