Re: Volume Manager information not persistent across reboots
- From: "gns" <gnskumar@xxxxxxxxx>
- Date: Tue, 28 Nov 2006 19:16:21 +0530
I had initiated this thread long back. I had managed to resolve this.
I came across same issues being raised by others in different forums and
hence thought of posting a response to this thread.
"gns" <gnskumar@xxxxxxxxx> wrote in message
news:ec1d93$7tp$1@xxxxxxxxxxxxxxxxxxxxxxxx
Dear Group members,longer
In a cluster system with solaris volume manager setup, on reboot, the
multi-owner information is getting lost. Also one of the nodes is no
identified as a member of the diskset.
(/var/run/nodelist was manually created and is not persistent across
reboots).
The missing link is that atleast one resource group needs to be created on
the cluster for the nodelist to get updated by the cluster software (bug?).
Oracle RAC setup procedure is given in two documents:
[1] Sun Cluster Data Service for Oracle Real Application Clusters Guide for
Solaris OS (available at docs.sun.com)
[2] Oracle Clusterware and Oracle RAC Installation Guide - 10g Release 2
(10.2) for Solaris OS (Available online in oracle documentation library)
It so happens that neither of them is complete in itself.
- Refer to [2] for the resource group creation.
- Refer to [1] for rest of the steps (note the IP requirements for
Clusterware).
With this one should be able to setup Oracle RAC.
regards
Shiv
.
- Prev by Date: Re: zones and 127.0.0.1
- Next by Date: Posix command locations
- Previous by thread: Configuring a StorEdge 3510 FC RAID array as a JBOD
- Next by thread: Posix command locations
- Index(es):
Relevant Pages
|
