Novell Home

Configuring GroupWise with NetWare Cluster Services

Novell Cool Solutions: Feature

Digg This - Slashdot This

Posted: 8 Dec 1999
 

Clustering is more than what people do whilst playing Twister or when they're cold. Or whatever. Clustering is a big deal when you're looking for the best configuration for your networking software. We've discovered a little-known chapter from the NetWare 5 documentation all about configuring GroupWise to take advantage of NetWare Cluster Services. It's a bit dry, but very useful. Enjoy.

GroupWise Installation and Configuration
To get the full benefit of using GroupWise with NetWare Cluster Services, GroupWise must be installed and configured to work in a cluster environment.

GroupWise Configuration Requirements
Before configuring NetWare Cluster Services to use GroupWise, GroupWise must be installed and configured properly. See the GroupWise documentation at for installation and setup instructions. In addition, ensure the following GroupWise configuration requirements are met:

  • GroupWise must be configured in Client/Server mode. GroupWise will not failover or failback in a cluster environment without Client/Server mode.

  • NetWare Cluster Services requires that all GroupWise post offices and domains be on shared volumes if GroupWise is to failover or failback. Cluster-enabling shared volumes is not recommended and not necessary for GroupWise to function with NetWare Cluster Services.

  • Since GroupWise agent software must be accessible from every server that will support the GroupWise post office and domain, you must generate startup files and install the GroupWise NLMs in the SYS:SYSTEM directory of each cluster server where GroupWise will run. Install and configure GroupWise on one server in your cluster. Be sure to create the post office, domain, and the software distribution directory on a shared volume and install Message Transfer Agents (MTAs) and Post Office Agents (POAs) on the same shared volume.

    Once GroupWise is installed and configured on one server in the cluster, you need to run the installation program again for each server in the cluster that will run GroupWise.

    Before you install GroupWise on additional servers, you must dismount the shared volume from the server where you last installed GroupWise and mount it on the next server you will install GroupWise on.

    When you install GroupWise on additional cluster servers, be sure to choose not to install MTAs, POAs, and software distribution, since these were installed on the shared volume the first time you installed GroupWise on a server in the cluster. The Install program creates and configures GroupWise startup files on each server.

  • The GroupWise (MTA) and each (POA) require their own IP addresses. These IP addresses float with the resource when a failover or failback occurs and are not bound to a specific server in the cluster.

    If you have trouble linking IP addresses to the MTA or a POA, check the GroupWise link definitions in NetWare Administrator by clicking Tools, GroupWise Utilities, Link Configuration. The links should be configured for TCP/IP, not UNC paths.

  • For each POA to reside on the same server as its volume, each post office should have its own volume. Protected mode must be used in order for each post office to have its own volume. GroupWise 5.5 Enhancement Pack is currently the only GroupWise version capable of running in protected mode.

    Multiple post offices can share the same volume, but all POAs for those post offices must be launched from the same resource, even if they use different protected address spaces.

Review GroupWise Startup Files
In many cases, GroupWise is started by running GRPWISE.NCF on the server where GroupWise is to be loaded. GRPWISE.NCF normally contains two lines.

  • LOAD GWMTA.NLM@<MTA Startup file>
  • LOAD GWPOA.NLM@<POA Startup file>

For example, if the MTA startup file is named GWDOMAIN.MTA and the POA startup file is named GWPOST.POA, the two lines read as follows:

  • LOAD GWMTA.NLM@GWDOMAIN.MTA
  • LOAD GWPOA.NLM@GWPOST.POA

Both the MTA and POA startup file have one required parameter:

  • /HOME<Server Name>/<Domain Directory>

In the MTA startup file, this parameter points the MTA to its domain. In the POA startup file, this parameter points the POA to its post office directory. For both the MTA and POA startup file, the server name must be specific to each server, so each server in the cluster will have slightly different startup files. The STARTUP.NCF file is automatically customized for each cluster server when you install GroupWise. After installation, you should review the STARTUP.NCF file on each server to ensure it is correct.

Optional GroupWise Configuration
Once the above GroupWise configuration requirements have been met, additional configuration might be necessary depending on how you will use GroupWise with NetWare Cluster Services. Optional configurations to consider are:

  • If you create multiple post offices on one server and configure them as separate resources for failover control, each POA must be loaded in protected memory space. If all POAs are loaded in kernel memory space, the first post office to unload will shutdown all other post offices. GroupWise 5.5 Enhancement Pack is currently the only GroupWise version capable of running in protected memory space. Ensure you have enough memory on each cluster server to accommodate the maximum number of POAs that could failover to that server.
  • You can have multiple POAs for one post office, but this requires that each POA be a separate resource, and that only one POA per post office mounts the volume which hosts the post office. This configuration is not encouraged.
  • You can associate the Message Transfer Agent (MTA) with a post office or configure it as a separate resource. Associating the MTA with a post office is the recommended configuration. Only one MTA runs on each domain.

Configure GroupWise for Protected Memory
Configuring GroupWise to use protected memory can give you a higher level of availability. It allows GroupWise to restart automatically on the same node after a failure instead of failing over to another node in the cluster. Using protected memory is required if you want to create multiple post offices on one server and configure them as separate resources for failover control. It is also recommended if you have GroupWise Post Offices configured to each have their own volume. GroupWise 5.5 Enhancement Pack is currently the only GroupWise version capable of running in protected memory.

To run GroupWise in protected memory,

  1. change the line that loads GWPOA.NLM in the GRPWISE.NCF file to LOAD ADDRESS SPACE=<NAME>RESTART GWPOA.NLM@<POA Startup file>
  2. Replace <NAME> with the name you want to give the protected memory space and <POA Startup file> with the name of the POA Startup file.

NetWare Cluster Services Configuration and Setup
Once GroupWise is installed and configured to run in a cluster environment, you must create and configure a GroupWise resource in NetWare Cluster Services. This includes configuring GroupWise load and unload scripts; setting GroupWise Start, Failover, and Failback modes; and assigning the GroupWise resource to specific servers in your cluster.

Creating a GroupWise Cluster Resource
NetWare Cluster Services includes a GroupWise resource template, which greatly simplifies the process for creating a GroupWise cluster resource. Much of the GroupWise cluster resource configuration is perfomed automatically by the GroupWise resource template.

  1. In ConsoleOne, browse and select the Cluster object.
  2. Click File, New, Cluster, Cluster Resource.
  3. Enter a name for the GroupWise cluster resource.
  4. Enter the GroupWise template name in the Inherit From Template field, or browse and select it from the list.
  5. Click Create.

The GroupWise resource template configures the GroupWise resource by automatically creating GroupWise load and unload scripts; setting Start, Failover and Failback modes and assigning GroupWise as a resource to all nodes in the cluster.

After creating a GroupWise resource, you should view or edit the GroupWise resource to ensure it is configured correctly for your specific needs.

Configuring GroupWise Load and Unload Scripts
The GroupWise resource template automatically creates a load and unload script to start and stop GroupWise on servers in your cluster. The load and unload scripts both contain IP address and volume-specific commands that you must customize for your specific GroupWise configuration.

To view or edit the GroupWise load and unload scripts:

  1. In ConsoleOne, right click the GroupWise Resource object and select Properties.
  2. Click the Load or Unload Script tab on the property page.
  3. Edit the IP address and volume-specific commands in both scripts. For example, you will need to edit the following two lines in the load script:
    • add secondary ipaddress A.B.C.D
    • nss /activate=VOLUME
  4. Replace A.B.C.D with the desired IP address and VOLUME with the name of the Groupwise volume on the shared disk system.
  5. Review the other lines in both scripts and edit them as necessary.

Setting GroupWise Start, Failover, and Failback Modes
The GroupWise resource template sets the GroupWise resource Start Mode and Failover Mode to AUTO, and the Failback Mode to DISABLE. With the GroupWise resource Start Mode set to AUTO, GroupWise will automatically start on a server when the cluster is first brought up. If the GroupWise Start Mode is set to MANUAL, you can manually start GroupWise on a server when you want, instead of having it automatically start when servers in the cluster are brought up.

With the GroupWise Failover Mode set to AUTO, GroupWise will automatically start on the next server in the Assigned Nodes list in the event of a hardware or software failure. If the GroupWise Failover Mode is set to MANUAL, you can intervene after a failure occurs and before GroupWise is moved to another node.

With the GroupWise Failback Mode set to DISABLE, Groupwise will not failback to its most preferred node when the most preferred node rejoins the cluster. If the Groupwise Failback Mode is set to AUTO, GroupWise will automatically failback to its most preferred node when the most preferred node rejoins the cluster. Set the GroupWise Failback Mode to MANUAL to prevent GroupWise from moving back to its preferred node when that node is brought back online, until you are ready to allow it to happen.

To view or change GroupWise Start, Failover, and Failback modes:

  1. In ConsoleOne, right click the GroupWise resource object and select Properties.
  2. Click the Policies tab on the property page.
  3. View or change the Start, Failover or Failback modes for GroupWise.

View or Edit GroupWise Resource Server Assignments
The GroupWise resource template automatically assigns the GroupWise resource to all nodes in the cluster. The order of assignment is the order the nodes appear in the resource list. To view or edit GroupWise resource node assignments or change the server failover order:

  1. In ConsoleOne, right click the GroupWise resource object and select Properties.
  2. Click the Nodes tab on the property page.
  3. View the current GroupWise resource server assignments, and if necessary, click the right or left arrow button to assign or unassign servers to the GroupWise resource.

You can click the up or down arrow button to change the failover order of the servers assigned to GroupWise.


Novell Cool Solutions (corporate web communities) are produced by WebWise Solutions. www.webwiseone.com

© 2014 Novell