| WARNING: No cluster interconnect has been specified [message #269727] |
Mon, 24 September 2007 05:43 |
ara.oracle
Messages: 142
Registered: August 2007
|
Senior Member |
|
|
RDBMS Version: Oracle 10g 10.2.0.1.0
Operating System and Version: Microsoft windows 2003
We had RAC setup consists of two nodes.
In alert log we had an error
Interface type 1 DB Switch 172.1.1.0 configured from OCR for use as a cluster interconnect
WARNING 172.1.1.0 could not be translated to a network address error 1
Interface type 1 DB Switch 172.1.1.0 configured from OCR for use as a cluster interconnect
WARNING 172.1.1.0 could not be translated to a network address error 1
Interface type 1 MZ-switch 192.168.1.0 configured from OCR for use as a public interface
Interface type 1 MZ-Switch 192.168.1.0 configured from OCR for use as a public interface
WARNING: No cluster interconnect has been specified. Depending on the communication driver configured Oracle cluster traffic
may be directed to the public interface of this machine.
Oracle recommends that RAC clustered databases be configured
with a private interconnect for enhanced security and performance.
I checked this error in meatlink.com and they given solution for this is
Cause:
The Clusterware does not see to be able to cope with card names that have dots included
If you run the command
oifcfg iflist :- the following is returned
Public_Vlan 1_165.17.2.xxx 165.17.2.0
Private_VLAN254_165.17.254.xxx 165.17.254.0
From the above you can see the Nic name 's are Public_vlan1_165.17.2 and Private_vlan254_165.17.254
i.e contain dots
If you rename the cards and remove the dots and update the names in the OCR , the clusteware selects the correct interfaces and works perfectly
Solution
1. Rename the network cards through the control panel network and remove any dots from the name
2. run the following to update the ocr :-
a) oifcfg delif
c) oifcfg getif
d) oifcfg setif -global "name 1 without dots"/ip address:public
e) oifcfg setif -global "name 2 without dots"/ipaddress:cluster_interconect
stop and retart the clusterware and database and the problem should no longer be reported in the alert logs
I just want to know what is the procedure for doing this,i.e
Wheather the db instance must be up or down
Cluster service must be down or up
Just give me the procedure and answer briefly
|
|
|
|
] 
Blog
Search
Help
Members
Register
Login
Home
