beautypg.com

Stopping a package, Failing over a package, Stopping a package 89 failing over a package 89 – HP StorageWorks 8000 NAS User Manual

Page 91

background image

Managing Your Storage 89

M

ana

ging

Yo
ur St

ora

ge

Stopping a Package

Note

This section applies only if you have purchased a high-

availability NAS solution.

The file system services, NFS, and CIFS are temporarily stopped; the volumes

within the package are unmounted; and NFS and CIFS are restarted. Once

the package has been stopped, it can be restarted on any active server in the

cluster. After a package is stopped, the volumes within the package are no

longer available to client systems until the package is restarted. Stopping a

package does not affect the server's cluster status and does not cause the

cluster to go down.

Caution

Check the Client Activity page before stopping the package to

make sure no one is accessing it. Stopping a package disrupts

service to users who are accessing the package(s) through the

virtual IP address. After you restart the package, the virtual IP

address is valid again.

To stop a package:

1

In the Command View NAS web interface, click the Storage tab.

2

Navigate down the tree and select Failover Packages.

3

Select the package you want to stop by clicking the row.

4

Select Actions > Stop Package.

Failing Over a Package

Note

This section applies only if you have purchased a high-

availability NAS solution.

The specified package fails over to another server in the cluster. The failover

consists of stopping the package on the primary server (node), then starting

the package on the secondary server. This action does not affect the

designated ownership of the package. In other words, the primary node

specified in the package configuration doesn't change. If the cluster were

stopped and restarted, the package would automatically migrate back to its

primary node. This action provides a mechanism for manually failing over

packages without requiring a failure condition. One use of this functionality is

to allow manual load-balancing without requiring the cluster or individual

nodes to be reconfigured and/or rebooted.