How to drop and recreate Oracle ASM Disk Groups in RAC

This post covers the simple commands on ow to drop and recreate Oracle ASM Disk Groups in RAC.
Warning: Running these commands shall result in Data Loss.
SQL> create pfile from spfile;
$ asmtool -delete \\.\ASMDISK01
$ asmtool -delete \\.\ASMDISK02
$ asmtool -delete \\.\ASMDISK03
$ asmcmd
ASMCMD> lsdg
$ srvctl status diskgroup -g DATA01
$ srvctl status diskgroup -g RECO01
$ srvctl stop diskgroup -g DATA01 -f
$ srvctl stop diskgroup -g RECO01 -f
SQL> drop diskgroup RECO01 force including contents;
Diskgroup dropped.
SQL> drop diskgroup DATA01 force including contents;
Diskgroup dropped.
Now that the ASM Disk Groups and constituent Disks have been dropped/erased, the new ASM Disk Groups can be recreated.
CREATE DISKGROUP DATA01 EXTERNAL REDUNDANCY
  DISK '/dev/asm/asmdisk*';
Resources:
How To Drop and Recreate ASM Diskgroup (Doc ID 563048.1)

How to modify an Oracle RAC Service to startup in a specific DB Mode (For Data Guard)?

Below, I have mentioned commands to modify an Oracle RAC Cluster Managed DB Service to startup in a specific DB Mode (for Data Guard).




Modify the Service to startup in a Primary Mode:
srvctl modify service -d RACDB1 -s RACDB1SERV1 -l primary


Modify the Service to startup in Physical Standby Mode:
srvctl modify service -d RACDB1 -s RACDB1SERV1 -l physical_standby
Modify the Service to o startup in a Snapshot Standby Mode:
srvctl modify service -d RACDB1 -s RACDB1SERV1 -l snapshot_standby


How to Check the configuration/status of the Cluster Managed DB Service:srvctl config service -d RACDB1 -s RACDB1SERV1




Note: Please remember to change the name of the DB Names/Services with your own.

RAC Installs: root.sh fails after installation on the first node

This post covers RAC Install failures: root.sh fails after installation on the first node


Excerpt from Problem Logs:
Disk Group GRID01 creation failed with the following message:
ORA-15018: diskgroup cannot be created
ORA-15017: diskgroup "GRID01" cannot be mounted
ORA-15003: diskgroup "GRID01" already mounted in another lock name space
Configuration of ASM ... failed

see asmca logs at /u01/app/grid/cfgtoollogs/asmca for details
Did not succssfully configure and start ASM at /u01/app/11.2.0/grid/crs/install/crsconfig_lib.pm line 6912.
/u01/app/11.2.0/grid/perl/bin/perl -I/u01/app/11.2.0/grid/perl/lib -I/u01/app/11.2.0/grid/crs/install /u01/app/11.2.0/grid/crs/install/rootcrs.pl execution failed


Cause: "This particular issue arises when root.sh is run concurrently on the first and other nodes on the RAC cluster. The correct approach is to run root.sh on the first node followed by the other nodes of the RAC cluster.\

Solution:
root.sh Fails on the First Node for 11gR2 Grid Infrastructure Installation (Doc ID 1191783.1)
Run the following as the root OS user on Node 1.
"$GRID_HOME/crs/install/rootcrs.pl -deconfig -force -verbose"
Followed by this command on the other nodes:
"$GRID_HOME/crs/install/rootcrs.pl -verbose -deconfig -force -lastnode"


How to Proceed from Failed 11gR2 Grid Infrastructure (CRS) Installation (Doc ID 942166.1)
Here is the master note for troubleshooting/fixing Grid Infrastructure Startup issues.
Troubleshoot Grid Infrastructure Startup Issues (Doc ID 1050908.1)

Valid Node Checking for Registration (VNCR) Implementation on Oracle

Valid Node Checking and Registration (VNCR) is a new feature introduced in versions 11.2.0.4 and 12c of Oracle. VNCR allows registrations to the Oracle Listener more secure such that, they are only allowed from known servers/nodes; hence the namesake. The main advantage of VNCR is to get by without implementing Class of Secure Transport (COST) configurations that, tend to be much more complicated and expensive. VNCR is relatively simple to implement - Following MOS Notes contain the directions and other useful resources.

• Valid Node Checking For Registration (VNCR) (Doc ID 1600630.1)
• How to Enable VNCR on RAC Database to Register only Local Instances (Doc ID 1914282.1)

The following example shows how to implement VNCR on RAC to register the local RAC instances:
VALID_NODE_CHECKING_REGISTRATION_LISTENER=1
VALID_NODE_CHECKING_REGISTRATION_LISTENER_SCAN1=1
REGISTRATION_INVITED_NODES_LISTENER_SCAN1=(<enter the list of public ip's of all nodes separated by commas>)
VALID_NODE_CHECKING_REGISTRATION_LISTENER_SCAN2=1
REGISTRATION_INVITED_NODES_LISTENER_SCAN2=(<enter the list of public ip's of all nodes separated by commas>)
VALID_NODE_CHECKING_REGISTRATION_LISTENER_SCAN3=1
REGISTRATION_INVITED_NODES_LISTENER_SCAN3=(<enter the list of public ip's of all nodes separated by commas>)

Oracle Real Application Clusters (RAC): Jumbo Frames Notes/Recommendations

In this post, i shall quickly summarize recommendations for implementing Jumbo Frames on the Oracle Real Application Clusters (RAC) Private Cluster Interconnect.

Notes/Recommendations:

• It is recommended to implement Jumbo Frames for the RAC Cluster Interconnect for better overall performance.
• Interconnect Latency is significantly reduced by implementing Jumbo Frames.
• Implementing Jumbo Frames reduces the UDP, TCP, Ethernet overhead.
• The ifconfig -mtu 9000 shall implement Jumbo Frames with an MTU of 9000 (1500 for smaller frames).
• Test the RAC cluster thoroughly after configuring Jumbo Frames.

References: 

• Recommendation for the Real Application Cluster Interconnect and Jumbo Frames (Doc ID 341788.1)
• How to Validate Network and Name Resolution Setup for the Clusterware and RAC (Doc ID 1054902.1)
• Problem - Dropped segments and bad checksums caused by Jumbo frames. (Doc ID 1073610.1)
• System experiences memory shortage when using nxge with jumbo frames and NFS (Doc ID 1353254.1)