Oracle RAC

Overview of Real Application Clustering

Features and Functionality
Overview
 What is RAC?
 Cache Fusion
 Failover and Load-balancing
 Transparent Application Failover (TAF)
 Other RAC Features
RAC – What is it?
 Multiple instances of Oracle running on up to 8 nodes
 Multiple instances share a single physical database
 All instances can simultaneously execute transactions
against the single database
 Caches are synchronized using Oracle’s Global
Cache Management technology (Cache Fusion)
History of Oracle RAC

 Previous Oracle Clustering Products

 Oracle Failsafe on Windows
 OPS (Oracle Parallel Server) on multiple
platforms
 OPS to RAC: 7.3 OPS  8i OPS  9i RAC  10g RAC
 The clustering mechanism used to be more
dependent on the Operating System.
 With 10g RAC, clustered database is built
into Oracle
Oracle RAC Features
 Full Cache Fusion
 Enhanced coordination of cache management and
distributed lock manager (DLM)
 Lock simplification and automation
 Global Cache Service coordinates local buffer cache
and remote block transfers
 Enhanced IPC
 Resource configuration simplification and
automation
 Improved cluster aware tools
 Enhanced DBCA
 Oracle Enterprise Manager and Grid Control
Integration Enhancements
RAC uses “Shared Everything”

Users

Server Server Server Server

Database
How RAC clustering is done
 One set of data
 All nodes in the cluster see the same set of data
 All nodes have access
 Any node can update the data
Increased Manageability
 One virtual system to configure and manage
 Single Oracle Database
 Single management console
 Single system image for the database integrated with the
cluster
 Cluster-wide monitoring and diagnostics
 Oracle Enterprise Manager Integration (9i)
 Oracle DBConsole and Grid Control Integration
(10g)
What’s shared; What’s not
 Shared
 Disk access
 Resources that manage data
 All instances have common data & controls files
 Not Shared
 Each node has its own dedicated:
 System memory
 Operating system
 Database instance
 Application software
 Each instance has individual
 Log files and
 Rollback segments
RAC can perform
 Load-balancing
 Failover
Load-Balancing through the Listener

Client

r Lis
ne te
r

e ne
ne

Lis
Lis r
te

t
s

en
Li

er
Node 1 Node 2 Node 3 Node 4

Database
How workload is balanced
 Nodes report CPU
usage to listeners
Node 1

Database
Client

Node 2
How workload is balanced
 Listeners choose least
busy node when
request comes in from
client Node 1

Client Database

Node 2
Load-Balancing

Users

Node 1 Node 2 Node 3 Node 4

Database
Failover

 If a node in the shared disk cluster fails, the system

dynamically redistributes the workload among the surviving
cluster nodes.
 RAC checks to detect node and network failures. A disk-
based heartbeat mechanism uses the control file to monitor
node membership and the cluster interconnect is regularly
checked to determine correct operation.
 Reduced time to recovery with concurrent resource
configuration and instance (cache) recovery
 Enhanced failover reliability in 10g with the use of Virtual IP
addresses (VIPs)
Failover

Users

X
Server Server Server Server

Database
Transparent Application Failover
 Masks failures to end users; they don’t need to log
back into the system
 Applications and users are transparently reconnected
to another node
 Applications and queries continue uninterrupted
 Transactions can failover and replay
 Login context maintained
 DML transactions are rolled back
RAC Improvements for Oracle 9i
 Full Cache Fusion
 Enhanced coordination of cache management and
distributed lock manager (DLM)
 Lock simplification and automation
 Global Cache Service coordinates local buffer cache
and remote block transfers
 Enhanced IPC (InterProcess Communication)
 Resource configuration simplification and automation
Oracle 10g RAC New Features
 Integrated Clusterware Management
 No third-party clusterware software required
 Automatic Workload Management
 Application workloads can be managed through named services
 Single System Image Management
 Enterprise Manager manages RAC instances as a single image
 Fast Connection Failover
 Fast recovery between the database and mid-tier applications
 Performance Improvements
 Reduced message traffic, memory usage, and other resources
 Zero Downtime Patching
 Patches may be applied one node at a time without downtime
 Cluster Verification and Improved Diagnostic Tools
 New cluster diagnostic tool and improved diagnostic tools
Full Cache Fusion
 Is a major feature of RAC starting with 9i
 The underlying technology that enables RAC
 Protocol that allows instances to combine their data
caches into a shared global cache
 Allows any node to get the most up-to-date data
information from the cache of any other node in the
cluster without having to access the disk drives
again.
 Improved performance with 10g
What is Cache Fusion? When do I
care about it?
 “Dirty” block of data is created
 Data from disk is read into memory on a node
 Data is updated on that node
 Data hasn’t been written to disk yet.
 Another node requests the data
“ABC” block of data written to the
disk drives in the database

Node A Node B

ABC Data
“ABC” block of data read into
memory on Node A

Node A Node B

ABC Data

ABC Data
“ABC” updated to “XYZ” in cache

Node A Node B

ABC Data
XYZ Data

ABC Data
Node B requests data block

I want
data!
Gimme!
Node A Node B Gimme!
ABC Data
XYZ Data

ABC Data
Node A must write data block to
disk drive
I want
data!
Gimme!
Node A Node B Gimme!
ABC Data
XYZ Data
A C
9 iR
t o
s
wr

o u
ite

v i
r e
ABC Data
XYZ Data P
Node B must read data block from
disk drive

Node A Node B

ABC Data
XYZ Data XYZ Data
A C
9 iR
t o
us
ad

io
re

e v
ABC Data
XYZ Data Pr
Now with RAC Cache Fusion

I want
data!
Gimme!
Node A Node B Gimme!
ABC Data
XYZ Data XYZ Data

 Data is transferred immediately

via the interconnect
 Shared cache minimizes
slow I/O

ABC Data
Shared Cache Across Nodes

Users

Server Server Server Server

Database
Resource Simplification and
Automation
 No init.ora parameters required
 Resource affinity
 to move the location of the resource masters for a
database file to the instance where block operations are
most frequently occurring. This optimizes
 Dynamic resource remastering
 Ability to move the ownership of a resource between
instances of Real Application Clusters.
 Dynamic resource remastering is used to implement
resource affinity for increased performance.
Review
 What does cache fusion avoid that was mandatory
in previous versions of Oracle Parallel Server?
 Which Oracle process is most important in
managing user session failover?
 If the purpose of the interconnect is NOT to serve
as a “heartbeat”, where is the heartbeat?
Summary
 New Features
 Shared Everything Clustering
 Cache Fusion
 RAC Clustering failover & load-balancing