WiFi AC9700: Understanding AC Backup

Сообщение **xor** » 12 мар 2023, 23:25

Source: https://support.huawei.com/enterprise/e ... -ac-backup

WiFi AC9700: Understanding AC Backup

Backup Implementation and Process
Basic Concepts of AC Backup
Two ACs are deployed for an AP. One AC manages the AP, and the other AC functions as the backup AC. When an AC fails, the backup AC takes over services to improve network reliability.

Figure 23-1: Basic AC backup networking

: Figure 23-1: Basic AC backup networking

In the AC backup networking shown in Figure 23-1, an AP is connected to AC1 (active) and AC2 (standby), which are in working and backup states, respectively.
AC roles are specified during data planning and will not change upon failures or recovery of the network and devices.

Active AC: manages APs and provides services for them. As shown in Figure 23-1, AC1 functions as the active AC and manage APs when the network and devices are working properly.
Standby AC: acts as the backup of the active AC. As shown in Figure 23-1, AC2 functions as the standby AC and will take over services from AC1 upon a failure of the network or devices.

To ensure service reliability of APs and STAs, the states of an AC may change due to failures or recovery of the network and devices.

Working state: An AC is AP is managing APs and providing services for them. As shown in Figure 23-1, AC1 is in working state before it fails. After AC1 fails, AC2 enters the working state. After AC1 recovers, AC1 switches to the working state again.
Backup state: An AC does not manage or provide services for APs. It only waits to take over services from the working AC if the working AC fails. As shown in Figure 23-1, AC2 stays in backup state before AC1 fails. After AC1 recovers, AC2 changes from the working state to the backup state.

Active/Standby and Load Balancing Networking Modes
Basic Concepts of AC Backup describes the concepts involved AC backup from the perspective of an AP. In active/standby mode, AC1 is the active AC and AC2 is the standby AC for all APs on the network. Generally, AC1 processes all services, and AC2 only backs up services but does not process services. Figure 23-2 shows the active/standby networking.

Figure 23-2: Active/standby networking

: Figure 23-2: Active/standby networking

In this networking, the active AC is heavily loaded with service data, while the standby AC is idle. Therefore, AC resources are not fully utilized. The load balancing mode is introduced to address this. In the load balancing networking shown in Figure 23-3, AC1 serves as the active AC and AC2 as the standby AC for some APs; while for other APs, AC2 serves as the active AC and AC1 as the standby AC. In this way, both AC1 and AC2 process service data, effectively leveraging resources.

Figure 23-3: Load balancing networking

: Figure 23-3: Load balancing networking

For AP1, AC1 is the active AC and stays in working state, and AC2 is the standby AC and stays in backup state.
For AP2, AC2 is the active AC and stays in working state, and AC1 is the standby AC and stays in backup state.

AC Backup Process
In both active/standby and load balancing modes, the AC backup process consists of active/standby negotiation, data backup, active/standby switchover, and active/standby switchback.

Table 23-4: Backup process

Backup Phase	Description
Active/standby negotiation	In active/standby negotiation, multiple ACs use Virtual Router Redundancy Protocol (VRRP) or CAPWAP to negotiate the working and backup states based on the active/standby data plan. APs go online on the working AC and are managed by it. If both the active and standby ACs can communicate with the AP, the active AC is in working state and the standby AC in backup mode after the negotiation. If APs lose connectivity with the active AC during active/standby negotiation, the standby AC is negotiated to the working state.
Data backup	The AC in working state backs up AP, STA, and CAPWAP link information to the AC in backup state. Data backup ensures information consistency between the two ACs. When the AC state changes, sessions are not interrupted, and the time for service recovery on STAs is reduced.
Active/standby switchover	When the standby AC detects a failure of the active AC or the CAPWAP link between the active AC and APs, the standby AC automatically switches to the working state to take over services from the active AC, ensuring service continuity on STAs. Before an active/standby switchover, the active AC is in working state and the standby AC in backup state. After the active/standby switchover, the standby AC enters the working state.
Active/standby switchback	After the active AC recovers, it takes over services from the AC in working state and becomes the active AC again. This process is called active/standby switchback. Before an active/standby switchback, the active AC is in backup state and the standby AC in working state. After the switchback, the active AC enters the working state and the backup AC restores to the backup state.

Essentially, active/standby negotiation, switchover, and switchback are different phases when the active and standby ACs switches between different states. In some documents, the switchover from the active AC to the standby AC or switchback refers to switching ACs to the working or backup.

Application Scenarios and Modes of AC Backup
Different backup modes are available to meet differentiated reliability requirements in various network scenarios. ACs support VRRP hot standby (HSB), dual-link HSB, dual-link cold backup, and N+1 backup.

VRRP HSB is applicable to campus networks with high network reliability requirements, such as enterprises and schools. In VRRP HSB, the switchover speed is fast, with little impact on services. If the active AC or CAPWAP link fails, the standby AC quickly takes over services from the active AC to manage APs, ensuring service continuity.

However, VRRP HSB requires the Layer 2 networking between the active and standby ACs but does not support AC load balancing. If you want to maximize the resource utilization of AC resources or no Layer 2 network is available between the active and standby ACs, dual-link HSB can be used.

In dual-link HSB, the switchover response is slower than that in VRRP HSB. Additionally, neither VRRP HSB or dual-link HSB is applicable to remote disaster recovery scenarios. If the active and standby ACs are deployed in different places and the reliability requirement is not high, dual-link cold backup can be used.

To reduce network deployment costs, you can use the N+1 backup mode. Compared with VRRP HSB, dual-link HSB, and dual-link cold backup, N+1 backup does not require a standby AC for each active AC, thereby reducing costs.

N+1 backup has the lowest reliability. To improve the reliability in N+1 backup, you can deploy both VRRP HSB and N+1 backup so that every two ACs form a VRRP group and are presented as one virtual device. N+1 backup works between different virtual devices.

For details about the differences between backup modes, see Comparison of AC Backup Modes.

HSB
Data backup in AC Backup Process is implemented through hot standby (HSB). When the active AC fails, service traffic can be switched to the standby AC only if the standby AC has the same session entries as the active AC. The active and standby ACs use the HSB mechanism to ensure the consistency of session entries.

HSB provides two types of public services: HSB service and HSB group.

HSB service
An HSB service establishes an HSB channel for transmitting packets of other services and maintains the link status by notifying the HSB group of the faulty link.
An HSB service provides the following functions:

Establishing an HSB channel: A TCP channel is established for sending HSB packets by setting the IP addresses and port numbers of the local and peer devices. The HSB service provides packet sending and receiving for other services and notifies link status changes.
Maintaining the link status of the HSB channel: HSB packets are sent and retransmitted to prevent long TCP interruption that is not detected by the protocol stack. If a device does not receive an HSB packet from the peer device within the period (retransmission interval x retransmission times), the local device receives a message indicating the exception and then re-establishes a channel to the peer.

HSB group
An HSB group instructs service modules to perform batch backup, real-time backup, and status synchronization. Service backup depends on the status negotiation and event notification mechanisms provided by an HSB group to synchronize services between the active and standby devices.
An HSB group synchronizes backup information and responds to link status changes through the HSB channel established by an HSB service. The HSB service needs to be bound to the HSB group so that the HSB group can work properly. In addition, the HSB group must be bound to a VRRP group to negotiate the service status based on the VRRP status. By monitoring the changes in the bound channel status and VRRP status, the HSB group instructs service modules to perform batch backup, real-time backup, and status synchronization.
HSB supports three data synchronization modes: batch backup, real-time backup, and periodic synchronization.

Table 23-5: Data synchronization mode

Mode	Description
Batch backup	When the active and standby ACs are determined, the active AC synchronizes the existing session entries to the new standby AC at a time to ensure that the session entries on the active and standby ACs are the same.
Real-time backup	When the active AC generates new session entries or modifies existing session entries, it synchronizes new or modified session entries to the standby AC in real time.
Periodic synchronization	To ensure that entries on the active and standby ACs are consistent, the standby AC checks whether session entries are the same as those on the active AC every 30 minutes. If session entries are inconsistent, the session entries on the active AC are updated to the standby AC.

Comparison of AC Backup Modes
ACs support VRRP HSB, dual-link HSB, dual-link cold backup, and N+1 backup. Table 23-6 compares these backup modes.

Table 23-6: Comparison of AC backup modes

Item	VRRP HSB	Dual-Link HSB	Dual-Link Cold Backup	N+1 Backup
Implementation	The active and standby ACs have independent IP addresses, which are virtualized into one using VRRP. APs set up CAPWAP links with this virtual IP address. The active AC backs up information about APs, STAs, and CAPWAP links, and synchronizes such information to the standby AC through the HSB service. If the active AC fails, the standby AC takes over services.	An AP sets up an active and a standby CAPWAP link with the active and standby ACs, respectively. The active AC backs up only STA information and synchronizes such information to the standby AC through the HSB service. If the active AC fails, APs connected to it switch to the standby links and the standby AC takes over services.	An AP sets up an active and a standby CAPWAP link with the active and standby ACs, respectively. ACs do not back up or synchronize information. If the active AC fails, APs connected to it switch to the standby links and the standby AC takes over services.	An AP sets up a CAPWAP link with only one AC. ACs do not back up or synchronize information. If the active AC fails, APs connected to it set up CAPWAP links with the standby AC that takes over services.
Switchover speed	Fastest The switchover speed is fast, with little impact on services. The configuration of the VRRP preemption delay implements a faster switchover than other backup modes.	Fast The AP status switchover is slow and occurs only when CAPWAP link disconnection timeout is detected. After the AP status is switched, STAs do not need to go offline and online again.	Slow The AP status switchover is slow and occurs only when CAPWAP link disconnection timeout is detected. STAs need to go online again, and services are interrupted for a short period of time.	Slowest The AP status switchover is slow and occurs only when CAPWAP link disconnection timeout is detected. APs and STAs need to go online again, and services are interrupted for a short period of time, which is longer than the service interruption period in dual-link cold backup mode.
Deployment of active and standby ACs at different places	Not supported VRRP is a Layer 2 protocol and does not support deployment of active and standby ACs at different places.	Not recommended	Supported	Supported
Constraints	The models and software versions of the active and standby ACs must be the same.	The models and software versions of the active and standby ACs must be the same.	The software versions of the active and standby ACs must be the same. No constraint is placed on the AC model.	The models and software versions of the active and standby ACs can be different. However, it is recommended that they be the same on the two ACs.
Applicable scope	Scenarios that require high reliability, without the need for AC deployment at different places	Scenarios that require high reliability and AC deployment at different places	Scenarios with low reliability requirements	Scenarios with low reliability but high cost control requirements