• Where should we store the Oracle binaries in Oracle Real Application Cluster environment?
    Post By: admin

White paper "Oracle Homes in an Oracle Real Application Clusters Environment" released by Oracle address this query in details. You can access the white paper here.

  • Where should we store the Oracle binaries in Oracle Real Application Cluster environment?
    Post By: admin

Oracle CRS aka Clusterware does not provide any pre configured cluster resources and wrappers that can be used to provide the cold fail over of single instance databases BUT the Oracle Clusterware framework itself allows users to add their own cluster resources to achieve this behavior.

  • How can I use Oracle CRS to provide cold fail over of my single instance databases?
    Post By: admin

Oracle CRS aka Clusterware does not provide any pre configured cluster resources and wrappers that can be used to provide the cold fail over of single instance databases BUT the Oracle Clusterware framework itself allows users to add their own cluster resources to achieve this behavior.

Please read this article for more information.

  • I have configured the user equivalence. Why am I getting "stdin: is not a tty" message when I execute any operating system command on the remote machine using the SSH?
    Post By: admin

Ensure that none of the login script such as bashrc, profile etc are executing mesg operating system command during the login process of operating system account being used for installing Oracle software.

Remove or comment the 'mesg' command from the login scripts.

 

  • What are the files stored under scls_scr directory on any Oracle Cluster system?
    Post By: admin

These are cluster control files stored under scls_scr/HOSTNAME/root directory on any Oracle Cluster node. Oracle uses these files to store the run time data for Oracle Cluster such as boot time information in the cssrun and cssfboot files.

Similarly Oracle uses other control files to store different information. Oracle stores the following files under scls_scr/HOSTNAME/root directory:

clsomonpid
crsstart
cssrun
noclsm on
oprocdpid
crsdboot
cssfboot
daemonpid
nocl svmon

 

  • What is Oracle RAC?
    Post By: admin

Oracle Real Application Clusters (RAC) is a database option to create cluster database with a shared cache architecture that overcomes the limitations of traditional shared-nothing and shared-disk approaches to provide highly scalable and available database solutions for all your business applications.

  • I recently upgraded from previous Oracle Clusterware version to Oracle Grid Infrastructure 11.2.0.2 and I can see that Oracle has configured a virtual private IP on top of the private interface. What is this extra IP? Is this a bug?
    Post By: admin

Nope, definitely this is not a bug as this is expected behaviour from Oracle Grid Infrastructure 11.2.0.2 onwards. Oracle has introduced feature to provide redundancy for the private interconnect and this is implementation of this feature. This extra IP is called Highly Available IP. Please refer to this post here.

  • Why there is a virtual IP running on private interface in Oracle Grid Infrastructure 11.2.0.2 where as there is no such virtual IP in Oracle Grid Infrastructure 11.2.0.1?
    Post By: admin

Starting with Oracle Grid Infrastructure 11.2.0.2, Oracle provide redundancy for private interconnect by using High Available IP cluster resources, which manage these extra private virtual IPs. Oracle can configure four private virtual IPs on a cluster server. This post may give you more information on this.

  • In general, Oracle reboot the failed server in cluster to fence node but I have observed weird behavior in Oracle RAC 11.2.0.2 that Oracle not always reboot the failed server because sometime it simply restart the Oracle Clusterware on the failed node
    Post By: admin

Oracle has improved node fencing in Oracle Grid Infrastructure 11g Release 2 (11.2.0.2) and onwards and try to kill the processes on the failed node that are capable of doing I/O and restart the Oracle Clusterware rather than rebooting the failed node.

Oracle will reboot the failed node only if it is not able to kill the processes and restart the Oracle Clusterware on the failed node.