Clustering

The 2021.1 release of File Director removed the requirement for an external SQL database, simplifying infrastructure requirements.
The sections of help content applicable to database configuration and earlier releases of File Director are identified. See also What's New.

File Director supports clustered infrastructures and failover processing and is fully scalable to meet the varying demands of organizations. Appliances within the cluster share the same configuration settings.

In this section:

Overview

If an appliance is taken offline, the File Director service is maintained by having users log on in the background to an alternate appliance in the cluster, according to the current network load balancing method. Although users are momentarily disconnected, they are automatically returned to the service without losing their session state.

(2020.3 and earlier) Any transactions that have not been committed to the database are rolled back.

When setting up the load balancer, ensure that session persistence is setup for the File Director cluster. It is recommended that the cookie insert method is used.

When clustering is enabled the following configuration is replicated between nodes:

  • DNS configuration
  • SSL certificate (not the pending CSR state, only when a certificate has been fully applied)
  • Syslog server
  • QoS configuration
  • License
  • Global Web Client access
  • HTTP mode
  • NTP configuration
  • Kerberos configuration
  • Settings such as Directory Services (LDAP), Cloud Connector configuration, Map Points and Policy (2021.1 and later)

The following configuration items are not replicated:

  • Maintenance mode
  • TLS configuration

Prerequisites

Before configuring clustering:

  • Ensure all appliances that will be in your cluster are of the same version.
  • Configure the switching environment to allow Broadcast traffic.
  • Ensure all cluster nodes that are to share common settings are available on the same network to allow low frequency broadcast discovery between the cluster peers.
  • Take a backup of the current appliance configuration.

Enable clustering on the first File Director appliance

Before you enable clustering it is strongly recommended you take an appliance snapshot.

  1. Open the Admin console for the first cluster node.
  2. Click the Cluster tab.
  3. Enter a cluster name.
  4. Enter a cluster port number. The default port is 49152 but you can use any port from 49152 to 65535.
  5. To save the cluster configuration, click Save.
  6. A cluster configuration change message is displayed prompting you to confirm your action. Click OK.

    If File Director detects your selected node has non-default configuration settings applied, a message is displayed. You will need to determine whether you want the values of your node configuration to be overwritten with those used by the cluster.

    • If the node is being added to an existing cluster it is likely that you want the existing cluster settings to overwrite those of the current node. Click Yes.
    • If the node is the first one to be added to a cluster click No.
Disable clustering

To disable clustering, follow the process above and at step 3, leave the cluster name field blank.

Configure the appliance on the initial cluster node

It is recommend that once you have enabled clustering on the first node, you configure the appliance settings or restore a backup with the required settings configured. When further nodes are added to the cluster, the appliance settings are automatically applied.

Check the Load Balancer Status

  1. Select Configuration > Advanced.
  2. Locate the Load Balancer Status section.

  3. Click the Status URL link to check the health status of a server in a load balanced environment. A status page is displayed showing one of the following:
    • Success - The server is functioning correctly within the load balancer pool.
    • Failure - The server is either offline or is not functioning correctly within the load balancer pool.

Configure additional cluster nodes

Once you have successfully configured the initial cluster node, configure all the nodes you want to be part of the cluster.

  1. Log on to the Admin console for the node you want to add.
  2. Upload a License File.
  3. Enter the name of the cluster.

    This must be the name you entered when setting up the initial cluster node (see Enable clustering on the first File Director appliance, above.

  4. If you are not using the standard port number (49152), enter the port number you are using for your cluster.
  5. To save the cluster configuration, click Save.

    If File Director detects your selected node has non-default configuration settings applied, a message is displayed. You will need to determine whether you want the values of your node configuration to be overwritten with those used by the cluster - or not.

    • If the node is being added to an existing cluster it is likely that you want the existing cluster settings to overwrite those of the current node. Click Yes.
    • If the node is the first one to be added to a cluster click No.

To confirm clustering is operating correctly, log on to the web admin for the node and make a simple change, such as changing the DNS settings. When you log on to another node in the cluster, the same change should be apparent.

You will need to repeat this process for each new node that you add to the cluster.

Manage cluster status in the Admin console

From the Status view you can perform various cluster management tasks. You can promote a node as a patch server; check the status of the nodes in your cluster; enable and disable maintenance mode; apply a patch to your cluster; and apply various node-specific action buttons.

Patch server

By nominating one of the nodes in a cluster as the patch server, you can apply patches to all nodes in the cluster from one console. Any active node in a cluster can be used as the patch server.

The Cluster tab within File Director lists all nodes within the cluster. If no patch server is set, a warning message is displayed at the top of the Cluster Configuration dialog. Alternatively, when a node is selected as the patch server it is easily identified within the listed nodes.

To make a node the patch server, log into the Admin console for that node. In the Clustering tab click Promote to Patch Server. The current node is now identified as the patch server.

Cluster status

The status shows the name of the cluster, the File Director server version, and which node in the cluster is currently the patch server.

If a patch server has not been set, these details are not displayed.

Node status

The state of each node in the cluster is denoted by the icon displayed in the Status column.

Icon Meaning
Active The node is online and using the correct File Director server version, determined for the cluster by the version applied to the patch server.

Warning This can signify one of the following states:

  • A patch server has not been set. Set one of the nodes as the patch server.
  • The node is in maintenance mode. To remove maintenance mode, check the Enabled option.
  • The node's File Director server version is different to that of the patch server.
    Reapply the current patch to the cluster. This updates only those nodes that are not at the File Director server version applied to the patch server. Nodes already at the correct version are unaffected by the update.
  • The node requires a reboot. Click the Reboot button for the node.
  • Component information cannot be retrieved.
Inactive The node is offline.

Enabled

The Enabled checkbox allows you to control the status of each node within your cluster. The node status toggles from enabled to disabled as the checkbox is selected or cleared. When cleared (or disabled), the node is placed in maintenance mode and is not available for load balancing. See Load Balancer status.

To enable or disable a selected node, check or clear the Enabled checkbox for each node required. To select all nodes within your cluster, click the menu icon and select Enable All Nodes or Disable All Nodes as required.

Host

Displays the name of each node in the cluster, identifies which node is the patch server, and which node you are currently accessing through the web Admin console. Click on a name to see details of the current component versions of that node and its patching history. The name of the current node and the patch server are annotated appropriately.

IP address

The IP address of each node in the cluster.

Action

You have a number of action buttons available in the Action column. When selected, the action is applied to the specific node only:

  • Logs

    Appliance logs are available for download for each appliance in your configuration. They enable auditing and trouble-shooting. Press Logs to download the appliance logs for each node required.

  • Reboot

    Remotely reboot the node.

  • Remove

    If a node is inactive press Remove to remove it from the list. If a removed node is subsequently restarted, it will automatically re-display in the list.

  • Shutdown

    Remotely shutdown the node.

Related topics

Configure the Appliance

Apply a File Director Patch

Diagnostics and Troubleshooting