-
How does clustering in Oracle 10g work?
Hi,
Sorry if this is a layman question here but I am interested to know the answer from the experts here as I didn't get any proper/exact information from the internet.
How does clustering work in Oracle 9i/10g? I read that Oracle 9i RAC uses shared disk architecture where the servers have access to whole database. If a server fails, another server (since it has access to the whole database, it takes over the load of the first server). my question is if these servers have access to database, where is this database stored? aren't these servers part of database? If they are, if a server fails, data stored in that server is not lost?
Thanks in advance,
Padma.
-
shard disk, the database is stored on shared disk which all servers have access to.
each instance then opens the same database
-
Hi Dave,
Thanks for the reply.
So, servers are not the storage units. is it so? they just carry the request, talk to the database and respond back with data. is that so?
Thanks,
Padma.
-
yes, servers dont store anything - disk does
-
so, how does this clustering concept really work with databases? How does clustering really help with respect to databases? I can understand how this helps with application servers like weblogic etc. But when there is only one database, anyway, how does clustering or having multiple server instances help?
when I have an application server and a database, why would I choose clustering at database level when I can do it at application server itself? the job of server instances in the cluster is anyway carrying the request and they don't really hold any data. so, if an instance carrying a request fails, another node in the application server will take the load. I don't really require clustering at Database to do this.
I am interested to know why Oracle has come up with clustering at Database level.
Thanks,
Padma.
-
Originally Posted by techie_invincib
so, how does this clustering concept really work with databases? How does clustering really help with respect to databases? I can understand how this helps with application servers like weblogic etc. But when there is only one database, anyway, how does clustering or having multiple server instances help?
It helps because the data (which is shared on disks, DAS,NAS,SAN,JBOD whatever) resides and lives always. Instances are added , instaances are dropped, instances die but the business continuity is not affected.
Originally Posted by techie_invincib
when I have an application server and a database, why would I choose clustering at database level when I can do it at application server itself? the job of server instances in the cluster is anyway carrying the request and they don't really hold any data. so, if an instance carrying a request fails, another node in the application server will take the load. I don't really require clustering at Database to do this.
Same as I asnwered above, your front-end in this case a app server (which I'd assume according to your argument is not only clustered but also NLB'ed or hardware load balanced) is clustered to keep your front end running, your database is clustered and load balanced to keep your back-end up with minimal downtime.
Hope this gives you a bit of an idea. Wanna learn more, search this site and I'm currently writing a RAC series where you can try a windows, linux or unix setup. Also read oracle's concepts manual and RAC manuals (all freely available at http://www.oracle.com/pls/db102/homepage)
Tarry Singh
I'm a JOLE(JavaOracleLinuxEnthusiast)
--- Everything was meant to be---
Posting Permissions
- You may not post new threads
- You may not post replies
- You may not post attachments
- You may not edit your posts
-
Forum Rules
|
Click Here to Expand Forum to Full Width
|