Redundancy Groups¶
Redundancy groups enable high-availability scenarios. In the event of a failure of the node acting as the master for a given redundancy group, the IP addresses assigned to that group are taken over by another node with the highest priority within that group. Assigning different priorities to redundancy groups on individual cluster nodes allows for static load balancing while maintaining high-availability functionality.
Note
The redundancy group configuration options are available after the cluster has been initialized.
Adding Redundancy Groups
To add a redundancy group, follow the instructions below.
From the left menu, select > .
Go to the Redundancy Groups tab.
Click .
Define the group parameters:
Name – the name of the redundancy group.
ID – the redundancy group identifier (allowed range: 1–255).
Priority – the redundancy group priority (allowed range: 0–254). A lower parameter value indicates a higher priority.
Network interface – the network interface used by the redundancy group to communicate with other cluster nodes.
Note
Redundancy group with higher priority assumes the master role and handles all requests to monitored servers accessed through IP addresses assigned to this group. In case given cluster node crashes, user requests are directed to on of the remaining nodes with the highest priority defined for given redundancy group.
The network interface (interlink interface) is used to monitor the state of a given redundancy group. The master node broadcasts keep-alive packets over the Layer 2 network to inform other nodes that it is active and operational. The remaining cluster nodes use the interlink interface to listen for these packets.
All cluster nodes must have their interlink interfaces connected to the same broadcast domain (VLAN). Additionally, in VMware-based environments, the vSwitch security settings must be adjusted to allow MAC address changes and forged transmits.
By default, once a node takes the master role, it will continue on indefinitely as the master node.
Click .
From the left menu, select > .
Click to add an IP address.
Enter the IP address and select the General function to assign the cluster attribute.
From the drop-down list, select the previously defined redundancy group.
Click .
Note
The cluster IP address must be defined on each cluster node.
Editing Redundancy Groups
To modify a redundancy group, follow the instructions below.
From the left menu, select > .
Go to the Redundancy Groups tab.
Modify the parameters of the selected redundancy group.
Click .
Removing Redundancy Groups
To remove a redundancy group, follow the instructions below.
From the left menu, select > .
Go to the Redundancy Groups tab.
Click for the chosen redundancy group.
Click .
Demoting a Redundancy Group
Note
Demoting a group switches the master role for the given redundancy group to another cluster node. The node on which the selected redundancy group has the highest priority will assume the master role.
To demote a redundancy group, follow the instructions below.
From the left menu, select > .
Go to the Redundancy Groups tab.
Edit the group you want to demote.
Click for the selected redundancy group.
Click .
Note
If, after demoting the group, none of the remaining nodes assumes the master role, it will be restored to the redundancy group on the edited node.
Forcing the Slave Role
Note
Forcing the slave role ensures that the redundancy group will never switch to master mode, regardless of the state of the other cluster nodes.
Forcing the slave role is recommended before performing maintenance work to ensure that network traffic is redirected to the remaining cluster nodes. Another use case is a cluster node deployed in a separate location without Layer 2 connectivity to the other cluster nodes.
To force the slave role for a selected redundancy group, follow the instructions below.
From the left menu, select > .
Go to the Redundancy Groups tab.
Locate the redundancy group and click its name to edit it.
From the Cluster communication interface drop-down list, select
Force slave mode.Click .
Related topics: