09-27-2010 12:59 PM
when delpoying oracle (11g) dataguard on VCS (red hat linux 5), I have read you can only have 1 standby per agent. Does anyone know if you can have multiple agents (multiple standbys) from the 1 primary cluster ?
Thanks all
Solved! Go to Solution.
11-02-2010 02:29 PM
From page 12 of “Veritas Cluster Server Agent for Oracle Data Guard Installation and Configuration Guide”
Typical Oracle data Guard setup in a VCS cluster.
Clustering in a Data Guard environment typically consists of the following
hardware infrastructure:
■ The primary database instance (db1) sends redo data across a TCP/IP link to
a standby database instance (db2). A local cluster protects the primary database
and makes it highly available.
■ The standby database instance applies the redo information to a physical copy
of the primary database.
■ The primary and standby sites must be connected through a single TCP/IP
network connection. This link can be shared with VCS global clusters for
heartbeat communication.
■ Network heartbeating between the two data centers to determine their health;
this network heartbeating could be LLT or TCP/IP.
Based on the above definition, with VCS one primary database can have only one dataguard resource setup.
If there are multiple primary databases in the same VCS, you can have multiple associated dataguard resources if your hardware can handle the work load.
09-27-2010 01:59 PM
I am not sure, if you are discussing about the VCS bundled agents/ enterpise agents... request you to explain a bit more about the query.
Regards
Rajesh
09-29-2010 07:18 AM
yes , I am referring to the VCS bundled agents for managing dataguard during a failover to another node in the VCS cluster. According to the documentation VCS can manage failover for 1 dataguard standby, my query is , can VCS manage multiple standbys ?
11-02-2010 02:29 PM
From page 12 of “Veritas Cluster Server Agent for Oracle Data Guard Installation and Configuration Guide”
Typical Oracle data Guard setup in a VCS cluster.
Clustering in a Data Guard environment typically consists of the following
hardware infrastructure:
■ The primary database instance (db1) sends redo data across a TCP/IP link to
a standby database instance (db2). A local cluster protects the primary database
and makes it highly available.
■ The standby database instance applies the redo information to a physical copy
of the primary database.
■ The primary and standby sites must be connected through a single TCP/IP
network connection. This link can be shared with VCS global clusters for
heartbeat communication.
■ Network heartbeating between the two data centers to determine their health;
this network heartbeating could be LLT or TCP/IP.
Based on the above definition, with VCS one primary database can have only one dataguard resource setup.
If there are multiple primary databases in the same VCS, you can have multiple associated dataguard resources if your hardware can handle the work load.