Hi,
I have a 2 node 10gR2 (10.2.0.2) Installation running on HP-UX (PA-RISC) 64 Bit, 11.23. HP Serviceguard Extensions for RAC (11.17) is also installed.
.
Due to maintenance both nodes were brought down. One they were brought up, I noticed that the vip for one node - node2, was running on interface lan0:2. lan0 is our public interface. Now typically, the vip runs on lan0:1. When a node is evicted from the cluster, the vip of the evicted node gets relocated from lan0:1 on the evicted node to lan0:2 on the surviving node.
My question basically is, does it matter which logical interface (:1,:2,:3,:4,etc) the vip runs on for any node - be it the evicted node, or the surviving (master) node. Essentially the mapping for the vip is defined in the OCR and the /etc/hosts file. There is no mention of whether this should run on :1,:2,:3,:4,etc - only that it should be on the public interface - lan0 - which it is.
A look into the ocr file, and this is all i can find:
.
[SYSTEM.css.interfaces.global.lan0.172|d25|d3|d0.1]
.
Not sure what d0.1 means?, the first three segments are those for the public network. .1 = :1 = lan0:1 ??
That said, whenever I have seen a relocation of the vip take place it goes to lan0:2 - However as mentioned earlier, lan0:2 is already in use by one of the nodes, if a relocation to this node were to take place would the the vip come up here as lan0:3 - unfortunately I can't test this out yet, but was wondering if anyone had an idea on this?
Thanks
Naqi
The all-new Yahoo! Mail goes wherever you go - free your email address from your Internet provider.
http://uk.docs.yahoo.com/nowyoucan.html
--
http://www.freelists.org/webpage/oracle-l
Received on Tue Mar 27 2007 - 16:13:25 CDT
