Configuring a Non-Global Zone on a Global-Cluster Node

This section provides the following procedures to create a non-global zone on a global-cluster node.

• How to Create a Non-Global Zone on a Global-Cluster Node

• How to Configure an HAStoragePlus Resource for a Cluster File System That is Used by Non-Global Zones

How to Create a Non-Global Zone on a Global-Cluster Node

Perform this procedure for each non-global zone that you create in the global cluster.

You can configure an Oracle Solaris Containers non-global zone, simply referred to as a zone, on a cluster node while the node is booted in either cluster mode or in noncluster mode.

• If you create a zone while the node is booted in noncluster mode, the cluster software discovers the zone when the node joins the cluster.

• If you create or remove a zone while the node is in cluster mode, the cluster software dynamically changes its list of zones that can master resource groups.

Before You Begin

Perform the following tasks:

• Plan your non-global zone configuration. Observe the requirements and restrictions in Guidelines for Non-Global Zones in a Global Cluster.

• Have available the following information:

  • The total number of non-global zones that you will create.

  • The public adapter and public IP address that each zone will use.

  • The zone path for each zone. This path must be a local file system, not a cluster file system or a highly available local file system.

  • One or more devices that should appear in each zone.

  • (Optional) The name that you will assign each zone.

• If you will assign the zone a private IP address, ensure that the cluster IP address range can support the additional private IP addresses that you will configure. Use the cluster show-netprops command to display the current private-network configuration.

  If the current IP address range is not sufficient to support the additional private IP addresses that you will configure, follow the procedures in How to Change the Private Network Configuration When Adding Nodes or Private Networks to reconfigure the private IP-address range.

For additional information, see Zone Components in System Administration Guide: Oracle Solaris Containers-Resource Management and Oracle Solaris Zones.

1. Become superuser on the global-cluster node where you are creating the non-voting node.

   You must be working in the global zone.

2. Verify on each node that multiuser services for the Service Management Facility (SMF) are online.

   If services are not yet online for a node, wait until the state changes to online before you proceed to the next step.

   phys-schost# svcs multi-user-server node
   STATE          STIME    FMRI
   online         17:52:55 svc:/milestone/multi-user-server:default

3. Configure, install, and boot the new zone.

   ---

   Note - You must set the autoboot property to true to support resource-group functionality in the non-voting node on the global cluster.

   ---

   Follow procedures in the Oracle Solaris documentation:

   1. Perform procedures in Chapter 18, Planning and Configuring Non-Global Zones (Tasks), in System Administration Guide: Oracle Solaris Containers-Resource Management and Oracle Solaris Zones.
   2. Perform procedures in Installing and Booting Zones in System Administration Guide: Oracle Solaris Containers-Resource Management and Oracle Solaris Zones.
   3. Perform procedures in How to Boot a Zone in System Administration Guide: Oracle Solaris Containers-Resource Management and Oracle Solaris Zones.
4. Verify that the zone is in the ready state.

   phys-schost# zoneadm list -v
   ID  NAME     STATUS       PATH
    0  global   running      /
    1  my-zone  ready        /zone-path

5. (Optional) For a shared-IP zone, assign a private IP address and a private hostname to the zone.

   The following command chooses and assigns an available IP address from the cluster's private IP-address range. The command also assigns the specified private hostname, or host alias, to the zone and maps it to the assigned private IP address.

   phys-schost# clnode set -p zprivatehostname=hostalias node:zone

   -p

   Specifies a property.

   zprivatehostname=hostalias

   Specifies the zone private hostname, or host alias.

   node

   The name of the node.

   zone

   The name of the global-cluster non-voting node.

6. Perform the initial internal zone configuration.

   Follow the procedures in Performing the Initial Internal Zone Configuration in System Administration Guide: Oracle Solaris Containers-Resource Management and Oracle Solaris Zones. Choose either of the following methods:

   • Log in to the zone.

   • Use an /etc/sysidcfg file.

7. In the non-voting node, modify the nsswitch.conf file.

   These changes enable the zone to resolve searches for cluster-specific hostnames and IP addresses.

   1. Log in to the zone.

      phys-schost# zlogin -c zonename

   2. Open the /etc/nsswitch.conf file for editing.

      sczone# vi /etc/nsswitch.conf

   3. Add the cluster switch to the beginning of the lookups for the hosts and netmasks entries, followed by the files switch.

      The modified entries should appear similar to the following:

      …
      hosts:      cluster files nis [NOTFOUND=return]
      …
      netmasks:   cluster files nis [NOTFOUND=return]
      …

   4. For all other entries, ensure that the files switch is the first switch that is listed in the entry.
   5. Exit the zone.
8. If you created an exclusive-IP zone, configure IPMP groups in each /etc/hostname.interface file that is on the zone.

   You must configure an IPMP group for each public-network adapter that is used for data-service traffic in the zone. This information is not inherited from the global zone. See Public Networks for more information about configuring IPMP groups in a cluster.

9. Set up name-to-address mappings for all logical hostname resources that are used by the zone.
   1. Add name-to-address mappings to the /etc/inet/hosts file on the zone.

      This information is not inherited from the global zone.

   2. If you use a name server, add the name-to-address mappings.

Next Steps

To install an application in a non-global zone, use the same procedure as for a stand-alone system. See your application's installation documentation for procedures to install the software in a non-global zone. Also see Adding and Removing Packages and Patches on an Oracle Solaris System With Zones Installed (Task Map) in System Administration Guide: Oracle Solaris Containers-Resource Management and Oracle Solaris Zones.

To install and configure a data service in a non-global zone, see the Oracle Solaris Cluster manual for the individual data service.

How to Configure an HAStoragePlus Resource for a Cluster File System That is Used by Non-Global Zones

Use this procedure to make a cluster file system available for use by a native brand non-global zone that is configured on a cluster node.

1. On one node of the global cluster, become superuser or assume a role that provides solaris.cluster.modify RBAC authorization.
2. Create a resource group with a node list of native brand non-global zones.
   • Use the following command to create a failover resource group:

     phys-schost# clresourcegroup create -n node:zone[,…] resource-group

     -n node:zone

     Specifies the names of the non-global zones in the resource-group node list.

     resource-group

     The name of the resource group that you create.

   • Use the following command to create a scalable resource group:

     phys-schost# clresourcegroup create -S -n node:zone[,…] resource-group

     -S

     Specifies that the resource group is scalable.

3. Register the HAStoragePlus resource type.

   phys-schost# clresourcetype register SUNW.HAStoragePlus

4. On each global-cluster node where a non-global zone in the node list resides, add the cluster file system entry to the /etc/vfstab file.

   Entries in the /etc/vfstab file for a cluster file system must contain the global keyword in the mount options.

   See Sample Entries in /etc/vfstab for Cluster File Systems in Oracle Solaris Cluster Data Services Planning and Administration Guide.

5. Create the HAStoragePlus resource and define the file-system mount points.

   phys-schost# clresource create -g resource-group -t SUNW.HAStoragePlus \
   -p FileSystemMountPoints="mount-point-list" hasp-resource

   -g resource-group

   Specifies the name of the resource group that the new resource is added to.

   -p FileSystemMountPoints="mount-point-list"

   Specifies one or more file-system mount points for the resource.

   hasp-resource

   The name of the HAStoragePlus resource that you create.

   The resource is created in the enabled state.

6. Add a resource to resource-group and set a dependency for the resource on hasp-resource.

   If you have more than one resource to add to the resource group, use a separate command for each resource.

   phys-schost# clresource create -g resource-group -t resource-type \
   -p Network_resources_used=hasp-resource resource

   -t resource-type

   Specifies the resource type that you create the resource for.

   -p Network_resources_used=hasp-resource

   Specifies that the resource has a dependency on the HAStoragePlus resource, hasp-resource.

   resource

   The name of the resource that you create.

7. Bring online and in a managed state the resource group that contains the HAStoragePlus resource.

   phys-schost# clresourcegroup online -eM resource-group

   -M

   Specifies that the resource group is managed.

Example 6-1 Configuring an HAStoragePlus Resource for a Cluster File System That is Used by Non-Global Zones

The following example creates a failover resource group, cfs-rg, to manage an HA-Apache data service. The resource-group node list contains two non-global zones, sczone1 on phys-schost-1 and sczone1 on phys-schost-2. The resource group contains an HAStoragePlus resource, hasp-rs, and a data-service resource, apache-rs. The file-system mount point is /global/local-fs/apache.

phys-schost-1# clresourcegroup create -n phys-schost-1:sczone1,phys-schost-2:sczone1 cfs-rg
phys-schost-1# clresourcetype register SUNW.HAStoragePlus

Add the cluster file system entry to the /etc/vfstab file on phys-schost-1
phys-schost-1# vi /etc/vfstab
#device            device         mount          FS    fsck    mount    mount
#to mount          to fsck        point          type  pass    at boot  options
#
/dev/md/kappa-1/dsk/d0 /dev/md/kappa-1/rdsk/d0 /global/local-fs/apache ufs 5 yes logging,global
Add the cluster file system entry to the /etc/vfstab file on phys-schost-2
phys-schost-2# vi /etc/vfstab

…
phys-schost-1# clresource create -g cfs-rg -t SUNW.HAStoragePlus \
-p FileSystemMountPoints="/global/local-fs/apache" hasp-rs
phys-schost-1# clresource create -g cfs-rg -t SUNW.apache \
-p Network_resources_used=hasp-rs apache-rs
phys-schost-1# clresourcegroup online -eM cfs-rg