Min requirements for clustering

1 Two or more systems

2 Dedicated Private Network

3 Shared storage box which can be accessed by all the nodes that can be disk array or SAN

4 Maintain trusted relationship between all the nodes

The maximum no of nodes in veritas cluster environment should not be exceed more than 32

CLUSTER

Is a collection of nodes that provide scalability and high availability for specified applications in 24*7

CLUSTER COMPONENTS

1 Service

2 Service Group

Service

Which provides a utility for end user

Service Group

Collection of independent utilities inside the frame work for example

1. Failover and switch over

2. Parallel

3. Hybrid

Failover

In a failover service group if once system goes down the other system in cluster nodes automatically takes the load

Switch over

In switch over service group if one system goes down you should manually switch the load to the other system this is usually preferred for maintenance purpose

Note Either in failover or switch over only one machine will be active to provide the resources

Parallel

This service group runs simultaneously one or more systems in the cluster environment. It is

basically used for load balancing

Hybrid

RESOURCE

A resource is a software and hardware utility which provides the services

Ex: NFS service depends on

IP-address---nic card—daemons—mount points

TYPES OF RESOURCES

on-off resource

It can be online position or it can be offline position VCS can start or stop

Ex: IP-address

on-only resource

VCS starts on-only resources but does not stop

Ex: daemon for VCS is HAD (high availability daemon) and HASHADOW

Persistent resource

This resource cannot be brought online or taken offline

Ex: NIC card

AGENTS

It is used to manage the resources

Ex;

Online brings a resource online from offline

Offline brings a resource offline from online

Monitor tests the status of resource whether it is online or offline

Clean cleans up after a resource fails to come online and fails to go offline

Info retrieves specific information for online resource

TYPES OF AGENTS

Bundled Agent

It will control disk, mount point, IP and various other resources types

Enterprise Agent

Enterprise Agent controls third party applications and are licensed separately this includes agents for

example ORACLE, NETBACKUP and SYBASE

Custom Agent

It will control the applications that is not covered by bundled and enterprise agents

PROTOCOLS

LLT (Low Latency Transport Protocol)

Used to transport the load and to check the heartbeat links

It is used to check the priority of nodes in cluster environment and transfer the load to other nodes

GAB and HASHADOW (Group membership or automatic broadcast)

It is a protocol which checks membership of nodes in the cluster environment and gives information

to LLT protocol

It is used for communication between all cluster nodes

Types of servers

master server

media server

clients

MASRER SERVER

DAEMONS

bprd and bpcd

bprd ( back up restore daemon )

bpcd (back up communication daemon)

Storage units (tape,disk, autoloader, NAS)

volume :- A volume contains one or more tapes

catalog :- It maintains databases of clients ,volumes and policies

policies :- The policies are when backup should run for autamatic backups

properties :-

Attributes à contains policies & storage units

schedule à contains backup time ,full backup, incremental

back up selection à which file system of clients

clients à contains client information .

The daemon that controls autoloader is iltid and ‘ vmd ‘