How to create an Airlock Failover Cluster setup

• To enhance the availability of a single Airlock installation, Airlock has the built-in capability to act as a cluster system. An Airlock cluster installation consists of two systems, where a passive system surveys its active partner to ensure availability, functionality and performance of the whole cluster.

We highly recommend creating a backup of the existing system before proceeding with following steps.

Prerequisites

• Use the identical hardware plattform for the two Airlock systems within a cluster and install the same Airlock version, including all updates on both systems.

Network infrastructure detail requirements

• New IP address for the new Airlocks Back-end interface.
• New IP address for the new Airlocks Management interface if required.
• New network segment with layer 3 connectivity for private failover IP addresses. This means there is no routing allowed between these IP addresses.
• New private failover IP address for each cluster member. They do not have to be reachable from the internet.
• If the failover cluster is deployed over two locations, a virtual switch is needed to garantie layer 3 connectivity.
• Please check that no port security is activated on the switch interfaces involved, since Airlock does move the Virtual IP addresses from on interface to the other in case of failover.
• License for an Airlock Failover Cluster. This one license needs at least one MAC address of each cluster member.

Configuration of first (exisiting) Airlock system

• Install and configure the first Airlock instance completely. At least one Virtual Host has to be configured.
• Test and verify your configuration and make sure that you monitor the alert messages.
• Go to "System Setup" - "Nodes" and configure a Private Failover IP address (e.g. 10.0.0.1/30) as well as for the failover partner a Mirror Failover IP address (e.g. 10.0.0.2/30). In the case of multiple external network interfaces, repeat these steps for each interface.
• Activate the new configuration on the first Airlock and verify that it works correctly.
• Go to "Configuration Files", select "Include private keys of certificates" and "Export" the activated configuration.

Configuring the second (new) Airlock system

• Install the second Airlock and login to the Configuration Center.
• Go to "Configuration Files" of the second Airlock and import the previously exported configuration file.
• Select "Set up a Failover Cluster".
• Activate the configuration on the second Airlock. This Airlock will stay "Passive" after the activation.
• Go to "Configuration Files" on the second Airlock, select "Include private keys of certificates" and "Export" the activated configuration which now contains the information for both hosts.
• Go back to "Configuration Files" on the first Airlock, "Import" the configuration file and activate it there as well. The two servers are now set up as a Failover Cluster. From now on you can use Remote Activation to activate modified configurations on both cluster servers simultaneously.

Now your two Airlock instances are working as an Airlock Failover Cluster. If one of the systems stops working, the standby system will take over, processing all requests. You can see the status of an Airlock in its Configuration Center, in the upper left corner underneath the Airlock logo.

Configuration changes on a running Failover Cluster

Starting with Airlock 4.2.3,  the same configuration can be activated on both systems of a Failover Cluster with one single step: The activation dialog gives a choice between local and remote activation. For example, in a cluster consisting of nodes "node1" and "node2", when clicking on "Activate" on node1, the activation dialog will give the options "Activate on node1" (local activation) and "Activate on node1 and node2" (remote activation). Remote activation simultaneously activates the current local configuration on both the local as well as the remote node.

Change Failover Healthcheck mechanism to a dedicated network infrastructure

To change the existing Failover Healthcheck mechanism from an existing network to a new dedicated network infrastructure do the following steps:

Start the migration on the passive node

• Log in to the passive Airlock node and add a new interface under "System Setup" > "Nodes" > "External network interface".
• On the just added interface define the new, different "Private Failover IP" and "Mirror Failover IP" available in the dedicated network infrastructure.
• Activate the configuration only on this passive node (at this point do NOT synchronize the configuration). After the activation the node should stay passive.
• Export the configuration (including the keys).
• Log off the passive node.

Finish the migration on the active node

• Log in the active node.
• Import the just saved configuration.
• Under "System Setup" > "Nodes" > "External network interface" select the new interface visible in the new dedicated network infrastructure.
• Remove the old existing setting for "Private Failover IP" and "Mirror Failover IP".
• Activate the configuration on both nodes. On the activation dialog select "Activate on node1 and node2". The configuration then will be synchronized.
• Check if the nodes remain in their existing failover state (passiv=passiv and active=active).