CN106603261B - Hot backup method, first main device, standby device and communication system - Google Patents

Abstract

Embodiments of the present invention provide a hot backup method, a first active device, a backup device and a communication system, which help to save system resources. The method includes: the first master device acquires the first link quality level from the first master device to the standby device, the X1th link quality level from the first master device to other N- ₁ master devices, and the first link quality level from the first master device to other N-1 master devices. The Y-th link quality level from N-1 active equipment to the backup equipment, when the sum of the first link quality level and the number of abnormal levels in the X1- _th link quality level is less than the abnormal level in the Y-th link quality level and the number of normal levels in the _X1th link quality level is greater than or equal to the number of normal levels in the Yth link quality level, and the link quality level of the first active device itself is an abnormal level, the first active device The device is hot backed up to the standby device; otherwise, the first active device remains in the state of the active device.

Description

Hot backup method, first active device, backup device and communication system

technical field

The embodiment of the present invention relates to the technical field of communication, and in particular to a hot backup method, a first active device, a backup device and a communication system.

Background technique

With the development of communication technology, more and more functions are integrated on the communication equipment in the network. The failure of communication equipment will cause the terminal equipment served by it to fail to communicate normally, and may even affect the normal operation of the entire network. Therefore, in general, each communication device is equipped with a backup channel to ensure that the communication device can still work normally when the main communication channel fails, and the network is also configured with a backup device so that the entire network can work normally when a certain device fails.

Hot backup technology is a reliable protection technology in the network. When a single communication device in the network fails, the business in the failed communication device will be quickly switched to the standby communication device, and the terminal device will not be aware of the communication device. failover. Hot backup generally has the form of dual-machine hot backup and multi-machine hot backup. Among them, dual-machine hot backup means that one backup communication device protects one main communication device; multi-machine hot backup means that one backup communication device protects N main communication devices. Obviously, dual-machine hot backup is a kind of multi-machine hot backup.

Hot backup has a wide range of application scenarios. For example, for a broadband remote access server (Broadband Remote Access Server, BRAS), a virtual router redundancy protocol (Virtual Router Redundancy Protocol, VRRP) is used to implement hot backup. Take two BRAS for dual-system hot backup as an example. When the active BRAS device fails, active-standby switchover occurs, and the data traffic on the active BRAS, whether uplink or downlink, will be switched to the standby BRAS. During the entire switchover process , the user device will not be aware, so there is no need for the user device to reconnect to the network. And when the failure of the active BRAS recovers, it will switch to the active BRAS to continue working.

The usual multi-machine hot backup technology can only make correct judgments on link interruptions, but it is easy to make wrong judgments on serious link packet loss, which may lead to waste of resources.

Contents of the invention

An embodiment of the present invention provides a hot backup method, a first active device, a backup device, and a communication system. By judging the link quality between each active device and the backup device, it is determined whether to back up, so that each communication in the backup system The devices can communicate normally, thus saving system resources.

In the first aspect, a hot backup method is provided, including:

The first active device acquires a first link quality level, where the first link quality level is used to represent the link quality between the first active device and the backup device;

When the first link quality level indicates that the link is abnormal, the first active device acquires the N-1 X _i -th links between the first active device and other N-1 active devices Road quality level, the X _i -th link quality level is used to indicate the link quality between the first active device and the i-th active device, N is greater than or equal to 2, i is greater than 1 and less than or equal to N ;

The first active device sends a request message to each of the other N-1 active devices, and the request message is used to request the other N-1 active devices Each active device in the above respectively sends the link quality level between the other N-1 active devices and the backup device;

The first active device respectively receives the M Y _j -th link quality levels between the j-th active device and the backup device sent by the M active devices among the other N-1 active devices, The Y _jth link quality level is used to represent the link quality between the jth active device and the backup device, j is greater than 1 and less than or equal to N, and M is greater than or equal to zero and less than or equal to N-1;

When the sum of the first link quality level and the number of link abnormalities in the M Y _j -th link quality levels is less than N-1 of the X _i -th link quality levels that indicate link abnormalities number, and the number of normal links in the M Y _j -th link quality levels is greater than or equal to the number of normal links in the N-1 X _i -th link quality levels, and the first When the service of the active device is damaged, the first active device switches the service to the backup device.

In combination with the first aspect, a first possible implementation of the first aspect is provided, which also includes:

When the sum of the first link quality level and the number of link abnormalities in the M Y _j -th link quality levels is greater than or equal to N-1 of the X _i -th link quality levels, the link When the number of abnormalities, or when the number of normal links in the M Y _j -th link quality levels is less than the normal number of links in the N-1 X _i -th link quality levels, or the When the service of the first active device is not damaged, the first active device does not back up data to the standby device within a preset silent time window.

In combination with the first possible implementation of the first aspect above, a second possible implementation of the first aspect is also provided, including:

When the silent time window expires, the first active device acquires the first link quality level again;

When the first link quality level indicates that the link is normal, the first active device notifies the backup device to switch services back to the first active device.

In combination with the above first aspect, the first possible implementation of the first aspect, or the second possible implementation of the first aspect, a third possible implementation of the first aspect is also provided, which further includes:

The first active device suspends sending backup data to the backup device within a waiting time window starting from sending the request message to each of the other N-1 active devices respectively .

In the second aspect, a hot backup method is provided, including:

The backup device acquires a first link quality level, where the first link quality level is used to indicate the link quality between the first active device and the backup device;

When the first link quality level indicates a link abnormality level, the backup device acquires the backup device and each of the other N-1 active devices except the first active device. N-1 Y _j -th link quality levels between devices, the Y _j -th link quality level is used to represent the link quality between the backup device and the j-th active device, and N is greater than or Equal to 2, j greater than 1 and less than or equal to N;

The backup device sends a request message to each of the other N-1 active devices, and the request message is used to request each of the other N-1 active devices One active device respectively sends the link quality levels between the other N-1 active devices and the first active device;

The backup device respectively receives the M X _i link quality levels between the i-th active device and the first active device sent by the other N-1 active devices, and the X _i link The road quality level is used to indicate the link quality between the i-th active device and the first active device, i is greater than 1 and less than or equal to N, and M is greater than or equal to zero and less than or equal to N-1;

When the sum of the first link quality level and the number of link abnormalities in the N-1 Y _j -th link quality levels is less than the number of link abnormalities in the X _i -th link quality levels number, and the quantity indicating that the link is normal in the N-1 said Y _jth link quality level is greater than or equal to the number indicating that the link is normal in the M said _Xi'th link quality level, and the backup device promotes the connection The access priority after the promotion is used to instruct the user of the first active device to switch to the backup device.

In combination with the second aspect above, a first possible implementation of the second aspect is also provided, including:

When the sum of the first link quality level and the number of N-1 Y _j -th link quality levels indicating link anomalies is greater than or equal to M said X _i -th link quality levels indicating link anomalies , or the number of normal links in the N-1 Y _jth link quality levels is less than the normal number of links in the M X _i link quality levels, and the backup device is in the pre-preparation Suspend receiving the backup data sent by the first master device within the preset silent time window.

In a third aspect, a first primary device is provided, including:

A first acquiring module, configured to acquire a first link quality level, where the first link quality level is used to represent the link quality between the first active device and the backup device;

A judging module, configured to judge whether the first link quality level indicates a link abnormality;

The second acquiring module is configured to acquire N-1 links between the first master device and other N-1 master devices when the judging module judges that the first link quality level indicates that the link is abnormal An X _i -th link quality level, the Xi- _th link quality level is used to represent the link quality between the first active device and the i-th active device, N is greater than or equal to 2, and i is greater than 1 and less than or equal to N;

A sending module, configured to send a request message to each of the other N-1 active devices, and the request message is used to request each of the other N-1 active devices Each active device sends the link quality level between the other N-1 active devices and the backup device;

The receiving module is configured to respectively receive the M Y _j link quality levels between the jth active device and the backup device sent by the M active devices among the other N-1 active devices, the The Y _j -th link quality level is used to represent the link quality between the j-th active device and the backup device, where j is greater than 1 and less than or equal to N, and M is greater than or equal to zero and less than or equal to N-1;

A processing module, configured to be used when the sum of the first link quality level and the number of link abnormalities in the M Y _j -th link quality levels is less than N-1 of the X _i -th link quality levels Represents the number of link abnormalities, and the number of normal links in the M Y _j link quality levels is greater than or equal to the normal number of links in the N- ₁ Xi link quality levels, And when the service of the first master device is damaged, switch the service to the backup device.

In combination with the above third aspect, a first possible implementation manner of the third aspect is provided, the processing module is further configured to, when the first link quality level and the M Y _j -th link quality levels When the sum of the numbers indicating link anomalies is greater than or equal to the number indicating link anomalies in N-1 said X _i link quality levels, or when M said Y _j link quality levels indicate link When the normal number is less than the N-1 number indicating that the link is normal in the X _i -th link quality level, or when the first active device's own service is not damaged, no Back up data to the backup device.

In combination with the first possible implementation of the third aspect above, a second possible implementation of the third aspect is also provided, the first acquiring module is further configured to acquire again after the silent time window expires the first link quality level;

The sending module is further configured to notify the standby device to switch services back to the first active device when the judging module judges that the first link quality level indicates that the link is normal.

In combination with the above third aspect, the first possible implementation manner of the third aspect, or the second possible implementation manner of the third aspect, a third possible implementation manner of the third aspect is also provided, the sending module, It is also used to suspend sending backup data to the backup device within a waiting time window starting from sending the request message to each of the other N-1 master devices.

In the fourth aspect, a backup device is provided, including:

A first obtaining module, configured to obtain a first link quality level, where the first link quality level is used to represent the link quality between the first active device and the backup device;

A judging module, configured to judge whether the first link quality level indicates a link abnormality;

The second acquisition module is configured to acquire the backup device and other N-1 masters except the first master device when the judging module judges that the first link quality level indicates a link abnormality level N-1 Y _j -th link quality levels between each active device in the active device, and the Y _j -th link quality level is used to indicate that between the backup device and the j-th active device The link quality of , N is greater than or equal to 2, j is greater than 1 and less than or equal to N;

A sending module, configured to send a request message to each of the other N-1 active devices, and the request message is used to request each of the other N-1 active devices Each active device respectively sends the link quality level between the other N-1 active devices and the first active device;

A receiving module, configured to respectively receive the M X _i -th link quality levels between the i-th active device and the first active device sent by the other N-1 active devices, and the X _i -th The link quality level is used to indicate the link quality between the i-th active device and the first active device, where i is greater than 1 and less than or equal to N, and M is greater than or equal to zero and less than or equal to N-1;

A processing module, configured to be used when the sum of the first link quality level and the number of abnormal links in the N-1 Y _j -th link quality levels is less than M of the X _i -th link quality levels Indicates the number of link abnormalities, and the number of normal links in the Y _jth link quality level of N-1 is greater than or equal to the number of normal links in the X _ith link quality level of M, and the improvement Access priority, where the elevated access priority is used to instruct a user of the first active device to switch to the standby device.

In combination with the fourth aspect above, a first possible implementation of the fourth aspect is provided, the processing module is further configured to, when the first link quality level and N-1 the Y _j -th link quality The sum of the number of link abnormalities in the levels is greater than or equal to the number of link abnormalities in the X _i -th link quality level, or the link is normal in the N-1 Y _j -th link quality levels The number is less than the number of the M X _i link quality levels indicating that the link is normal, and the reception of the backup data sent by the first master device is suspended within the preset silent time window.

A fifth aspect provides a communication system, including: N master devices and the backup device provided in the fourth aspect above or the first possible implementation manner of the fourth aspect above, where the N master devices Any one master device is the first master device provided in the above third aspect or any possible implementation manner of the above third aspect.

The hot backup method, the first active device, the backup device and the communication system provided by the embodiment of the present invention count the link states between the first active device and other active devices, and count the links between each active device and the other active devices. The link status between the backup devices can accurately determine the cause of the link problem between the first active device and the backup device, so that a more reasonable hot backup process can be made, and it will not be caused by the first active device. Error hot backup processing caused by link packet loss between the backup device and the backup device, saving system resources.

Description of drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description These are some embodiments of the present invention. For those skilled in the art, other drawings can also be obtained according to these drawings on the premise of not paying creative efforts.

Fig. 1A is a schematic diagram of a BRAS dual-machine hot backup;

Fig. 1B is another schematic diagram of BRAS dual-machine hot backup;

Figure 2 is a schematic diagram of a multi-machine hot standby network including N active devices and one backup device;

FIG. 3 is a flowchart of a hot backup method provided by an embodiment of the present invention;

FIG. 4 is a flowchart of a hot backup method provided by an embodiment of the present invention;

FIG. 5 is a flowchart of a hot backup method provided by an embodiment of the present invention;

FIG. 6 is a flowchart of a hot backup method provided by an embodiment of the present invention;

FIG. 7 is a flowchart of a hot backup method provided by an embodiment of the present invention;

FIG. 8 is a flowchart of a hot backup method provided by an embodiment of the present invention;

FIG. 9 is a schematic structural diagram of a first master device provided by an embodiment of the present invention;

FIG. 10 is a schematic structural diagram of a backup device provided by an embodiment of the present invention;

FIG. 11 is a schematic structural diagram of a first master device provided by an embodiment of the present invention;

FIG. 12 is a schematic structural diagram of a backup device provided by an embodiment of the present invention;

FIG. 13 is a schematic structural diagram of a communication system provided by an embodiment of the present invention.

Detailed ways

In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments It is a part of embodiments of the present invention, but not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

Hot backup is a commonly used technology in communication networks. Take BRAS implementing hot backup using VRRP protocol as an example. The basic flow of BRAS dual-system hot backup is shown in Figure 1A and Figure 1B. 1A is a schematic diagram of a BRAS dual-machine hot backup; FIG. 1B is a schematic diagram of another BRAS dual-computer hot backup.

In Fig. 1A and Fig. 1B, terminal 11 communicates with switch 13 through access device 12, switch 13 can communicate with BRAS14 and BRAS 15, BRAS 14 access core network 17 through core router (Core Router, CR) 16, BRAS 15 The core network 17 is accessed through the CR 16 .

Among them, BRAS 14 is the active BRAS, and BRAS 15 is the backup BRAS. BRAS 14 is responsible for terminal 11 access and bearing user services, and BRAS 15 synchronizes database information such as user sessions from BRAS 14, such as user media access control (Media Access Control, MAC), network protocol (Internet Protocol, IP), dynamic host configuration Agreement (Dynamic Host Configuration Protocol, DHCP) lease term, DHCP option 82 (option82), etc. Through the routing policy, the priority of the user route issued by BRAS 14 can be higher than that of BRAS 15. In this way, the downlink traffic from the network side to terminal 11 is preferably forwarded by BRAS 14, as shown in path 18 in FIG. 1A.

When BRAS 14 fails, the failure may be a link failure or a complete machine failure, VRRP and redundant user information (Redundant User Information, RUI) undergo active-standby switchover, BRAS 15 acts as the new active BRAS, and BRAS 14 acts as the new active BRAS. Standby BRAS. BRAS 15 will publish the Internet Protocol (Internet Protocol, IP) route of the high-priority user, and BRAS 14 will cancel the user's IP route or reduce the priority of the route, so that the traffic downlinked from the network side to the terminal 11 is switched to the BRAS 15. On the user access side, the BRAS 15 device will send an Address Resolution Protocol (Address Resolution Protocol, ARP) message to modify the MAC table in the switch 13, and the upstream traffic of the terminal 11 will also be switched to the BRAS 15, as shown in the path 19 in Figure 1B . The above switching protection process will not be perceived by the terminal 11, that is, the terminal 11 does not need to redial to go online. After the fault of BRAS14 recovers, the users who access BRAS15 can switch back to BRAS14.

BRAS uses VRRP protocol to realize multi-device hot backup and has several important modules:

Module 1: VRRP module. Deploy the VRRP protocol between every two BRASs, and elect the determined active and standby roles. After VRRP is deployed, the active BRAS will send VRRP heartbeat packets to the standby BRAS. The packets are Layer 2 multicast packets. Therefore, a Layer 2 channel needs to be established between the active BRAS and the standby BRAS. If both the active BRAS and the standby BRAS are connected to a layer 2 network, such as a switch, there is a layer 2 channel between the active BRAS and the standby BRAS. If the distance between the active BRAS and the standby BRAS is relatively long, or there is no channel through the Layer 2 network between the active BRAS and the standby BRAS, a Layer 2 virtual private network that traverses the MAN needs to be deployed between the active BRAS and the standby BRAS (Layer 2 Virtual Private Network, L2VPN) tunnel. The VRRP module can track the state of the access port, such as available (up) or unavailable (down), to dynamically adjust the priority parameters of the VRRP protocol. At the same time, the VRRP module can be linked with detection technologies such as Bidirectional Forwarding Detection (BFD) to trigger re-election of the active BRAS and standby BRAS, and link services for switching processing.

Module 2: Remote Backup Service (RBS) module. The RBS module is used to transmit backup data between two BRASs. The RBS module is implemented using Transmission Control Protocol (TCP), and provides batch backup and real-time backup mechanisms for services. Batch backup occurs after the TCP connection is successfully established. The real-time backup occurs when the user accesses successfully, the attributes of the user are changed successfully after being online, or the backup cycle of user statistical data arrives.

Module 3: RUI processing module. The RUI processing module is responsible for backing up user information, tracking the status of VRRP, determining the status of the primary device and backup device for accessing users, and controlling the switching of forwarding routes and charging control based on the status.

The VRRP protocol can effectively determine the active/standby relationship between two BRASs, and combined with BFD and other detection technologies, it can ensure that the standby BRAS can quickly sense and switch quickly after the active BRAS fails. Specifically, as shown in Figure 1A and Figure 1B, VRRP heartbeat packets are forwarded through the switches connected to the two BRASs. At the same time, BFD tracks the link on the network side of the active BRAS. The standby BRAS can quickly detect faults and link faults with the CR device on the network side. In a relatively complex multi-device hot standby network, rapid detection and switching of faults is also achieved through the technology of VRRP combined with BFD detection.

The VRRP combined with BFD detection technology in the current multi-device hot backup technology can only deal with link faults, such as link interruption and complete machine failure. For faults such as link packet loss, erroneous judgments may be made. Fig. 2 is a schematic diagram of a multi-machine hot standby network including N active devices and one backup device. In the network shown in Figure 2, take the multi-machine hot standby network including 3 active BRAS and one standby BRAS as an example, the three active BRAS are BRAS 21, BRAS 22 and BRAS 23, and the standby BRAS is BRAS 24. BRAS 21, BRAS 22, and BRAS 23 respectively establish VRRP and RBS with BRAS 24 to form three device pairs, and there is an active-standby relationship between two BRASs included in each of the three device pairs. Any BRAS in each device pair can judge the status of the peer end according to VRRP and BFD, and determine whether the RBS is established or not according to whether the route is reachable. The detection technology can only take effect when the link fails or the whole machine fails, and any BRAS in each device pair cannot accurately judge the status of the opposite end when the intermediate network suffers serious packet loss.

Specifically, if the RBS repeatedly disconnects or BFD flaps, the active BRAS still considers itself to be the VRRP master and maintains the role of the VRRP master; The VRRP state keeps fluctuating. However, based on this information, neither the active BRAS nor the standby BRAS can determine whether there is a problem with the peer end or the network where the peer end is located. The backup BRAS cannot accurately judge whether there is a problem with itself, unless there are non-RUI users in the backup BRAS, and a large number of non-RUI users are disconnected, the backup BRAS can learn that its own network is faulty.

In the following three scenarios, VRRP cannot handle it correctly. (1) Packets are lost on the network where the active BRAS is located, causing RBS to flap. The active BRAS should be switched to the standby BRAS for RUI switchover, but in fact, because only packet loss, not link unavailability, VRRP switchover will not be triggered. (2) The network where the backup BRAS is located fails and loses packets, causing RBS to fluctuate. The active and standby state of the RUI remains unchanged, and it is reasonable not to back up data to the standby BRAS, but in fact, as long as the RBS recovers, the backup is started, resulting in waste of resources. (3) The network failure between the active BRAS and the standby BRAS causes RBS to fluctuate. At this time, both parties cannot guarantee that the links of all users are reachable, the links between some users and the active BRAS are reachable, and the links between some users and the backup BRAS are reachable. At this time, the RUI should meet the services of all users. But in fact, VRRP has been fluctuating, unable to stabilize the active BRAS and the standby BRAS, and the active BRAS and the standby BRAS are always backing up data.

It can be seen that VRRP can only make the correct backup selection when the whole machine fails or the link is interrupted. waste.

In a multi-device hot backup scenario, the link status between multiple active devices and a backup device may have problems to varying degrees, but only through the detection of the two devices themselves, it may not be possible to determine which device is the cause. cause link problems. Therefore, in order to make a correct backup selection in a multi-device hot backup scenario, it is first necessary to correctly determine the device causing the link problem. In the N:1 multi-device hot backup scenario, when the active BRAS finds that the packet loss rate between itself and other active BRASs is small, but the packet loss rate between itself and the standby BRAS is high, it means that the standby BRAS is causing a link problem. device of. When all active BRASs and standby BRASs find that the packet loss rate of themselves and a certain active BRAS is very high, it means that the certain active BRAS is the device causing the link problem. Therefore, a performance detection protocol can be configured between all BRASs. When VRRP flapping or RBS flapping occurs, the active BRAS finds that the packet loss rate between itself and other active BRASs is very small, but the packet loss rate between itself and the standby BRAS is very high. If the value is large, it indicates that the standby BRAS is the device causing the link failure. When all active BRASs and standby BRASs find that they have a high packet loss rate with a certain active BRAS, it indicates that the certain active BRAS is a device causing a link problem.

Based on the above idea, the embodiment of the present invention provides a hot backup method, which can accurately determine the device causing the link failure, and then reasonably perform hot backup processing.

Fig. 3 is a flow chart of the hot backup method provided by the embodiment of the present invention. The hot backup method provided by this embodiment is applied to a multi-machine hot backup system in which the number ratio of the active device to the backup device is N:1, where N≥2 . The hot backup method provided by the embodiment of the present invention will be described in detail below with reference to FIG. 3 .

S301. The first active device acquires a first link quality level, where the first link quality level is used to represent the link quality between the first active device and the backup device.

Specifically, the hot backup method provided in this embodiment is applied to a multi-machine hot backup system including N master devices and one backup device, where N≥2.

The execution subject of the hot backup method provided in this embodiment is any active device, which is called the first active device. First, the first active device needs to obtain a first link quality level from the first active device to the backup device. The first link quality level is used to indicate the link quality between the first active device and the backup device. The first link quality level may be represented by various parameters, such as the packet loss rate and data throughput rate between the first active device and the backup device. The first active device can learn whether the link between the first active device and the standby device is normal through the first link quality level. The first link quality level can be divided into several levels, such as normal, abnormal, unavailable, etc., where unavailable is a state when the first active device cannot obtain the first link quality level. Alternatively, the first link quality level may be a numerical value or a proportional value, and corresponding value ranges may be set for normal, abnormal, and unavailable states, which will not be described one by one here.

S302. When the first link quality level indicates that the link is abnormal, the first active device obtains the N-1 X _i link quality levels between the first active device and other N-1 active devices, The X _{i th} link quality level is used to indicate the link quality between the first active device and the i th active device, N is greater than or equal to 2, and i is greater than 1 and less than or equal to N.

Specifically, after the first active device obtains the first link quality level, if the first link quality level indicates that the link is normal, it means that the link status between the first active device and the backup device is normal, It can be processed according to the normal hot backup process.

When the first link quality level indicates that the link is abnormal, it means that there is a problem with the link between the first active device and the backup device, and the first active device needs to determine whether the link problem is caused by the backup device failure or the second A link problem caused by the failure of the active device.

For example, the first active device first obtains the N-1 X _i link quality levels between the first active device and other N-1 active devices, and the X _i link quality level is used to represent the The link quality between an active device and the i-th active device, where N is greater than or equal to 2, and i is greater than 1 and less than or equal to N. Since the system includes N−1 active devices in addition to the first active device, the first active device will obtain N−1 X _{i th} link quality levels. If the W link quality level of the Xi th link quality level in the N-1 _{xi th} link quality levels is abnormal, and the S th link quality level of the Xi _th link is normal, then the sum of W and S is N-1.

S303, the first active device sends a request message to each of the other N-1 active devices respectively, and the request message is used to request each of the other N-1 active devices The active device sends the link quality levels between the other N-1 active devices and the backup device respectively.

Specifically, the first active device also needs to know the link quality level between other N-1 active devices and backup devices, that is, to know the link status between other N-1 active devices and backup devices . The first active device sends request packets to each of the other N-1 active devices respectively.

It should be noted that the cause of abnormal link status between the first active device and the backup device may be caused by the failure of the first active device. The link between the first active device and other N-1 active devices The link of may be abnormal due to the failure of the first active device. Therefore, the request message sent by the first active device may not be received by other N-1 devices.

S304. The first active device respectively receives the M Y j-th link quality levels between the _j -th active device and the standby device sent by M active devices among the other N-1 active devices, and the Y _j -th link quality level is The link quality level is used to indicate the link quality between the jth active device and the standby device, where j is greater than 1 and less than or equal to N, and M is greater than or equal to zero and less than or equal to N-1.

Specifically, after the first master device sends request messages to other N-1 master devices, it will receive the Y _j -th link quality level sent by the j-th master device, where the Y _j -th link The quality level is used to indicate the link quality between the jth active device and the backup device. Since there may be problems with the link between the first active device and the other N-1 active devices, the first active device may not receive the message sent by the jth active device among the other N-1 active devices. The Y _jth link quality, that is, the N-1 link quality level may not be received, and the Y _j link quality level in the N-1 link quality levels is used to represent the jth active device and the backup Link quality between devices. If the first active device has received M link quality levels in total, M is less than or equal to N-1. Suppose P link quality levels are abnormal, and Q link quality levels are normal, then the sum of P and Q is less than or equal to N-1.

S305, when the sum of the first link quality level and the number of link abnormalities in the M Y _j -th link quality levels is less than the number of link abnormalities in the N-1 X _i -th link quality levels, and M When the number of normal links in the Y _j -th link quality level is greater than or equal to the number of normal links in the N-1 X _i -th link quality level, and the service of the first active device is damaged, the A master device switches services to the backup device.

For example, the M Y _j -th link quality levels represent the link quality level between each active device and the backup device among the M active devices. If M is greater than ₃ , then the M Y _jth link quality levels are _Y2th link quality level, Y3rd link quality level... _YM+1th link quality level. Among them, the Y _2nd link quality level represents the link status from the second master device to the backup device; the Y _3rd link quality level represents the link status from the third master device to the backup device; the Y _M+1 link The link quality level indicates the link status from the M+1th master device to the backup device. When M is other values, the specific content of the M Y _j -th link quality level will not be illustrated here.

For example, the N−1 X _{i th} link quality levels indicate the link quality level between each active device among the N−1 active devices and the first active device. If N is greater than 3, then the N−1 X _i -th link quality levels are respectively the X2 _- th link quality level, the _X3 -th link quality level...the _XN -th link quality level. Wherein, the _X2th link quality level indicates the link status between the first active device and the second active device. The _X3rd link quality level indicates a link state between the first active device and the third active device. The X _Nth link quality level indicates a link state between the Nth active device and the first active device. When N is other values, the specific content of the N−1 X _{i th} link quality level will not be illustrated here.

Specifically, after the first active device obtains the first link quality level, the N-1 X _i link quality levels, and the M Y _j link quality levels, it can determine whether the first active device and The link problem between the backup devices is due to whom. If the sum of the first link quality level and the number of link abnormalities in the M Y _j -th link quality levels is less than the number of link abnormalities in the N-1 X _i -th link quality levels, that is, P and The sum of 1 is less than W. The sum of P and 1 is because the first active device also believes that the link quality level between the backup device and the backup device indicates that the link is abnormal. The link quality level of the device indicates the number of link abnormalities; and the number of link normal levels in the M Y _j -th link quality level is greater than or equal to the N-1 X _i -th link quality level indicating that the link is normal The number of levels, that is, Q≥S, means that the link quality level of the backup device is considered to indicate that the link is normal, which is greater than or equal to the number of the link quality level of the first active device that is considered to be normal. That is to say, the link quality level of the standby device indicates that the link is less likely to be abnormal than the link quality level of the first active device indicates that the link is abnormal, and the link quality level of the standby device indicates that the link is normal The possibility of being greater than or equal to the link quality level of the first active device indicates the possibility that the link is normal. At the same time, when the first active device finds that its own service is damaged, the first active device will determine that the link problem between the first active device and the backup device is caused by its own fault, so the first active device Use the equipment to switch the service to the standby equipment.

In addition to judging the quality level of each link, the first active device also needs to judge that its own business is damaged before making a business switchover. This is because in the N:1 multi-device hot backup system, the backup device is always in the standby state , only when the first master device does have a problem, it will switch to the backup device, and the backup device is not suitable for long-term role as the master device in the network. If the first active device is not damaged in its own business, it can still provide network services for users, but the hot backup of data may be affected due to possible problems in links with other active devices and backup devices. In order to ensure that the users in the system are not frequently switched and ensure the stability of the system, only when the service of the first active device is damaged and the link problem between the first active device and the backup device is caused by the first The service of the first active device is switched to the backup device only when the fault of the active device is caused. The first active device can determine whether its own business is damaged by detecting whether the connected user is disconnected, whether there is no data transmission for a long time, or whether the data transmission with other connected users is normal. In other words, the first active device determines that its own service is damaged when it detects that the connected user is offline, there is no data transmission for a long time, or the data transmission with other connected users is abnormal.

The first active device can use a common service switching method to complete the service switching to the standby device, which will not be repeated here.

In the hot backup method provided in this embodiment, the link status between the first active device and other active devices is counted, and the link status between each active device and the backup device is counted, so that it can accurately Determine the cause of the link problem between the first active device and the backup device, so that a more reasonable hot backup process can be made without causing packet loss on the link between the first active device and the backup device. Wrong hot backup processing saves system resources.

Further, in the embodiment shown in FIG. 3 , after the first active device switches services to the backup device, the first active device will switch to the standby device state, and the backup device will switch to the active device state. At this time, the first active device may also start a preset silent time window, and within the silent time window, the first active device will not switch back to the active device state. This avoids frequent switching between active and standby devices.

Further, when the silent time window expires, the first active device will obtain the first link quality level with the backup device again, and if the first link quality level is normal, the first active device will notify the backup device The device switches services back to the active device. The time of the silent time window can be set according to empirical values, and the silent time window can generally be set as the average fault recovery time of the system.

Further, in the embodiment shown in FIG. 3 , the triggering method for the first active device to switch services to the backup device may include: the first active device lowers the access priority or initializes, and triggers access to the first active device The user device of the user is switched to the backup device. That is to say, when the first active device determines that its own failure causes a link problem with the backup device, the first active device can actively lower its own access priority, so that the user equipment connected to the first active device switches to Backup equipment to complete service switching. Alternatively, the first active device may also perform initialization processing, which may also switch the user equipment connected to the first active device to the backup device, thereby completing service switching. The principle of lowering the access priority of the first active device is that the access priority needs to be lowered to be lower than that of the standby device, so that the user who originally accessed the first active device can switch to the standby device.

Further, after the first active device lowers the access priority or initializes, the hot backup method provided in this embodiment further includes: the first active device sends the first active device lowering the access priority or initializing news. In this way, the backup device can also learn the information about the priority reduction or initialization of the first active device, so as to prepare for the access of the user equipment that originally accessed the first active device. It should be noted that although the first active device sent a message that the first active device lowered the access priority or initialized to the backup device, due to a problem with the link between the first active device and the backup device, the backup The device may not receive the message, but because the first active device has lowered the access priority, even if the backup device cannot receive the information sent by the active device, the handover of the user equipment can still be completed.

FIG. 4 is a flowchart of a hot backup method provided by an embodiment of the present invention. The hot backup method provided in this embodiment, on the basis of the method provided in the embodiment shown in Figure 3, may also include:

S401, when the sum of the first link quality level and the number of link abnormalities in the M Y _j -th link quality levels is greater than or equal to the number of link abnormalities in the N-1 X _i -th link quality levels , or when the number of normal links in the M Y _j -th link quality levels is less than the normal number of links in the N-1 X _i -th link quality levels, or the service of the first active device is not affected In the event of a loss, the first active device does not back up data to the backup device within the preset silent time window.

For example, after the first active device acquires the first link quality level, the N-1 X _i -th link quality levels, and the M Y- _th link quality levels, the following judgments may also be made. If the sum of the first link quality level and the number of abnormal link levels in the N-1 X _i link quality levels is greater than or equal to the number of link abnormal levels in the M Y _j link quality levels, that is The sum of P and 1 is greater than or equal to W. P+1 is because the first active device also thinks that the link quality level between the backup device and the backup device indicates that the link is abnormal), that is, the link quality level of the backup device indicates that the number of link abnormalities is greater than or equal to that of the first active device. The link quality level of the device indicates the number of abnormal links, so the first active device can consider that the link status between the backup device and other N-1 active devices is worse than that between the first active device and other N-1 active devices. If the link status between 1 master device is known, then the first master device can determine that the link problem between the first master device and the backup device is caused by a failure of the backup device.

For example, if the number of N-1 X _i -th link quality levels indicating the normal level of the link is less than the number of M Y- _th link quality levels indicating the normal level of the link, that is, Q<S, it is considered a standby device The link quality level of the link indicates that the number of normal links is less than the link quality level of the first active device indicates that the number of links is normal, then the first active device can be considered as a backup device to other N-1 active devices is worse than the link status of the first master device to other N-1 master devices, then the first master device can also determine that the link problem between the first master device and the backup device is due to the backup device caused by a malfunction. That is to say, the voting opinion of each device is that the link quality level of the backup device indicates that the possibility of link abnormality is greater than or equal to the link quality level of the first active device. The link quality level indicates that the link is normal than the link quality level of the first active device indicates that the link is normal.

Then, when the first active device determines that the link problem between the first active device and the backup device is caused by the failure of the backup device, if the first active device continues to send backup data to the backup device, it may be due to link loss. Backup failures or repeated backups due to reasons such as data packets, resulting in waste of resources. Then at this time, the first active device will maintain the state of the active device without performing hot backup, that is, not backing up data to the standby device.

Further, the first active device may also start a silent time window, and within the silent time window, the first active device does not perform hot backup processing, that is, no longer backs up data to the standby device. After the silent time window expires, the first master device may perform the hot backup processing in the embodiment shown in FIG. 3 again. The time of the silent time window can be set according to empirical values, and the silent time window can generally be set as the average fault recovery time of the system.

Further, in the embodiment shown in FIG. 4 , the first active device determines that the link status between the backup device and the other N-1 active devices is worse than that between the first active device and the other N-1 active devices. After the link state, it also includes: the first active device sends a silent time window to the backup device, and the silent time window is used to prevent the backup device from processing received backup data within the silent time window.

Specifically, when the first master device determines that the link problem between the first master device and the backup device is caused by a failure of the backup device, the first master device determines to maintain the state of the master device, and the first master device initiates a silent After the time window, the first active device may also send the silent time window to the backup device. After receiving the silent time window, the backup device will determine that the first active device will not perform hot backup within the time window. At this time, even if the backup device receives the backup data, it will not process it.

Further, the method provided in the embodiment corresponding to FIG. 3 or the method provided in the embodiment corresponding to FIG. 4 further includes: the first master device sends the slave to each of the other N-1 master devices respectively. During the waiting time window at the beginning of the request message, the sending of backup data to the backup device is suspended.

Specifically, after the first active device sends a request message to other N-1 active devices, the first active device is not yet sure whether the link problem between the first active device and the backup device is due to Which device is faulty, the first active device may also suspend sending backup data to the backup device within a preset waiting time window. In this way, resource waste caused by the first active device continuing to back up data to the backup device due to a failure of the backup device is avoided.

FIG. 5 is a flowchart of a hot backup method provided by an embodiment of the present invention. The hot backup method provided in this embodiment is applied to a multi-machine hot backup system in which the ratio of active equipment to standby equipment is N:1, where N is greater than or equal to two. The method provided by the embodiment of the present invention will be described below with reference to FIG. 5 .

S501. The backup device acquires a first link quality level, where the first link quality level is used to represent the link quality between the first active device and the backup device.

The execution subject of the hot backup method provided in this embodiment is the backup device. For the meaning of the first link quality level, refer to the corresponding content in the embodiment corresponding to FIG. 3 , which will not be repeated here.

S502. When the first link quality level indicates the link abnormality level, the backup device obtains the N- 1 Y _jth link quality level, the Y _jth link quality level is used to indicate the link quality between the backup device and the jth active device, N is greater than or equal to 2, and j is greater than 1 and less than or equal to N.

Specifically, after the backup device obtains the first link quality level, if the first link quality level is , indicating that the link is normal, it means that the link status between the first active device and the backup device is normal, and the Normal hot backup process for processing.

When the first link quality level indicates that the link is abnormal, it means that there is a problem with the link between the first active device and the backup device, and the backup device needs to determine whether the link problem is caused by the backup device failure or the first active Link problems caused by device failures. The backup device obtains N-1 link quality levels between the backup device and other N-1 active devices, and the Y _j -th link quality level in the N-1 link quality levels is used to represent the backup device and the first link quality level. j Link quality between active devices, N is greater than or equal to 2, and j is greater than 1 and less than or equal to N. The meanings of the N-1 Y _j -th link quality levels are the same as those in the embodiment corresponding to FIG. 3 , and will not be repeated here. The standby device will obtain N-1 quality levels of Y _j -th links, wherein, among the N-1 quality levels of Y _j -th links, p Y _j - _th link quality levels are abnormal, and q Y-th link quality levels are abnormal. The road quality level is normal, p+q=N-1.

S503. The backup device sends a request message to each of the other N-1 active devices, and the request message is used to request each of the other N-1 active devices to Send the link quality levels between the other N-1 active devices and the first active device.

Specifically, the standby device learns the link quality level between the other N-1 active devices and the first active device by sending a request message.

It should be noted that the abnormal link status between the first active device and the backup device may be caused by a failure of the backup device, so the request message sent by the backup device may not be received by all other N-1 devices. received.

S504. The backup device respectively receives the M X _i -th link quality levels between the i-th active device and the first device sent by other N-1 active devices, and the X _i -th link quality level is used to represent the i-th link quality level The link quality between the active device and the first active device, i is greater than 1 and less than or equal to N, and M is greater than or equal to zero and less than or equal to N-1.

Specifically, the meanings of the M X _i -th link quality levels are the same as those in the embodiment corresponding to FIG. 3 , and will not be repeated here. After the standby device sends the request message, it will receive the link quality levels between each active device and the first active device sent by other active devices. Since the backup device may have a link problem with the other N-1 active devices, the backup device may not receive the N-1 link quality levels sent by the other N-1 active devices. The device uses the device to receive a total of M X _i -th link quality levels, and M is less than or equal to N-1. It is assumed that the quality level of the wth Xi link is abnormal, and the quality level of the s Xi link is normal, _w + _s≤N -1.

S505, when the sum of the first link quality level and the number of link abnormalities in the N-1 Y _jth link quality levels is less than the number of link abnormalities in the M _Xith link quality levels, and N - The number of normal links in the Y _jth link quality level is greater than or equal to the M number of normal links in the X _i link quality level, the backup device increases the access priority, and the upgraded access priority Level is used to instruct the user of the first master device to switch to the backup device.

Specifically, after the backup device acquires the first link quality level, the N-1 Y _j -th link quality levels, and the M X- _th link quality levels, it can determine the difference between the first active device and the backup device. The link problem between them is caused by whom. If the sum of the first link quality level and the number of link abnormalities in the N-1 Y _j -th link quality levels is less than the number of link abnormalities in the M X _i -th link quality levels, that is, p+1 <w. p+1 is because the backup device also thinks that the link quality level between the backup device and the primary active device indicates that the link is abnormal. The link quality level of the device indicates the number of link abnormalities; and the number of normal links in the N-1th link quality level of Y _j is greater than or equal to the number of normal links in the Mth link quality level of X _i The number, that is, q≥s, means that the number of link quality levels of the backup device that indicate a normal link is greater than or equal to the number that the link quality level of the first active device indicates that a link is normal. That is to say, the voting opinion of each device is that the link quality level of the standby device indicates that the possibility of link abnormality is lower than that of the first active device, and the link quality level of the standby device indicates the possibility of link abnormality. The quality level indicates that the possibility of the link being normal is greater than or equal to the link quality level of the first active device, indicating the possibility of the link being normal. Then the backup device will determine that the link problem between the first master device and the backup device is caused by the failure of the first master device, so the backup device will determine that the service of the first master device needs to be switched to the backup device.

For example, if the backup device wants to switch the service of the first active device to the backup device, it can be realized by increasing the access priority. When the backup device raises the access priority, the user equipment originally connected to the first active device can be switched to the backup device. The principle for the backup device to increase the access priority is that the access priority needs to be increased to be higher than that of the first active device, so that the user who originally accessed the first active device can switch to the backup device.

In the hot backup method provided in this embodiment, the link status between the backup device and other active devices is counted, and the link status between each active device and the first active device is counted, so that it can accurately Determine the cause of the link problem between the first active device and the backup device, so that a more reasonable hot backup process can be made without causing packet loss on the link between the first active device and the backup device. Wrong hot backup processing saves system resources.

Fig. 6 is a flow chart of the hot backup method provided by the embodiment of the present invention. The hot backup method provided by this embodiment, on the basis of the method provided by the embodiment corresponding to Fig. 5, further includes:

S601, when the sum of the first link quality level and the number of link abnormalities in the N-1 Y _jth link quality levels is greater than or equal to the number of link abnormalities in the M _Xith link quality levels, or The number of normal links in the N-1 Y _jth link quality level is less than the normal number of links in the M X _ith link quality level, and the backup device suspends receiving the first link within the preset silent time window. Backup data sent by the active device.

Specifically, after the standby device acquires the first link quality level, the N-1 Y _j link quality levels, and the M X _i -th link quality levels, the following judgments may also be made. If the sum of the first link quality level and the number of link abnormalities in the N-1 Y _j link quality levels is greater than or equal to the number of link abnormalities in the M X _i link quality levels, that is, p+1 ≥w (p+1 is because the backup device also thinks that the link quality level between the backup device and the first active device indicates link abnormality), that is, the link quality level of the backup device indicates that the number of link exceptions is greater than or equal to If the link quality level of the first active device is considered to be abnormal, then the backup device can consider that the link status between the backup device and other N-1 active devices is worse than that between the first active device and other N-1 active devices. If there is no link status between 1 active device, then the backup device can determine that the link problem between the first active device and the backup device is caused by the failure of the backup device. If the number of normal links in the N-1 Y _j link quality levels is less than the normal number of links in the M X _i -th link quality levels, that is, q<s, the link quality level of the backup device is considered If the number of normal links is less than the number of normal links indicated by the link quality level of the first active device, then the backup device can consider that the link status between the backup device and other N-1 active devices is worse than The link status between the first active device and other N-1 active devices, then the backup device can also determine that the link problem between the first active device and the backup device is caused by the failure of the backup device. That is to say, the link quality level of the standby device indicates that the possibility of link abnormality is greater than or equal to the link quality level of the first active device indicates the possibility of link abnormality, or the link quality level of the standby device indicates that the link A possibility that the link is normal is lower than the link quality level of the first active device indicates the possibility that the link is normal.

Then, when the backup device determines that the link problem between the first active device and the backup device is caused by the failure of the backup device, if the backup device continues to receive the backup data sent by the first active device, it may be due to link packet loss, etc. Reasons cause backup failure or repeated backup, resulting in waste of resources. Then, at this time, the standby device suspends receiving the backup data sent by the first active device. Further, the standby device may also start a silent time window, and within the silent time window, the standby device suspends receiving the backup data sent by the first active device. After the silent time window expires, the backup device can perform the hot backup processing in the embodiment shown in FIG. 5 again. The time of the silent time window can be set according to empirical values, and the silent time window can generally be set as the average fault recovery time of the system.

Further, in the embodiment shown in FIG. 5 or FIG. 6 , the backup device judges whether to switch services between the first master device and the backup device according to the link status between itself and each master device. However, there is another possibility that the first active device judges whether to start service switching, and after the first active device makes the judgment, if the first active device determines that the link problem between the first active device and the backup device is If it is determined not to switch services due to the backup device, the first active device can send the silent time window to the backup device; when the backup device receives the silent time window sent by the first active device, it can know the first The active device has determined not to perform service switching, and it is determined that the link problem between the first active device and the backup device is caused by a failure of the backup device, then the backup device does not process the received backup data within the silent time window. That is to say, after the first active device determines that there is no service switching, it can send the silent time window to the backup device, then the backup device will not need to make a judgment, but will not process the received backup device within this time window. data. When the silent time window expires, the backup device may perform processing according to the method of the embodiment shown in FIG. 5 or FIG. 6 .

In addition, when the first active device judges whether to start service switching, and after the first active device makes the judgment, if the first active device determines that the link problem between the first active device and the backup device is caused by the first active device If it is determined to perform service switching due to the failure of the active device, the first active device may send a message that the first active device lowers the access priority or initializes to the backup device. When the standby device receives the message that the first active device lowers the access priority or initializes the message sent by the first active device, it can know that the first active device has determined that it needs to perform service switching, and then the standby device can accept the first active device. A service switching of the master device. That is to say, after the first active device determines to trigger service switching, it can send a message to the backup device that the first active device lowers the access priority or initializes, then the backup device does not need to make a judgment, but can Directly accept the service switching of the first master device.

Further, in the hot backup method provided by the embodiments shown in FIG. 3 to FIG. 6 , the active device and the backup device are BRASs, and the first active device is hot-backed up to the backup device through VRRP. Taking BRAS hot backup through VRRP as an example, the hot backup method provided by the embodiment of the present invention will be further described below.

In order to apply the hot backup method provided by this embodiment under the existing system architecture shown in Figure 1A and Figure 1B, it is necessary to deploy additional functional modules in the BRAS-based system, including: adding a deployment performance detection module between BRAS (All BRASs are deployed in pairs, including between every two active BRASs, and between each active BRAS and the corresponding standby BRAS). Add a module "inter-BRAS link quality detection" module in the system, set a detection period D, such as D=10 seconds, collect the packet loss detection (mainly Throughput rate index) results, the detection results are converted into a drop packet ratio (DropPacketRate) between 0-100, the larger the value, the more serious the packet loss.

Add a module "BRAS terminal disconnection rate detection" module inside the system, set a statistical cycle P, for example P=60 seconds; set a detection cycle E, for example E=10 seconds. This module calculates according to the period of E, and the average offline rate (DownRate) value per second due to detection timeout in the period of P. The system sets a maximum offline rate (MaxDownRate), for example, MaxDownRate=200 sessions (Sesson)/second. Set the offline health index (CientDownHealthTarget), CientDownHealthTarget=DownRate*100/MaxDownRete; the value range is 0-100, and the larger the value, the more serious the offline situation is.

Add a module "RBS flash detection" module inside the system, set a statistical cycle Q, such as Q=1800 seconds (30 minutes); set a detection cycle F, such as F=10 seconds. This module calculates according to the cycle of F, the average RBS shaking frequency (RbsShakeRate) per fixed time (for example, every 300 seconds) in the Q cycle, RBS changes from offline (Down) to online (Up), or from online (Up) to It counts as a shock for going offline (Down). The value range of RbsShakeRate is, for example, 0-25 (300/12). Set a maximum shaking frequency (MaxRbsShakeRate), for example, MaxRbsShakeRate=300 (calculated on the basis of 1 shaking per second). Convert an RBS health index (RBSHealthTarget) RBSHealthTarget=RbsShakeRate*100/MaxRbsShakeRate, the value is between 0-100, the larger the value, the lower the RBS health index.

Moreover, a new device health advertising protocol (Device Health Advertising Protocol, DHAP) is defined, and the content of the protocol is shown in Table 1.

Table 1 DHAP protocol

In Table 1, the version number (Version), role (Role), message type (Message Type), and notification cycle (Interval) fields each occupy 1 byte, and the health indicator (HealthTarget) and reserved fields (Reserved) each occupy 2 bytes. bytes, and the System ID field occupies 4 bytes.

The Version field is the version number of the protocol. For example, if the current version is the first version, the value is 1. The Role field is the role of the device, for example, the value 1 is the active device, and the value 2 is the backup device; the role is set according to whether the device is currently the active device or the backup device in the system. The MessageType field is a message type: for example, a value of 1 is a Health Advertisement, and of course other types can also be used. Interval is the notification period, the unit is second, the default value is 1, and the maximum range is 1 second to 255 seconds. Reserved is a reserved field. The SystemID field is the system ID, and an IP address of the device is used to uniquely identify the device. The IP address, for example, can select the source IP address of the RBS communication. The HealthTarget field is the health index of the device, and the value is DropPacketRate+CientDownHealthTarget+RBSHealthTarget; the smaller the value, the healthier it is, and the value range is 0-300.

The transport layer of the DHAP protocol may be, for example, a User Datagram Protocol (User Datagram Protocol, UDP) or a Transmission Control Protocol (Transmission Control Protocol, TCP), which is implemented, for example, according to UDP in the embodiments of the present invention.

It is also necessary to design and detect the detection parameters of the peer device for the DHAP protocol, including the advertisement interval (Advertisement Interval), the adjustment time (Skew_Time), the peer offline detection interval (PeerDownInterval), etc. The notification period is the period for the DHAP protocol message to be notified to the outside, and the unit is second. For example, the default is 1 second, and the configuration range can be from 1 second to 255 seconds. The adjustment time is the adjustment time of the peer offline detection period, and the unit is second, and its value may be ((256-HealthTarget)/256), for example. The peer offline detection period is the offline detection period of the peer device, in seconds, and its value may be (3×Advertisement_Interval)+Skew_time), for example.

Two states can be defined for the DHAP protocol, which are the Initialize state and the Running state, wherein, in the Initializing state, each device configures neighbor information and then switches to the Running state. In the running state, each device receives an Advertisement message from a neighbor device, and responds to an Advertisement message in real time, that is, it regards the Advertisement message from the other party as a request, and The health index (HealthTarget) is recorded in the database to periodically monitor the state of the neighboring device in the database. If the notification from the neighboring device cannot be received within the time limit, the neighbor is regarded as unavailable, and the peer health index (PeerHealthTarget) parameter is set to the maximum value. 65535.

Table 2 shows the definition of neighbor device health information database (HealthTargetDB) information collected by each device.

Table 2 HealthTargetDB information

The health information of the device itself is also recorded in the database.

Define the variable self-system identification (MySystemId) to record the system identification (Identity, ID) of the device. Define the variable number of all devices (TotalDeviceNumber), record the number of all devices, including this device. Define health threshold: normal (Normal)<=50, 50<abnormal (Abnormal)<=100, unavailable (Unaviailable)>100. Define the offline threshold of terminal equipment: normal (Noraml)<=100; abnormal (Abnormal)>100. Define RBS flicker threshold, normal (Noraml) = 0, abnormal (Abnormal) > 2 (more than 2 times per minute).

After the above-mentioned modules and protocols are set in the system, hot backup processing can be performed according to the method shown in FIG. 7 or FIG. 8 , wherein FIG. 7 is the processing flow of the active BRAS, and FIG. 8 is the processing flow of the standby BRAS.

FIG. 7 is a flow chart of the hot backup method provided by the embodiment of the present invention. This embodiment is applied to a multi-machine hot backup scenario including N active BRASs and one backup BRAS through VRRP hot backup. The execution subject of this embodiment is Any active BRAS is referred to as the first active BRAS. As shown in Figure 7, the hot backup method of the present embodiment includes:

S701. The first active BRAS detects the HealthTarget of the standby BRAS. When the first active BRAS finds that the HealthTarget of the standby BRAS has changed to an abnormal (Abnormal) level, continue to execute S702. Otherwise, the first active BRAS will be processed according to the existing hot backup process. The HealthTarget of the standby BRAS detected by the first active BRAS is the first link quality level.

S702. The first active BRAS compares the HealthTargets of other N-1 active BRASs, and obtains that the IP FPM indicators of the W equipment and its own are degraded, and the S equipment and its own IP FPM indicators are normal; W+S=N-1 . The HealthTarget of the other N-1 active BRASs detected by the first active BRAS is the X _{i th} link quality level.

S703, the first active BRAS sends a first request message to other N-1 active BRASs, requesting the other party to respond to whether the standby BRAS is degraded; and notifies itself of the health degree of the standby BRAS, and waits for a response message A time window Tw, within which the data backup to the standby BRAS is suspended.

S704, within a waiting time window Tw, the first active BRAS receives L copies of response messages. Among them, P replies that the health degree of the peer end to the standby BRAS has deteriorated, and Q replies say that the health degree of the peer end to the standby BRAS has not deteriorated, and P+Q=L, L<=N-1; vote by show of hands. The response message received by the first active BRAS is the Y _jth link quality level.

S705, if (P+1>=W or Q<S). The standby BRAS is more unstable than the first active BRAS to other active BRASs. Vote for the failure of the standby BRAS: VRRP does not switch over silently, and at the same time starts the silent time window Ts, and suspends backup data to the standby BRAS within the time window Ts; The backup BRAS sends a message to inform the VRRP of silence (this message may not be received by the slave).

S706, if (P+1<W and Q>=S) and at the same time it finds that its own health is abnormal (Abnormal) (service damage, such as BRAS RUI user detection disconnection, ARP RUI detection unreachable, etc.). The backup BRAS is more stable than the first active BRAS to other active BRASs, and votes for its own failure: immediately lower the priority of its own VRRP or initialize VRRP, trigger VRRP/RUI switchover; send a VRRP heartbeat message to the standby BRAS, which carries the current Priority or VRRP initialization message (this message may not be received by the standby BRAS), and start the silent time window Ts at the same time.

S707, in cases other than S705 and S706, the first active BRAS maintains the state of the VRRP active device, starts the silent time window Ts, and does not back up data to the opposite end.

Among them, in S705 to S707, the first active BRAS is within the silent time window Ts, and the following judgment is made, if the throughput rate of the first active BRAS and the standby BRAS returns to normal: then the silent state is released, and VRRP communicates with the opposite end in the current state Negotiation, keep the state of the active device unchanged or switch back to the state of the active device after a delay. Otherwise, if the first silent time window Ts times out, repeat S701.

S708. Restore the RUI user backup.

Figure 8 is a flow chart of the hot backup method provided by the embodiment of the present invention. This embodiment is applied to a multi-machine hot backup scenario including N active BRASs and one backup BRAS through VRRP hot backup. The execution subject of this embodiment is Standby BRAS. As shown in Figure 8, the hot backup method of this embodiment includes:

S801, the standby BRAS detects whether the RBS health index (RBSHealthTarget) of the first active BRAS exceeds a threshold, and when the RBS health index (RBSHealthTarget) of the standby BRAS and the first active BRAS exceeds the threshold, continue to execute S802. Otherwise, it will be processed according to the existing hot backup process. The HealthTarget of the first active BRAS detected by the standby BRAS is the first link quality level.

S802, if the standby BRAS receives the VRRP silence message sent by the first active BRAS, transfer to S810.

S803, if the standby BRAS receives the VRRP initialization message sent by the first active BRAS or the low-priority VRRP, transfer to S811.

S804, the standby BRAS compares the IP FPM packet loss ratio (DropHealthTarget) of other N-1 active BRASs, and obtains that the IPFPM index of the p equipment and itself is degraded, and the IP FPM index of the q equipment and itself is normal; p+q= N-1. The DropHealthTarge of the other N-1 active BRASs detected by the standby BRAS is the Y _jth link quality level.

S805, the standby BRAS sends a second request message to other N-1 active BRASs, requesting the other side to answer whether the link to the first active BRAS is degraded; rate degradation) message (through the HealthTarget attribute of the message), and wait for a time window Tw of a response message, and suspend receiving the backup data sent by the first active BRAS within this time window Tw.

S806, within a waiting time window Tw, the standby BRAS receives one reply message. Wherein w replies say that the throughput rate from the peer to the first active BRAS is degraded, s replies say that the health index from the peer to the first active BRAS is not degraded, w+s=l, l<=N-1; Do a show of hands. The response message received by the standby BRAS is the X _i link quality level.

S807, if (p+1>=w or q<s), the backup BRAS is sent to another active BRAS, which is more unstable than the first active BRAS, and votes for a failure of the backup BRAS: transfer to S810.

S808, if (p+1<w and q>=s), the standby BRAS is sent to other active BRASs, which is more stable than the first active BRAS, and votes for the failure of the first active BRAS: immediately raise the VRRP priority of the standby BRAS , transfer to S811.

S809, in cases other than S807 and S808, each device maintains a VRRP state, either as an active device state or as a standby device state, but does not back up data to the peer end in the active device state.

S810, the VRRP of the standby BRAS is transferred to an initialization state, and the RBS backup message is not processed even if it receives the RBS backup message.

S811, the VRRP of the standby BRAS is upgraded from the state of the standby equipment to the state of the active equipment.

The embodiment of Fig. 7 and Fig. 8 is only a kind of implementation mode that uses BRAS to pass VRRP hot backup, but the hot backup method that the embodiment of the invention provides is not limited to this, wherein, the link quality level between each BRAS can include: At least one of packet loss ratio, timeout offline rate ratio, remote backup shock ratio; abnormal level includes at least one of packet loss ratio, timeout offline rate ratio, remote backup shock ratio or any sum exceeding the preset Anomaly ratio threshold.

Fig. 9 is a schematic structural diagram of the first master device provided by the embodiment of the present invention. The first master device provided by this embodiment is applied to a multi-machine hot backup system in which the master device and the backup device are N: 1, where N≥ 2. As shown in Figure 9, the first active device provided in this embodiment includes:

The first obtaining module 91 is configured to obtain a first link quality level, where the first link quality level is used to represent the link quality between the first active device and the backup device.

A judging module 92, configured to judge whether the first link quality level indicates a link abnormality.

The second obtaining module 93 is configured to obtain N-1 links between the first active device and other N-1 active devices when the judging module 92 judges that the first link quality level indicates that the link is abnormal. An X _i -th link quality level, the Xi- _th link quality level is used to represent the link quality between the first active device and the i-th active device, N is greater than or equal to 2, and i is greater than 1 And less than or equal to N.

The sending module 94 is configured to send a request message to each of the other N-1 active devices respectively, and the request message is used to request that one of the other N-1 active devices Each of the active devices respectively sends the link quality level between the other N-1 active devices and the backup device.

The receiving module 95 is configured to respectively receive the M Y _j link quality levels between the jth active device and the backup device sent by the M active devices among the other N-1 active devices, so The Y _j -th link quality level is used to represent the link quality between the j-th active device and the backup device, j is greater than 1 and less than or equal to N, and M is greater than or equal to zero and less than or equal to N-1.

A processing module 96, configured to be used when the sum of the first link quality level and the number of link abnormalities in the M Y _j -th link quality levels is less than N-1 of the X _i -th link quality levels Indicates the number of abnormal links, and the number of normal links in the M Y _j link quality levels is greater than or equal to the normal number of links in the N-1 X _i link quality levels , and when the service of the first master device is damaged, switch the service to the standby device.

The first master device provided in this embodiment is used to implement the technical solution of the method embodiment shown in FIG. 3 , and its implementation principle and technical effect are similar, and will not be repeated here.

Further, in the embodiment shown in FIG. 9 , the processing module 96 is also configured to start a preset silent time window after switching the service to the backup device, and not to switch back to the active device within the silent time window. device status.

Further, in the embodiment shown in FIG. 9 , the processing module 96 is specifically configured to lower the access priority or initialize, and trigger the user accessing the first active device to switch services to the standby device.

Further, in the embodiment shown in FIG. 9 , the sending module 94 is further configured to send the first active device lowering the access priority or initialization message.

Further, in the embodiment shown in FIG. 9 , the first acquiring module 91 is further configured to acquire the first link quality level again after the silent time window expires. The sending module 94 is further configured to notify the standby device to switch services back to the first active device when the judging module 92 judges that the first link quality level indicates that the link is normal.

Further, in the embodiment shown in FIG. 9 , the processing module 96 is further configured to be used when the sum of the numbers indicating link abnormalities in the first link quality level and the M Y _j -th link quality levels is greater than or When it is equal to the number of abnormal links in the N-1 said X _i link quality levels, or when the number of normal links in the M said Y _j -th link quality levels is less than the N-1 said When the link quality level X _i indicates a normal number of links, or when the service of the first active device is not damaged, it does not back up data to the standby device within a preset silent time window.

Further, in the embodiment shown in FIG. 9 , the sending module 94 is also used to determine in the processing module 96 that the link status between the backup device and the other N-1 master devices is worse than that of the first master device. After the device is in the link state of the other N-1 active devices, send the silent time window to the standby device, and the silent time window is used to prevent the standby device from Process the received backup data.

Further, in the embodiment shown in FIG. 9 , the sending module 94 is further configured to wait for a waiting time window starting from sending the request message to each of the other N-1 master devices respectively. , suspend sending backup data to the backup device.

Further, in the embodiment shown in FIG. 9, the master device and the backup device are BRAS;

The processing module 96 is specifically configured to perform hot backup to the backup device through VRRP.

Further, the link quality level in the embodiment shown in FIG. 9 includes: at least one of a packet loss ratio, a timeout offline rate ratio, and a remote backup vibration ratio. The link quality level may be the first link quality level, the X _{i th} link quality level or the Y _{j th} link quality level. The abnormal level includes that at least one of the packet loss ratio, the timeout offline rate ratio, and the remote backup shock ratio exceeds a preset abnormal ratio threshold, or the sum of any two of the above exceeds a preset abnormal ratio threshold.

Fig. 10 is a schematic structural diagram of the backup device provided by the embodiment of the present invention. The backup device provided in this embodiment is applied to a multi-machine hot backup system in which the master device and the backup device are N: 1, where N≥2, as shown in Fig. 10 As shown, the backup equipment provided in this embodiment includes:

The first acquiring module 101 is configured to acquire a first link quality level, where the first link quality level is used to represent the link quality between the first master device and the backup device.

A judging module 102, configured to judge whether the first link quality level indicates a link abnormality.

The second acquiring module 103 is further configured to acquire the backup device and other N-1 units except the first active device when the judging module 102 judges that the first link quality level indicates a link abnormality level N-1 Y _j -th link quality levels between each active device in the active device, and the Y _j -th link quality level is used to indicate that between the backup device and the j-th active device The quality of the link between them, N is greater than or equal to 2, and j is greater than 1 and less than or equal to N.

The sending module 104 is configured to send a request message to each of the other N-1 active devices, and the request message is used to request the other N-1 active devices Each of the active devices respectively sends the link quality level between the other N-1 active devices and the first active device.

The receiving module 105 is configured to respectively receive the M X _i link quality levels between the i-th active device and the first device sent by the other N-1 active devices, and the X _i -th link quality levels The road quality level is used to indicate the link quality between the i-th active device and the first active device, i is greater than 1 and less than or equal to N, and M is greater than or equal to zero and less than or equal to N-1.

A processing module 106, configured to be used when the sum of the first link quality level and the number of link abnormalities in the N-1 Y _jth link quality levels is less than M said X _i link quality levels Indicates the number of link abnormalities, and the number of normal links in the Y _j link quality level of N-1 is greater than or equal to the normal number of links in the X _i link quality level of M, Raising the access priority, where the raised access priority is used to instruct a user of the first active device to switch to the standby device.

The standby device provided in this embodiment is used to implement the technical solution of the method embodiment shown in FIG. 5 , and its implementation principle and technical effect are similar, and details are not repeated here.

Further, in the embodiment shown in FIG. 10 , the processing module 106 is further configured to: when the first link quality level and the sum of the numbers representing link abnormalities in the N-1 Y _j -th link quality levels greater than or equal to the number of M said X _i link quality levels indicating that the link is abnormal, or the number of N-1 said Y _j link quality levels indicating that the link is normal is less than the M said X _i link quality level The link quality level indicates the number of normal links, and the receiving of the backup data sent by the first active device is suspended within the preset silent time window.

Further, in the embodiment shown in FIG. 10 , the receiving module 105 is also configured to receive the silent time window sent by the first master device; the processing module 106 is also configured not to process the received silent time window within the silent time window. Backup data.

Further, in the embodiment shown in FIG. 10 , the receiving module 105 is also configured to receive a message that the first master device lowers the access priority or initializes the message sent by the first master device; the processing module 106 also uses After receiving the message that the first active device lowers the access priority or initializes, the user accepting the first active device switches to the standby device.

Further, in the embodiment shown in FIG. 10 , the active device and the backup device are BRASs; the first active device is hot-backed up to the backup device through VRRP.

Further, the link quality level in the embodiment shown in FIG. 10 includes: at least one of a packet loss ratio, a timeout offline rate ratio, and a remote backup vibration ratio. The link quality level may be the first link quality level, the X _{i th} link quality level or the Y _{j th} link quality level. The abnormal level includes that at least one of the packet loss ratio, the timeout offline rate ratio, and the remote backup shock ratio exceeds a preset abnormal ratio threshold, or the sum of any two of the above exceeds a preset abnormal ratio threshold.

FIG. 11 is a schematic structural diagram of a first master device provided by an embodiment of the present invention. As shown in FIG. 11 , the first master device in this embodiment includes: a processor 111 , a transmitter 112 , and a receiver 113 . Optionally, the first master device may further include a memory 114 . Wherein, the processor 111, the transmitter 112, the receiver 113 and the memory 114 can be connected by a system bus or in other ways. In FIG. Peripheral Component Interconnect (PCI) bus or Extended Industrial Standard Architecture (Extended Industrial Standard Architecture, EISA) bus, etc. The system bus can be divided into address bus, data bus, control bus and so on. For ease of representation, only one line is used in FIG. 11 , but it does not mean that there is only one bus or one type of bus. The memory 114 is used to store computer programs. The processor 111 can read the code corresponding to the stored computer program from the memory 114, and perform the following operations:

Acquiring a first link quality level, where the first link quality level is used to represent the link quality between the first active device and the backup device;

judging whether the first link quality level indicates a link abnormality;

When it is judged that the first link quality level indicates that the link is abnormal, acquiring the N-1 _th link quality levels between the first active device and other N-1 active devices, the The Xi- _th link quality level is used to indicate the link quality between the first active device and the i-th active device, N is greater than or equal to 2, and i is greater than 1 and less than or equal to N;

Through the transmitter 112, a request message is sent to each of the other N-1 active devices, and the request message is used to request the other N-1 active devices. Each active device sends the link quality level between the other N-1 active devices and the backup device;

Through the receiver 113, respectively receive the M Y _j -th link quality levels between the j-th active device and the backup device sent by the M active devices among the other N-1 active devices, the The Y _j -th link quality level is used to represent the link quality between the j-th active device and the backup device, where j is greater than 1 and less than or equal to N, and M is greater than or equal to zero and less than or equal to N-1.

When the sum of the first link quality level and the number of link abnormalities in the M Y _j -th link quality levels is less than N-1 of the X _i -th link quality levels that indicate link abnormalities number, and the number of normal links in the M Y _j -th link quality levels is greater than or equal to the number of normal links in the N-1 X _i -th link quality levels, and the first When the service of the active device is damaged, the service is switched to the backup device.

The first active device in this embodiment is used to implement the hot backup method shown in FIG. 3 , and its implementation principle and technical effect are similar, and will not be repeated here.

In the first master device provided in this embodiment, the processor 111 is configured to implement the processing of the judgment module 92, the processing module 96, the first acquisition module 91, and the second acquisition module 93 in the first master device shown in FIG. 9; The transmitter 112 is used to realize the processing of the sending module 94 in the first master device shown in FIG. 9 ; the receiver 113 is used to realize the processing of the receiving module 95 in the first master device shown in FIG. 9 .

FIG. 12 is a schematic structural diagram of a backup device provided by an embodiment of the present invention. As shown in FIG. 12 , the backup device in this embodiment includes: a processor 121 , a transmitter 122 , and a receiver 123 . Optionally, the backup device may also include a memory 124 . Wherein, the processor 121, the transmitter 122, the receiver 123 and the memory 124 can be connected through a system bus or other means, and the system bus connection is taken as an example in FIG. The system bus can be divided into address bus, data bus, control bus and so on. For ease of representation, only one line is used in FIG. 12 , but it does not mean that there is only one bus or one type of bus. The memory 124 is used to store computer programs. The processor 121 can read the code corresponding to the stored computer program from the memory 124, and perform the following operations:

Acquiring a first link quality level, where the first link quality level is used to represent the link quality between the first active device and the backup device;

judging whether the first link quality level indicates a link abnormality;

When it is used to determine that the first link quality level indicates a link abnormality level, obtain a link between the backup device and each of the other N-1 active devices except the first active device. N-1 Y _j -th link quality levels between, the Y _j -th link quality level is used to represent the link quality between the backup device and the j-th active device, and N is greater than or equal to 2 , j is greater than 1 and less than or equal to N;

Send a request message to each of the other N-1 active devices through the transmitter 122, and the request message is used to request each of the other N-1 active devices Each active device respectively sends the link quality level between the other N-1 active devices and the first active device;

Through the receiver 123, respectively receive the M X _i -th link quality levels between the i-th active device and the first active device sent by the other N-1 active devices, and the X _i -th The link quality level is used to indicate the link quality between the i-th active device and the first active device, where i is greater than 1 and less than or equal to N, and M is greater than or equal to zero and less than or equal to N-1;

When the sum of the first link quality level and the number of link abnormalities in the N-1 Y _j -th link quality levels is less than the number of link abnormalities in the X _i -th link quality levels number, and the number of normal links in the Y _jth link quality level of N-1 is greater than or equal to the normal number of links in the X _i link quality level of M, and the access priority is increased, The elevated access priority is used to instruct a user of the first active device to switch to the backup device.

In the backup device provided in this embodiment, the processor 121 is used to implement the processing of the first acquisition module 101, the second acquisition module 103, the judgment module 102, and the processing module 106 in the backup device shown in FIG. 10; the transmitter 122 is used to implement The processing of the sending module 104 in the backup device shown in FIG. 10 ; the receiver 123 is used to implement the processing of the receiving module 105 in the backup device shown in FIG. 10 .

Figure 13 is a schematic structural diagram of Embodiment 1 of the communication system provided by the embodiment of the present invention. The system includes N active devices 131 and one backup device 132 .

The master device 131 includes the first master device as shown in FIG. 9 or FIG. 11 ; the backup device 132 includes the backup device as shown in FIG. 10 or 12 .

Those of ordinary skill in the art can understand that all or part of the steps for implementing the above method embodiments can be completed by program instructions and related hardware. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above-mentioned method embodiments; and the aforementioned storage medium includes: ROM, RAM, magnetic disk or optical disk and other various media that can store program codes.

Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present invention, rather than limiting them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: It is still possible to modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the technical solutions of the various embodiments of the present invention. scope.

Claims (13)

1. A hot backup method, characterized in that the method comprises: The first active device acquires a first link quality level, where the first link quality level is used to represent the link quality between the first active device and the backup device; When the first link quality level indicates that the link is abnormal, the first active device acquires the N-1 X _i -th links between the first active device and other N-1 active devices Road quality level, the X _i -th link quality level is used to indicate the link quality between the first active device and the i-th active device, N is greater than or equal to 2, i is greater than 1 and less than or equal to N ; The first active device sends a request message to each of the other N-1 active devices, and the request message is used to request the other N-1 active devices Each active device in the above respectively sends the link quality level between the other N-1 active devices and the backup device; The first active device respectively receives the M Y _j -th link quality levels between the j-th active device and the backup device sent by the M active devices among the other N-1 active devices, The Y _j -th link quality level is used to represent the link quality between the j-th active device and the backup device, where j is greater than 1 and less than or equal to N, and M is greater than zero and less than or equal to N-1; When the sum of the first link quality level and the number of link abnormalities in the M Y _j -th link quality levels is less than N-1 of the X _i -th link quality levels that indicate link abnormalities number, and the number of normal links in the M Y _j -th link quality levels is greater than or equal to the number of normal links in the N-1 X _i -th link quality levels, and the first When the service of the active device is damaged, the first active device switches the service to the backup device. 2. The method according to claim 1, characterized in that the method further comprises: When the sum of the first link quality level and the number of link abnormalities in the M Y _j -th link quality levels is greater than or equal to N-1 of the X _i -th link quality levels, the link When the number of abnormalities, or when the number of normal links in the M Y _j -th link quality levels is less than the normal number of links in the N-1 X _i -th link quality levels, or the When the service of the first active device is not damaged, the first active device does not back up data to the standby device within a preset silent time window. 3. The method according to claim 2, further comprising: When the silent time window expires, the first active device acquires the first link quality level again; When the first link quality level indicates that the link is normal, the first active device notifies the backup device to switch services back to the first active device. 4. The method according to any one of claims 1 to 3, further comprising: The first active device suspends sending backup data to the backup device within a waiting time window starting from sending the request message to each of the other N-1 active devices respectively . 5. A hot backup method, characterized in that the method comprises: The backup device acquires a first link quality level, where the first link quality level is used to indicate the link quality between the first active device and the backup device; When the first link quality level indicates a link abnormality level, the backup device acquires the backup device and each of the other N-1 active devices except the first active device. N-1 Y _j -th link quality levels between devices, the Y _j -th link quality level is used to represent the link quality between the backup device and the j-th active device, and N is greater than or Equal to 2, j greater than 1 and less than or equal to N; The backup device sends a request message to each of the other N-1 active devices, and the request message is used to request each of the other N-1 active devices One active device respectively sends the link quality levels between the other N-1 active devices and the first active device; The backup device respectively receives the M X _i link quality levels between the i-th active device and the first active device sent by the other N-1 active devices, and the X _i link The road quality level is used to indicate the link quality between the i-th active device and the first active device, where i is greater than 1 and less than or equal to N, and M is greater than zero and less than or equal to N-1; When the sum of the first link quality level and the number of link abnormalities in the N-1 Y _j -th link quality levels is less than the number of link abnormalities in the X _i -th link quality levels number, and the quantity indicating that the link is normal in the N-1 said Y _jth link quality level is greater than or equal to the number indicating that the link is normal in the M said _Xi'th link quality level, and the backup device promotes the connection The access priority after the promotion is used to instruct the user of the first active device to switch to the backup device. 6. The method according to claim 5, further comprising: When the sum of the first link quality level and the number of N-1 Y _j -th link quality levels indicating link anomalies is greater than or equal to M said X _i -th link quality levels indicating link anomalies , or the number of normal links in the N-1 Y _jth link quality levels is less than the normal number of links in the M X _i link quality levels, and the backup device is in the pre-preparation Suspend receiving the backup data sent by the first master device within the preset silent time window. 7. A first active device, characterized in that the first active device comprises: A first acquiring module, configured to acquire a first link quality level, where the first link quality level is used to represent the link quality between the first active device and the backup device; A judging module, configured to judge whether the first link quality level indicates a link abnormality; The second acquiring module is configured to acquire N-1 links between the first master device and other N-1 master devices when the judging module judges that the first link quality level indicates that the link is abnormal An X _i -th link quality level, the Xi- _th link quality level is used to represent the link quality between the first active device and the i-th active device, N is greater than or equal to 2, and i is greater than 1 and less than or equal to N; A sending module, configured to send a request message to each of the other N-1 active devices, and the request message is used to request each of the other N-1 active devices Each active device sends the link quality level between the other N-1 active devices and the backup device; The receiving module is configured to respectively receive the M Y _j link quality levels between the jth active device and the backup device sent by the M active devices among the other N-1 active devices, the The Y _j -th link quality level is used to represent the link quality between the j-th active device and the standby device, where j is greater than 1 and less than or equal to N, and M is greater than zero and less than or equal to N-1; A processing module, configured to be used when the sum of the first link quality level and the number of link abnormalities in the M Y _j -th link quality levels is less than N-1 of the X _i -th link quality levels Represents the number of link abnormalities, and the number of normal links in the M Y _j link quality levels is greater than or equal to the normal number of links in the N- ₁ Xi link quality levels, And when the service of the first master device is damaged, switch the service to the backup device. 8. The first active device according to claim 7, wherein the processing module is further configured to indicate when the first link quality level and the M Y _j -th link quality levels When the sum of the number of link abnormalities is greater than or equal to the number of link abnormalities in the N-1 said X _i link quality levels, or when the link is normal in the M said Y _j link quality levels When the number is less than the N-1 number indicating that the link is normal in the X _i -th link quality level, or when the first active device's own business is not damaged, it does not send a link within the preset silent time window. The backup device backs up data. 9. The first active device according to claim 8, wherein the first acquiring module is further configured to acquire the first link quality level again after the silent time window expires; The sending module is further configured to notify the standby device to switch services back to the first active device when the judging module judges that the first link quality level indicates that the link is normal. 10. The first master device according to any one of claims 7-9, characterized in that the sending module is further configured to slave to each of the other N-1 master devices Within the waiting time window after the user device sends the request message respectively, the sending of the backup data to the backup device is suspended. 11. A backup device, characterized in that the backup device comprises: A first obtaining module, configured to obtain a first link quality level, where the first link quality level is used to represent the link quality between the first active device and the backup device; A judging module, configured to judge whether the first link quality level indicates a link abnormality; The second acquisition module is configured to acquire the backup device and other N-1 masters except the first master device when the judging module judges that the first link quality level indicates a link abnormality level N-1 Y _j -th link quality levels between each active device in the active device, and the Y _j -th link quality level is used to indicate that between the backup device and the j-th active device The link quality of , N is greater than or equal to 2, j is greater than 1 and less than or equal to N; A sending module, configured to send a request message to each of the other N-1 active devices, and the request message is used to request each of the other N-1 active devices Each active device respectively sends the link quality level between the other N-1 active devices and the first active device; A receiving module, configured to respectively receive the M X _i -th link quality levels between the i-th active device and the first active device sent by the other N-1 active devices, and the X _i -th The link quality level is used to indicate the link quality between the i-th active device and the first active device, i is greater than 1 and less than or equal to N, and M is greater than zero and less than or equal to N-1; A processing module, configured to be used when the sum of the first link quality level and the number of abnormal links in the N-1 Y _j -th link quality levels is less than M of the X _i -th link quality levels Indicates the number of link abnormalities, and the number of normal links in the Y _jth link quality level of N-1 is greater than or equal to the number of normal links in the X _ith link quality level of M, and the improvement Access priority, where the elevated access priority is used to instruct a user of the first active device to switch to the standby device. 12. The standby device according to claim 11, wherein the processing module is further configured to indicate a link in the first link quality level and the N-1 Y _jth link quality levels The sum of the number of path abnormalities is greater than or equal to the number of abnormal links in the X _i -th link quality level, or the number of normal links in the N-1 Y- _th link quality level is less than M In the X _{i th} link quality level, the number indicates that the link is normal, and the reception of the backup data sent by the first master device is suspended within the preset silent time window. 13. A communication system, characterized in that the communication system comprises N main devices and the backup device according to claim 11 or 12, any one of the N main devices is the right The first active device described in any one of 7-10 is required.