MetroCluster fail-over?

[Chaiyuth_1]

MetroCluster env are;
DC1 --> has 1xHP-UX and 1xManagement Server
DC2 -->has 1xHP-UX and 1xManagement Server
DC3 --> Quorum Server on Linux
Patch install for SG 11.18 is PHSS_36460 and MC is PHSS_35714
EVA 8100 comply with MC compatibility doc for CV and firmware etc.. And Data sync with 2 EVA site is ok with EVA Read/Write on DC1 and DC2 Read only mode
Using Oracle10g for pkg
After I have created Cluster config and pkg and completed to starting

The questions are below;
1. I have check from MC manual it's has /usr/sbin/DRCheckDiskStatus scripts But I cannot found it after install this install MC software.How I can verify it?
2. I have to start pkg from DC1 is OK.But when manual halt pkg (cmhaltpkg ) and manual start pkg (cmrunpkg ) on DC2 is fails (EVA disk did not fail-over from Rear/Write mode from DC1 to DC2 when I have check from CV GUI on Management Server).Is it true scenario for MC testing??
Or the scenario for testing is starting pkg on DC1 and shutdown HP-UX server then pkg will automatic fail-over to DC2

Thanks,
Chaiyuth

[melvyn burnard]

Have you installed the MetroCluster EVA Toolkit on each server?
Have you created your caeva.env file for the package, and copied it to each servers package directory?
Have you used the tools and created the /etc/dtsconf/caeve.map fiel and copied that to the servers?
What does your package log say about "entering data replication enabler"

[Olivier Masse]

Last time I checked last week, the 8100 was not officially supported yet. And with 11.18 it won't support modular packages but it can be easily modified to work anyway. I suggest you do your first cluster with the help of HP Services, especially with a 0/19 ratio.