as circumstances require. If the cluster node owning the group of file shares should be s hut down or
fail, the other node in the cluster will begin sharing the directories until the original owner node is
brought back on line. At tha t time, ownership of the group and its resources can be brought back
to the o riginal own er node.
Resource planning
1. Create a cluster group for each node in the cluster with an IP address resource and a n etwork
name resourc
e.
Cluster resource groups are used to balance the processing load on the servers. Distribute ownership
of the groups bet ween the virtual ser vers.
2. For NFS environments, configure the NFS server.
NFS specific procedures include entering audit and file lock information as well as setting up client
groups and user name mappings. These procedures are not unique to a clustered deployment and
are detailed in the Microsoft Services for NFS section within the “O ther network fi le and p rint services”
chapter. C
hanges to NFS setup information are automatically replicated to all nodes in a cluster.
3. Create the
file share resources.
4. Assign ownership of the file share resources to the resource groups.
a. Divide ownership of the file share resource between the resource groups, which are in turn
distributed bet ween the virtual servers, for effective load balancing.
b. Verify that the physical disk resource for this file share is also included in this group.
c. Verify t
hat the resources are dependent on the virtual ser vers and physical disk resources
from which the file share was created.
Permissions and access rights on share resources
File Share and NFS Share permissions must be managed using the Cluster Administrator tool versus the
individual shares on the file system themselves via Windows Explorer. Administering them through
the Cluster Administrator tool allows the permissions to migrate from one node to other. In addition,
permissions established using Explorer are lost after the share is failed or taken offline.
NFS cluster-specificissues
For convenience, all suggestions are listed below:
• Back
up user and group mappings.
To av
oidlossofcomplexadvancedmappingsinthecaseofasystemfailure,backupthemappings
whenever the mappings have been edited or new mappings have been added.
• Map consistently.
Grou
ps that are mapped to each other should contain the same users and the members of the groups
sho
uld be properly mapped to each other to ensure proper file access.
• Map properly.
• Valid UN IX users should be m apped to valid Windows users.
•Val
id UNIX groups should be mapped to valid Windows groups.
• Mapped Windows user must have the “Access this computer from the Network privilege”
or the mapping will be squashed.
•Th
e m apped Windows user must have an active password, o r the mapping will be squashed.
• In a clustered de ployment, create user name mappings using domain user accounts.
Because the security identifiers of local a ccounts are recognized only by the local server, other nodes
in the cluster will not be able to resolve those accounts during a failover. Do not create m appings
using local user and group accounts.
110
Cluster administration
Comments to this Manuals