ArubaOS 8.6.0.0Help Center
You are here: Home > Cluster > Cluster Deployment Scenarios

Cluster Deployment Scenarios

Clusters can be deployed in four different scenarios. The following section describes the guidelines for these different cluster deployment scenarios.

Arubarecommends you to enable AP load balancing for cluster using theCLICommand-Line Interface. A console interface with a command line shell that allows users to execute text input as commands and convert these commands to appropriate functions.commandactive-ap-lb. It is enabled by default.

Ifactive-ap-lbis disabled,LMSLocal Management Switch. In multi-controller networks, each controller acts as an LMS and terminates user traffic from the APs, processes, and forwards the traffic to the wired network.is used for the initial termination.LMSLocal Management Switch. In multi-controller networks, each controller acts as an LMS and terminates user traffic from the APs, processes, and forwards the traffic to the wired network.抢占用作described in all scenarios mentioned.

Scenario 1: Cluster with Virtual IP Setup

In this scenario, an AP performs a cluster failover to the S-AAC if the A-AAC (LMSLocal Management Switch. In multi-controller networks, each controller acts as an LMS and terminates user traffic from the APs, processes, and forwards the traffic to the wired network.) is down. The APs perform internal rebootstrap if both A-AAC and S-AAC are down at the same time. If the AP reboots on any node including theLMSLocal Management Switch. In multi-controller networks, each controller acts as an LMS and terminates user traffic from the APs, processes, and forwards the traffic to the wired network., the AP remembers the nodelist and tries all the entries in the nodelist. The AP performs a legacy rebootstrap only when it cannot reach any of the nodes.

Following are the guidelines to ensure a successful deployment of the cluster in a Virtual IP :

Master of the APs must be configured as the virtual IP on the cluster nodes.

If the cluster hasVRRPVirtual Router Redundancy Protocol. VRRP is an election protocol that dynamically assigns responsibility for a virtual router to one of the VRRP routers on a LAN.IP configured, set theVRRPVirtual Router Redundancy Protocol. VRRP is an election protocol that dynamically assigns responsibility for a virtual router to one of the VRRP routers on a LAN.IP inLMSLocal Management Switch. In multi-controller networks, each controller acts as an LMS and terminates user traffic from the APs, processes, and forwards the traffic to the wired network.IP address and backup-LMSLocal Management Switch. In multi-controller networks, each controller acts as an LMS and terminates user traffic from the APs, processes, and forwards the traffic to the wired network.IP地址。

Nodelist from the cluster node is saved on the AP. If the A-AAC and S-AAC are down at the same time, the AP will perform an internal rebootstrap and tries different nodes from the nodelist till the nodelist is exhausted.

Scenario 2: Cluster with Multiple Master via DNS resolution

In this scenario, an AP will perform a cluster failover to the S-AAC if the A-AAC (LMSLocal Management Switch. In multi-controller networks, each controller acts as an LMS and terminates user traffic from the APs, processes, and forwards the traffic to the wired network.) is down. The AP internally rebootstraps if both A-AAC and S-AAC are down at the same time and the AP tries to contact another node in the cluster till it is unable to reach the entire nodelist in the cluster. If the AP reboots on any node including theLMSLocal Management Switch. In multi-controller networks, each controller acts as an LMS and terminates user traffic from the APs, processes, and forwards the traffic to the wired network., the AP remembers the nodelist and tries all the entries in the nodelist. The AP performs a legacy rebootstrap only when it cannot reach any of the nodes.

Following are the guidelines to ensure for the successful deployment of the cluster in a multiple master viaDNSDomain Name System. A DNS server functions as a phone book for the intranet and Internet users. It converts human-readable computer host names into IP addresses and IP addresses into host names. It stores several records for a domain name such as an address 'A' record, name server (NS), and mail exchanger (MX) records. The Address 'A' record is the most important record that is stored in a DNS server, because it provides the required IP address for a network peripheral or element.resolution setup:

APs must get multiple masters using theDNSDomain Name System. A DNS server functions as a phone book for the intranet and Internet users. It converts human-readable computer host names into IP addresses and IP addresses into host names. It stores several records for a domain name such as an address 'A' record, name server (NS), and mail exchanger (MX) records. The Address 'A' record is the most important record that is stored in a DNS server, because it provides the required IP address for a network peripheral or element.resolution.

If the cluster hasVRRPVirtual Router Redundancy Protocol. VRRP is an election protocol that dynamically assigns responsibility for a virtual router to one of the VRRP routers on a LAN.IP configured, set theVRRPVirtual Router Redundancy Protocol. VRRP is an election protocol that dynamically assigns responsibility for a virtual router to one of the VRRP routers on a LAN.IP inLMSLocal Management Switch. In multi-controller networks, each controller acts as an LMS and terminates user traffic from the APs, processes, and forwards the traffic to the wired network.IP address and backup-LMSLocal Management Switch. In multi-controller networks, each controller acts as an LMS and terminates user traffic from the APs, processes, and forwards the traffic to the wired network.IP地址。

Nodelist from the cluster node is saved on the AP. If both A-AAC and S-AAC are down at the same time, the AP performs an internal rebootstrap and tries different nodes from the nodelist till the nodelist is exhausted.

In a large deployment,Arubarecommends this configuration to avoid large failure domain.

Scenario 3: Cluster with Virtual IP via DNS Resolution Across Data Centers

在这个场景中,当一个A-AAC下来,美联社ils over to an S-AAC. An AP internally rebootstraps if both A-AAC and S-AAC are down at the same time and the AP tries to contact another node in Cluster1 till all the nodes are exhausted in the Cluster1 nodelist. If the AP is not able to reach Cluster1, it fails over to the backupLMSLocal Management Switch. In multi-controller networks, each controller acts as an LMS and terminates user traffic from the APs, processes, and forwards the traffic to the wired network..

IfLMSLocal Management Switch. In multi-controller networks, each controller acts as an LMS and terminates user traffic from the APs, processes, and forwards the traffic to the wired network.preemption is enabled, APs preempt to Cluster1 when the primaryLMSLocal Management Switch. In multi-controller networks, each controller acts as an LMS and terminates user traffic from the APs, processes, and forwards the traffic to the wired network.node is up on Cluster1. The APs remain on Cluster2 if theLMSLocal Management Switch. In multi-controller networks, each controller acts as an LMS and terminates user traffic from the APs, processes, and forwards the traffic to the wired network.preemption is disabled even though the Cluster1 is up.

If the AP reboots on any node including theLMSLocal Management Switch. In multi-controller networks, each controller acts as an LMS and terminates user traffic from the APs, processes, and forwards the traffic to the wired network., the AP remembers the nodelist and tries all the entries in the nodelist. The AP performs a legacy rebootstrap only when it cannot reach any of the nodes.

Following are the guidelines to ensure a successful deployment of the cluster with Virtual IP viaDNSDomain Name System. A DNS server functions as a phone book for the intranet and Internet users. It converts human-readable computer host names into IP addresses and IP addresses into host names. It stores several records for a domain name such as an address 'A' record, name server (NS), and mail exchanger (MX) records. The Address 'A' record is the most important record that is stored in a DNS server, because it provides the required IP address for a network peripheral or element.resolution across data centers:

AP boots up and has two masters (one from each cluster) resolved from theDNSDomain Name System. A DNS server functions as a phone book for the intranet and Internet users. It converts human-readable computer host names into IP addresses and IP addresses into host names. It stores several records for a domain name such as an address 'A' record, name server (NS), and mail exchanger (MX) records. The Address 'A' record is the most important record that is stored in a DNS server, because it provides the required IP address for a network peripheral or element.server. The master of the AP is resolved to virtual IP of Cluster1 and virtual IP of Cluster2.

Nodelist from the cluster node is saved on the AP. If both A-AAC and S-AAC are down at the same time, the AP performs an internal rebootstrap and tries different nodes from the nodelist till the nodelist is exhausted.

If AP load balancing is disabled,LMSLocal Management Switch. In multi-controller networks, each controller acts as an LMS and terminates user traffic from the APs, processes, and forwards the traffic to the wired network.of the ap-group or ap-name must be the IP address of the Cluster1 node and the backup-LMSLocal Management Switch. In multi-controller networks, each controller acts as an LMS and terminates user traffic from the APs, processes, and forwards the traffic to the wired network.must be the IP address of the other node in the Cluster2. That is,LMSLocal Management Switch. In multi-controller networks, each controller acts as an LMS and terminates user traffic from the APs, processes, and forwards the traffic to the wired network.of the ap-group or ap-name must be configured to the Cluster1 node and the backup-LMSLocal Management Switch. In multi-controller networks, each controller acts as an LMS and terminates user traffic from the APs, processes, and forwards the traffic to the wired network.must be configured to the Cluster2 node.

In a large deployment,Arubarecommends this configuration to avoid large failure domain.

Scenario 4: Cluster with Multiple Master via DNS Resolution Across Data Centers

在这个场景中,当一个A-AAC下来,美联社ils over to an S-AAC. An AP internally rebootstraps if both A-AAC and S-AAC are down at the same time and the AP tries to contact another node in Cluster1 till all nodes in the nodelist of Cluster1 are exhausted. If the AP is unable to reach Cluster1, it fails over to the backupLMSLocal Management Switch. In multi-controller networks, each controller acts as an LMS and terminates user traffic from the APs, processes, and forwards the traffic to the wired network..

APs terminate on the node of Cluster2, which is configured as a backup-LMSLocal Management Switch. In multi-controller networks, each controller acts as an LMS and terminates user traffic from the APs, processes, and forwards the traffic to the wired network.using legacy failover.

IfLMSLocal Management Switch. In multi-controller networks, each controller acts as an LMS and terminates user traffic from the APs, processes, and forwards the traffic to the wired network.preemption is enabled, APs will preempt to Cluster1 when the primaryLMSLocal Management Switch. In multi-controller networks, each controller acts as an LMS and terminates user traffic from the APs, processes, and forwards the traffic to the wired network.node is up on Cluster1. APs remain on Cluster2 if theLMSLocal Management Switch. In multi-controller networks, each controller acts as an LMS and terminates user traffic from the APs, processes, and forwards the traffic to the wired network.preemption is disabled even though the Cluster1 is up.

If the AP reboots on any node including theLMSLocal Management Switch. In multi-controller networks, each controller acts as an LMS and terminates user traffic from the APs, processes, and forwards the traffic to the wired network., the AP remembers the nodelist and tries all the entries in the nodelist. The AP will perform a legacy rebootstrap only when it cannot reach any of the nodes.

Following are the guidelines to ensure a successful deployment of the cluster with multiple master viaDNSDomain Name System. A DNS server functions as a phone book for the intranet and Internet users. It converts human-readable computer host names into IP addresses and IP addresses into host names. It stores several records for a domain name such as an address 'A' record, name server (NS), and mail exchanger (MX) records. The Address 'A' record is the most important record that is stored in a DNS server, because it provides the required IP address for a network peripheral or element.resolution across data centers:

AP boots up and has four masters (two from each cluster) resolved from theDNSDomain Name System. A DNS server functions as a phone book for the intranet and Internet users. It converts human-readable computer host names into IP addresses and IP addresses into host names. It stores several records for a domain name such as an address 'A' record, name server (NS), and mail exchanger (MX) records. The Address 'A' record is the most important record that is stored in a DNS server, because it provides the required IP address for a network peripheral or element.server. The master of the AP must be resolved to two nodes in Cluster1 and two nodes in Cluster2.

Nodelist from the cluster node is saved on the AP. If both A-AAC and S-AAC are down at the same time, the AP performs an internal rebootstrap and will try different nodes from the nodelist till the nodelist is exhausted.

If AP load balancing is disabled,LMSLocal Management Switch. In multi-controller networks, each controller acts as an LMS and terminates user traffic from the APs, processes, and forwards the traffic to the wired network.of the ap-group or ap-name must be the IP address of the Cluster1 node and the backup-LMSLocal Management Switch. In multi-controller networks, each controller acts as an LMS and terminates user traffic from the APs, processes, and forwards the traffic to the wired network.should be the IP address of the other node in the Cluster2. That is,LMSLocal Management Switch. In multi-controller networks, each controller acts as an LMS and terminates user traffic from the APs, processes, and forwards the traffic to the wired network.of the ap-group or ap-name must be configured to the Cluster1 node and the backupLMSLocal Management Switch. In multi-controller networks, each controller acts as an LMS and terminates user traffic from the APs, processes, and forwards the traffic to the wired network.must be configured to the Cluster2 node.

Maximum of 10 entries are supported for master resolution. Combination of nodes from Cluster1 and Cluster2 can be used inDNSDomain Name System. A DNS server functions as a phone book for the intranet and Internet users. It converts human-readable computer host names into IP addresses and IP addresses into host names. It stores several records for a domain name such as an address 'A' record, name server (NS), and mail exchanger (MX) records. The Address 'A' record is the most important record that is stored in a DNS server, because it provides the required IP address for a network peripheral or element.master resolution.

/*]]>*/
Baidu