Resource configuration planning

Resources in the Windows HA cluster include the application and all the shared file systems that belong to it. Windows HA provides an environment by identifying a set of cluster-wide resources essential to uninterrupted processing defining relationships among these nodes ensuring resources are available to client processes. When a cluster node fails or detaches from the cluster for a scheduled outage, the Cluster Manager redistributes its resources among any number of the surviving nodes.

Each resource in a cluster is defined as part of a Resource Group or a Role. This allows you to combine related resources that need to be clustered to provide a particular service. A Role includes the list of nodes that can acquire those resources and serve them to clients. Resources can either be cascading or rotating.

When configuring the InfoPrint Manager clustering, you have to add the InfoPrint services as resources in the created Resource Group or Role.

Before adding the Infoprint Manager services as resources, the InfoPrint Manager shared folders need to be created within the shared disk in the cluster (CSV) and the cluster nodes need to be rearranged.

Changes within the primary node

  1. Mount the CSV.

    Failover forces the CSV to be mounted within the other node.

  2. On the CSV storage, create this folder structure:

    • RICOH\InfoPrint Manager
    • RICOH\InfoPrint Manager\var\pd
    • RICOH\InfoPrint Manager\var\psf
    • RICOH\InfoPrint Manager\var\pddir

  3. After you create the folder structure, start installing InfoPrint Manager for Windows on the primary node and on all existing secondary nodes.

  4. On the first panel, the installer asks you to choose between Standard or HA installation. To install InfoPrint Manager on a Windows Server Cluster for High Availability, select HA.

  5. On the next installer panel, you can choose between HA primary and HA failover installation.

      Note:

    • When installing InfoPrint Manager on the primary node, select HA primary.

    • When installing InfoPrint Manager on any of the secondary nodes, select HA failover.

  6. Continue with the installation in the usual manner. See RICOH InfoPrint Manager for Windows: Getting Started for more information.

  7. After creating the folder structure and installing InfoPrint Manager for Windows, follow these steps:

    1. Stop all the InfoPrint Manager services within the primary node, if they are running. By default for a HA installation, the InfoPrint Manager services do not start automatically after the installation.

    2. After you create these folders, move the content from the primary node to the CSV:

      • Move the primary node \var\pd content to the \var\pd folder you created on the cluster shared volume. Within the primary node, the var\pd folder must become a Windows junction to the \var\pd shared volume folder.

      • Move the primary node \var\psf to the \var\psf folder you created on the cluster shared volume. Within the primary node, the \var\psf folder must become a Windows junction to the \var\psf shared volume folder.

      • Move the primary node \var\pddir to the \var\pddir folder you created on the cluster shared volume, and then the primary node \var\pddir must become a Windows junction to the \var\pddir folder you created on the cluster shared volume.

    3. Start the InfoPrint Manager services and check if they are running after the primary node modifications.

    4. Stop all the InfoPrint Manager services.

Changes within the failover node

Note: By default, a HA failover installation does not create the \var\pd, \var\pddir and \var\psf folders. In this case, you must create the \var\psf folder and \var\psf\segments subfolder manually. For each node, you must create a unique segment with a unique name.

  1. Remove the content of the failover node, \var\pd folder, if it exists.

  2. Remove the content of the failover node, \var\psf folder.

  3. Remove the content of the failover node, \var\pddir folder, if it exists.

  4. Move the shared volume to the failover node using the Cluster Manager.

  5. Create a Windows junction from the failover nodes \var\pd to the \var\pd folder you created on the cluster shared volume.

  6. Create Windows junctions from the failover node \var\psf to the \var\psf folder you created on the cluster shared volume.

  7. Create a Windows junction from the \var\pddir folder to the \var\pddir folder you created on the cluster shared volume.

  8. To test if the InfoPrint Manager services start properly after making these changes within the failover node, change the state of the InfoPrint Manager services to Manual mode.

  9. Stop all the InfoPrint Manager services if you have started them and change the state of the InfoPrint Manager services to Disabled mode.

  10. Move the shared disk back to the primary node by using the Cluster Manager.

File system configuration planning

InfoPrint Manager shared file systems include these:

  • /var/pd

  • /var/psf

  • /var/pddir/

The cluster administrator must configure the \var\psf\segments folder path on each node before starting InfoPrint Manager. Each node uses its own segments directory. The first node uses C:\ProgramData\RICOH\InfoPrint Manager\var\psf\segments1 folder, the second node uses C:\ProgramData\RICOH\InfoPrint Manager\var\psf\segments2 folder, and so on.

You must configure each node through the Management Console application. Select Option from the menu and choose the Customize Input Manager option.

Add resources to the InfoPrint Manager cluster group

  1. Right click on the group you have created and select Add a resource->Generic Service.

    Note: All InfoPrint Manager services must be added as resources within the created group.

  2. The InfoPrint Manager services are displayed on the next panel.

Cluster resource configuration

After you add all the InfoPrint Manager resources to the cluster, you must configure them.

First, you must configure the dependencies between resources, the same as for a normal installation on a Windows machine.

These services or cluster resources are dependent on the service license:

  • InfoPrint Manager Service

  • InfoPrint Manager LPD

  • InfoPrint Manager MVS Download Receiver

  • InfoPrint Manager SAP

You must configure them individually, by right clicking on the resource you want to configure. Select the Properties tab and add the license service as a dependency.
Note: The services or cluster resources that are not listed above do not need any dependency configuration.

After you configure the service dependencies, you must configure the registry entries. Right-click each cluster group resource, select the Option tab, and add these registry entries:

  • InfoPrint Manager Registry replication.

    • HKEY_LOCAL_MACHINE\SOFTWARE\RICOH\Infoprint Manager\Configuration

    • HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\services\InfoprintManager

  • InfoPrint Manager LPD Service Registry replication

    • HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\services\Infoprint Manager LPD

  • InfoPrint Manager MVS Download Receiver Service Registry replication

    • HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\services\Infoprint Manager MVS Download

    • HKEY_LOCAL_MACHINE\SOFTWARE\RICOH\InfoPrint Manager\MVS Download

  • InfoPrint Manager DPF Service Registry replication

    • HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\services\InfoprintManagerDPF

  • InfoPrint Manager IPP Gateway Registry replication

    • HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\services\InfoprintManagerIPPGateway

  • License Service Registry replication

    • HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\services\InfoprintManagerLicenseService

  • InfoPrint Manager SAP Service Registry replication

    • HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\services\InfoprintManagerSAP

  • InfoPrint Manager Notification Service Registry replication

    • HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\services\InfoprintManagerNotification

  • InfoPrint Manager WebServer Service Registry replication

    • HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\services\InfoPrintManagerWebServer

Note: After you configured the cluster, you can start or stop the InfoPrint Manager services from the Cluster Manager.

To check if the InfoPrint Manager services were moved on the failover node, you can do a manual failover procedure within the Cluster Manager.

The server name must not be changed when InfoPrint Manager runs on the cluster.

The jobs will fail if you do this chain action: send the jobs with rip-and-hold, change the server name from the console, and you release the jobs

Printers and Ports replication

Cluster administrator must assure that all printers are correctly replicated within the all failover nodes.

Note: InfoPrint manager provides a printer migration script to help InfoPrint manager printers to migrate. It is a sample script and the cluster administrator can customize this migration script. By default, the name of the script is IPMHAReplicate.ps1 and its log is located within the cluster shared volume. When configuring the cluster, make sure that all cluster nodes in the printers, ports, and drivers are identically configured.
Printers and ports can be created on the node where InfoPrint Manager is running. They must be replicated by the cluster administrator within all failover nodes. This operation should happen before any failover transition.

The default paths internally used by the IPMHAReplicate.ps1 script:

  • "C:\ClusterStorage\Volume1\RICOH\InfoPrint Manager\printers.csv" – default place for printers.csv file

  • "C:\ClusterStorage\Volume1\RICOH\InfoPrint Manager\migrateLog.log" – default place for script log

  • “C:\ClusterStorage\Volume1\RICOH\InfoPrint Manager\IPMPorts.reg" – default place for ports exported file

If the cluster is set using a different path, “C:\ClusterStorage\DISK1\RICOH\InfoPrint Manager”, modify the script default path in the IPMHAReplicate.ps1 script.

The IPMHAReplicate.ps1 script default output:

  • Printers.csv contains serialized printers.

  • migrateLog.log script log file.

  • IPMPorts.reg contains exported registers required for InfoPrint Manager.

On the machine where you created printers, export them by running the IPMHAReplicate.ps1 script usage as follows:

  1. Run .\ IPMHAReplicate.ps1 export -ports without parameters on the node where the Cluster Shared Volume is mapped.

    On the node where the Cluster Shared Volume is not mapped, run .\ IPMHAReplicate.ps1 export -ports with parameters, <printers.csv file> <logfile>.

  2. Run .\ IPMHAReplicate.ps1 export -printers without parameters on the node where the CSV volume mapped.

    On the node where the Cluster Shared Volume is not mapped, run .\ IPMHAReplicate.ps1 export -printers with parameters, <printers.csv file> <logfile>.

The IPMHAReplicate.ps1 script usage, on the machine where you want to replicate the printers, follow these steps:

  1. Run .\ IPMHAReplicate.ps1 import -ports without parameters on the node where the CSV volume mapped.

    On the node where the Cluster Shared Volume is not mapped, run .\ IPMHAReplicate.ps1 import -ports with parameters, <portfile> <logfile>.

  2. Run .\ IPMHAReplicate.ps1 import -printers without parameters on the node where the CSV volume mapped.

    On the node where the Cluster Shared Volume is not mapped, run .\ IPMHAReplicate.ps1 import -printers with parameters, <portfile> <logfile>.

The delete printers action is optional.

Run .\ IPMHAReplicate.ps1 delete-printers without parameters on the node where the CSV volume mapped.

On the node where the Cluster Shared Volume is not mapped, run .\ IPMHAReplicate.ps1 delete-printers with parameters, <logfile>.

The default script usage:

  • Export ports on primary node.

  • Export printers on primary node.

  • Place the generated outputs, IMPorts.reg and printres.csv, on the failover node or nodes on a custom path.

  • Import ports on failover node or nodes. Use the command with parameters.

  • Import printers on failover node or nodes. Use the command with parameters.

    Note: Do not import printers before ports as the command may fails to create printers on the failover node or nodes due to missing ports on that machine.

Parent topic: Planning and implementing for Windows