CN117857233A - Multicast high-availability market distribution method and device - Google Patents

Abstract

The invention discloses a method and a device for distributing multicast high availability quotation, which realize optimal multicast replication by only sending one multicast flow from a quotation center to public clouds, save the bandwidth cost from the quotation center to each member center, creatively propose an elastic telescopic group of multicast receivers, select a member of one AZ of the elastic telescopic group as the multicast receiver nearby in order to save the bandwidth among AZ, select another member of the telescopic group only when members of the same telescopic group and AZ fail, save the bandwidth and achieve the effect of high availability crossing AZ.

Description

A method and device for multicast high-availability market information distribution

Technical Field

The present invention relates to the technical field of network address translation in data communication, and in particular to a method and device for multicast high-availability market information distribution.

Background technique

When the customer's securities and stock business is deployed on a cloud host in a public cloud, it is necessary to support multicast functions in the public cloud. At present, major market centers have realized disaster recovery of market sources. For example, the five major futures exchanges have established trading main centers and trading backup centers. Service providers currently use IDC computer rooms to achieve high availability of market centers, but the multicast disaster recovery system of member units themselves is not sound:

(1) When a member unit’s computer room fails, the entire member unit cannot receive market information.

(2) Secondly, each member unit needs to deploy two dedicated lines to connect to the production center and the disaster recovery center, which is also costly. The market information required by other member units is the same, and the market information center sends a copy of the market information to each member unit, which also consumes a lot of network bandwidth.

(3) Third, if each member unit deploys two quotation servers in two computer rooms, firstly, the number of dedicated lines will increase, and secondly, the two quotation servers will be separated from each other. Two sets of certificates will also be required, which will increase the cost. In addition, the two quotation servers will consume bandwidth to receive quotation information.

Summary of the invention

The purpose of this section is to summarize some aspects of the embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the abstract and invention title of this application to avoid blurring the purpose of this section, the abstract and the invention title, and such simplifications or omissions shall not be used to limit the scope of the present invention.

In view of the problems existing in the above-mentioned existing method and device for multicast high-availability market information distribution, the present invention is proposed.

Therefore, the object of the present invention is to provide a method and device for multicast high-availability market information distribution, which is to solve the above-mentioned problems.

In order to solve the above technical problems, the present invention provides the following technical solutions:.

As a method for multicast high-availability market information distribution described in the present invention, the process method is as follows:

S1. Deploy a multicast gateway MCGW in a cluster across multiple AZs. For example, deploy one MCGW in AZ1 and one MCGW in AZ2.

S2. The multicast gateway MCGW synchronizes the multicast entry information through the DCI interconnection link of the two AZs;

S3. Each member unit will encrypt and pre-install the authentication certificate on the MCGW multicast gateway, which will then be sent to various market centers for certificate authentication.

S4. Each member unit creates a multicast receiver elastic scaling group. Each scaling group has at least one quotation server in each AZ. Set the priority of the elastic scaling group members. For example, the multicast receiver elastic scaling group 1 of member unit 1 has two members, and the priority of quotation server 1 is 100, and the priority of quotation server 2 is 50, indicating that quotation server 1 is selected first to receive quotation information.

S5. Each member in the elastic scaling group joins the multicast group according to the following logic. It innovatively proposes to judge the reachability of the network card and MCGW according to the query message sent by the multicast gateway received by the network card of the market information server within a certain period of time. In this way, only one network card of the same market information server joins the multicast group at the same time, ensuring that the market information server will not receive the same multicast message from two network cards at the same time, which will cause abnormal processing of the market information server.

S6. Configure the IP addresses of the production center and the disaster recovery center on the multicast gateway MCGW, and set the multicast source priority. For example, set the production center to 100 and the disaster recovery center to 50. It is assumed that the larger the number, the higher the priority.

S7. According to the priority of the multicast source, the production center is selected to initiate unicast certificate authentication. When the unicast of the production center is unreachable, MCGW attempts to initiate certificate authentication to the disaster recovery center. At this time, since the unicast route from MCGW2 to the disaster recovery center has a higher priority, MCGW2 will be used to initiate authentication.

S8, the multicast gateway uses VIP and POP devices to establish a vxlan unicast tunnel. The multicast traffic enters from the POP devices of the two AZs and is encapsulated in the vxlan tunnel to enter the multicast gateway. When the production center sends market information as a multicast source, MCGW1 and MCGW2 both publish VIP BGP routes to POP, but POP1 receives the VIP route from MCGW1 because the AS number is shorter and is preferred. Therefore, MCGW1 receives the multicast market information of the production center.

S9. After receiving the message, MCGW can query the AZ and scaling group information described in the market server according to the OVS address information, and after recording the multicast group IGMP report message, it will generate a multicast table entry.

S10. Each member of MCGW periodically checks the multicast members through IGMP query. If there is no response after the timeout, the multicast group will be deleted.

S11. When the network card of market server 1 in elastic scaling group 1 fails, the IGMP query message sent by MCGW1 and MCGW2 times out without response. When the failure is detected, network card A in the multicast table is deleted. At the same time, the market server of the member unit will also detect that it is unreachable, and use another network card to send IGMP Join to MCGW.

S12, when the member unit server 1 fails, network card A and network card B are unreachable. When the member unit server 1 fails, network card A and network card B are unreachable, MCGW1 will select network card E of AZ2. The multicast market information will pass through the production center to MCGW1 and then to the DCI interconnection link, and then to network card E. At the same time, an alarm will be sent, and a new virtual machine will be automatically pulled up as the market information server of the member unit of AZ1;

S13, when the entire AZ1 fails, MCGW2 detects that member MCGW1 is no longer active through the timeout of the inter-group synchronization message, and initiates unicast authentication to the disaster recovery center based on the certificate installed by the member unit. It also detects that network cards A, B, C, and D are all unreachable, queries the multicast table entry, selects another AZ, namely the receiver of AZ2, changes the multicast table entry of AZ2 to receive multicast, and selects to send multicast market data to network cards E and G.

As a preferred solution of the method for multicast high-availability market information distribution described in the present invention, wherein: multiple member units are placed on a public cloud platform, and resources are isolated through VPC. In order to save dedicated line costs and achieve AZ and high availability at the same time, the production center and AZ1 are interconnected, and the disaster recovery center of the market information center and AZ2 are interconnected.

As a preferred solution of the multicast high-availability market information distribution method described in the present invention, the elastic scaling group of member unit 1 is market information servers 1-100 and market information servers 3-50, and the elastic scaling group 2 of member unit 2 is market information servers 2 and market information servers 4.

As a preferred solution of the method for multicast high-availability market information distribution described in the present invention, the market information server of each member unit will select the network card with the highest member priority to forward the multicast market information, such as member unit 1 selects network card A of market information server 1 to receive market information.

As a preferred solution of the method for multicast high-availability market distribution described in the present invention, wherein: each market server in the multicast elastic scaling group selects a network card to send an IGMP Join multicast group joining message to the outside. After this IGMP join message arrives at the OVS of the market server, the OVS adds vxlan encapsulation according to the flow table, the outer vxlan address source IP uses the OVS address, and the destination IP address uses the MCGW cluster VIP address.

As a preferred solution of the method for multicast high-availability market information distribution described in the present invention, the market information is transmitted from the disaster recovery center through the multicast gateway member MCGW2 to the network card E and the network card G, thereby realizing the continuous reception of market information of member units 1 and 2.

As a preferred solution of the method for multicast high-availability market information distribution described in the present invention, wherein: the VXLAN is the abbreviation of Virtual eXtensible LAN, that is, extensible virtual local area network, and the IGMP is the abbreviation of Internet Group Management Protocol, that is, Internet Group Management Protocol snooping.

As a preferred solution of the method for multicast high-availability market distribution described in the present invention, the VTEP is the abbreviation of VXLAN Tunnel End Point, that is, the VXLAN tunnel endpoint, the VSI is the abbreviation of Virtual SwitchInstance, that is, the virtual switching instance, and the VTEP is a virtual switching instance that provides a Layer 2 switching service for a VXLAN.

As a preferred solution of the method for multicast high-availability market distribution described in the present invention, the method uses a device for multicast high-availability market distribution, including multiple switches and routers. When the server composed of the switches and routers is running, the method for multicast high-availability market distribution described in any one of claims 1-8 is used.

The present invention also provides a device, including: a device for multicast high-availability market information distribution as described in claim 9, wherein the device is mainly composed of a plurality of switches and routers, and is externally assembled with a server housing.

Beneficial effects of the present invention:

1. Only one copy of multicast traffic needs to be sent from the market center to the public cloud, achieving optimal multicast replication and saving bandwidth costs from the market center to each member center;

2. The elastic scaling group of multicast receivers is proposed. In order to save bandwidth between AZs, a member of an AZ in the elastic scaling group is selected as the multicast receiver. Only when all members in the same AZ in the same scaling group fail will a member receiver of another AZ in the scaling group be selected. This not only saves bandwidth but also achieves high availability across AZs.

3. Implement cross-AZ deployment of multicast gateways, establish unicast vxlan tunnels with POP access points, and automatically select the nearest optimal reception source based on the length of the BGPAZ number;

4. Reusing the existing IGMP query and report mechanisms, innovatively implementing the reverse health check of the multicast receiver on the multicast gateway, without adding additional detection messages, the multicast gateway can switch the receiver, and at the same time achieve the dual effect of switching the network card of the market information server, achieving high availability of the AZ level for member units to receive market information;

5. The multicast gateway obtains the multicast group joining information from the IGMP report joining message carrying vxlan encapsulation sent by the market server network card, and can query the market server AZ information based on the vxlan encapsulation. Then the multicast gateway preferentially receives the market information from the market server with the same AZ as the multicast source, thereby reducing the occupancy of the DCI cross-AZ link bandwidth.

BRIEF DESCRIPTION OF THE DRAWINGS

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative labor. Among them:

FIG1 is a flow chart of a method for multicast high-availability market information distribution proposed by the present invention;

FIG2 is a schematic diagram of the logic flow of the elastic scaling group of the present invention;

FIG3 is a schematic diagram of the processing logic of the multicast gateway of the present invention;

FIG4 is a schematic diagram of the detection flow when AZ1 fails in the present invention.

Detailed ways

In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

Secondly, the present invention is described in detail with reference to schematic diagrams. When describing the embodiments of the present invention in detail, for the sake of convenience, the cross-sectional diagrams showing the device structure will not be partially enlarged according to the general scale, and the schematic diagrams are only examples, which should not limit the scope of protection of the present invention. In addition, in actual production, the three-dimensional dimensions of length, width and depth should be included.

1-4, the present invention provides a method for multicast high-availability market information distribution, the process method is as follows:

S1. Deploy a multicast gateway MCGW in a cluster across multiple AZs. For example, deploy one MCGW in AZ1 and one MCGW in AZ2.

S2. The multicast gateway MCGW synchronizes the multicast entry information through the DCI interconnection link of the two AZs;

S3. Each member unit will encrypt and pre-install the authentication certificate on the MCGW multicast gateway, which will then be sent to various market centers for certificate authentication.

S4. Each member unit creates a multicast receiver elastic scaling group. Each scaling group has at least one quotation server in each AZ. Set the priority of the elastic scaling group members. For example, the multicast receiver elastic scaling group 1 of member unit 1 has two members, and the priority of quotation server 1 is 100, and the priority of quotation server 2 is 50, indicating that quotation server 1 is selected first to receive quotation information.

S5. Each member in the elastic scaling group joins the multicast group according to the following logic. It innovatively proposes to judge the reachability of the network card and MCGW according to the query message sent by the multicast gateway received by the network card of the market information server within a certain period of time. In this way, only one network card of the same market information server joins the multicast group at the same time, ensuring that the market information server will not receive the same multicast message from two network cards at the same time, which will cause abnormal processing of the market information server.

S6. Configure the IP addresses of the production center and the disaster recovery center on the multicast gateway MCGW, and set the multicast source priority. For example, set the production center to 100 and the disaster recovery center to 50. It is assumed that the larger the number, the higher the priority.

S7. According to the priority of the multicast source, the production center is selected to initiate unicast certificate authentication. When the unicast of the production center is unreachable, MCGW attempts to initiate certificate authentication to the disaster recovery center. At this time, since the unicast route from MCGW2 to the disaster recovery center has a higher priority, MCGW2 will be used to initiate authentication.

S8, the multicast gateway uses VIP and POP devices to establish a vxlan unicast tunnel. The multicast traffic enters from the POP devices of the two AZs and is encapsulated in the vxlan tunnel to enter the multicast gateway. When the production center sends market information as a multicast source, MCGW1 and MCGW2 both publish VIP BGP routes to POP, but POP1 receives the VIP route from MCGW1 because the AS number is shorter and is preferred. Therefore, MCGW1 receives the multicast market information of the production center.

S9. After receiving the message, MCGW can query the AZ and scaling group information of the market server according to the OVS address information, and after recording the multicast group IGMP report message, it will generate a multicast table entry.

S10. Each member of MCGW periodically checks the multicast members through IGMP query. If there is no response after the timeout, the multicast group will be deleted.

S11. When the network card of market server 1 in elastic scaling group 1 fails, the IGMP query message sent by MCGW1 and MCGW2 times out without response. When the failure is detected, network card A in the multicast table is deleted. At the same time, the market server of the member unit will also detect that it is unreachable, and use another network card to send IGMP Join to MCGW.

S12, when the member unit server 1 fails, network card A and network card B are unreachable. When the member unit server 1 fails, network card A and network card B are unreachable, MCGW1 will select network card E of AZ2. The multicast market information will pass through the production center to MCGW1 and then to the DCI interconnection link, and then to network card E. At the same time, an alarm will be sent, and a new virtual machine will be automatically pulled up as the market information server of the member unit of AZ1;

S13, when the entire AZ1 fails, MCGW2 detects that member MCGW1 is no longer active through the timeout of the inter-group synchronization message, and initiates unicast authentication to the disaster recovery center based on the certificate installed by the member unit. It also detects that network cards A, B, C, and D are all unreachable, queries the multicast table entry, selects another AZ, namely the receiver of AZ2, changes the multicast table entry of AZ2 to receive multicast, and selects to send multicast market data to network cards E and G.

Among them, multiple member units are placed on the public cloud platform, and resources are isolated through VPC. In order to save dedicated line costs and achieve AZ and high availability at the same time, the production center and AZ1 are interconnected, and the disaster recovery center of the market center and AZ2 are interconnected.

Furthermore, the elastic scaling group of member unit 1 is market servers 1-100 and market servers 3-50, and the elastic scaling group 2 of member unit 2 is market servers 2 and market servers 4.

Furthermore, the market information server of each member unit will select the network card with the highest member priority to forward the multicast market information. For example, member unit 1 selects network card A of market information server 1 to receive the market information.

Furthermore, each market server in the multicast elastic scaling group selects a network card to send an IGMPJoin multicast group joining message. After this IGMP joining message reaches the OVS of the market server, OVS adds vxlan encapsulation according to the flow table. The outer vxlan address source IP uses the OVS address, and the destination IP address uses the MCGW cluster VIP address.

Furthermore, the market information is transmitted from the disaster recovery center to the network cards E and G via the multicast gateway member MCGW2, thereby achieving continuous reception of market information of the member units 1 and 2.

Furthermore, VXLAN is the abbreviation of Virtual eXtensible LAN, which is an extensible virtual local area network, and IGMP is the abbreviation of Internet Group Management Protocol, which is Internet Group Management Protocol.

Furthermore, VTEP is the abbreviation of VXLAN Tunnel End Point, that is, the VXLAN tunnel endpoint, and VSI is the abbreviation of Virtual Switch Instance, that is, the virtual switching instance. VTEP is a virtual switching instance that provides a Layer 2 switching service for a VXLAN.

Furthermore, the method uses a device for multicast high-availability market distribution, including multiple switches and routers. When the server composed of the switches and routers is running, a method for multicast high-availability market distribution as described in any one of claims 1-8 is used.

The present invention also provides a device, including a multicast high-availability market information distribution device as described in claim 9, wherein the device is mainly composed of a plurality of switches and routers, and is externally assembled with a server housing.

It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims (10)

1. A method for multicast high-availability market information distribution, characterized in that: the process method is as follows: S1. Deploy a multicast gateway MCGW in a cluster across multiple AZs. For example, deploy one MCGW in AZ1 and one MCGW in AZ2. S2. The multicast gateway MCGW synchronizes the multicast entry information through the DCI interconnection link of the two AZs; S3. Each member unit will encrypt and pre-install the authentication certificate on the MCGW multicast gateway, which will then be sent to various market centers for certificate authentication. S4. Each member unit creates a multicast receiver elastic scaling group. Each scaling group has at least one quotation server in each AZ. Set the priority of the elastic scaling group members. For example, the multicast receiver elastic scaling group 1 of member unit 1 has two members, and the priority of quotation server 1 is 100, and the priority of quotation server 2 is 50, indicating that quotation server 1 is selected first to receive quotation information. S5. Each member in the elastic scaling group joins the multicast group according to the following logic. It innovatively proposes to judge the reachability of the network card and MCGW according to the query message sent by the multicast gateway received by the network card of the market information server within a certain period of time. In this way, only one network card of the same market information server joins the multicast group at the same time, ensuring that the market information server will not receive the same multicast message from two network cards at the same time, which will cause abnormal processing of the market information server. S6. Configure the IP addresses of the production center and the disaster recovery center on the multicast gateway MCGW, and set the multicast source priority. For example, set the production center to 100 and the disaster recovery center to 50. It is assumed that the larger the number, the higher the priority. S7. According to the priority of the multicast source, the production center is selected to initiate unicast certificate authentication. When the unicast of the production center is unreachable, MCGW attempts to initiate certificate authentication to the disaster recovery center. At this time, since the unicast route from MCGW2 to the disaster recovery center has a higher priority, MCGW2 will be used to initiate authentication. S8, the multicast gateway uses VIP and POP devices to establish a vxlan unicast tunnel. The multicast traffic enters from the POP devices of the two AZs and is encapsulated in the vxlan tunnel to enter the multicast gateway. When the production center sends market information as a multicast source, MCGW1 and MCGW2 both publish VIP BGP routes to POP, but POP1 receives the VIP route from MCGW1 because the AS number is shorter and is preferred. Therefore, MCGW1 receives the multicast market information of the production center. S9. After receiving the message, MCGW can query the AZ and scaling group information described in the market server according to the OVS address information, and after recording the multicast group IGMP report message, it will generate a multicast table entry. S10. Each member of MCGW periodically checks the multicast members through IGMP query. If there is no response after the timeout, the multicast group will be deleted. S11. When the network card of market server 1 in elastic scaling group 1 fails, the IGMP query message sent by MCGW1 and MCGW2 times out without response. When the failure is detected, network card A in the multicast table is deleted. At the same time, the market server of the member unit will also detect that it is unreachable, and use another network card to send IGMP Join to MCGW. S12, when the member unit server 1 fails, network card A and network card B are unreachable. When the member unit server 1 fails, network card A and network card B are unreachable, MCGW1 will select network card E of AZ2. The multicast market information will pass through the production center to MCGW1 and then to the DCI interconnection link, and then to network card E. At the same time, an alarm will be sent, and a new virtual machine will be automatically pulled up as the market information server of the member unit of AZ1; S13. When the entire AZ1 fails, MCGW2 detects that member MCGW1 is no longer active through the timeout of the inter-group synchronization message. It will initiate a unicast authentication to the disaster recovery center based on the certificate installed by the member unit, and detect that network cards A, B, C, and D are all unreachable. It queries the multicast table entry and selects another AZ, namely the receiver of AZ2, and changes the multicast table entry of AZ2 to receive multicast, and selects to send multicast market data to network cards E and G. 2. According to claim 1, a method for multicast high-availability market distribution is characterized in that: multiple member units are placed on a public cloud platform, and resources are isolated through VPC. In order to save dedicated line costs and achieve AZ and high availability at the same time, the production center and AZ1 are interconnected, and the disaster recovery center of the market center and AZ2 are interconnected. 3. According to a method for multicast high-availability market information distribution according to claim 2, it is characterized in that the elastic scaling group of member unit 1 is market information servers 1-100 and market information servers 3-50, and the elastic scaling group 2 of member unit 2 is market information servers 2 and market information servers 4. 4. A method for multicast high-availability market information distribution according to claim 3, characterized in that: the market information server of each member unit will select the network card with the highest member priority to forward the multicast market information, such as member unit 1 selects network card A of market information server 1 to receive the market information. 5. According to the method of multicast high-availability market distribution described in claim 4, it is characterized in that: each market server in the multicast elastic scaling group selects a network card to send an IGMP Join multicast group joining message to the outside. After this IGMP join message arrives at the OVS of the market server, the OVS adds vxlan encapsulation according to the flow table, the outer vxlan address source IP uses the OVS address, and the destination IP address uses the MCGW cluster VIP address. 6. A method for multicast high-availability market information distribution according to claim 5, characterized in that: the market information is transmitted from the disaster recovery center through the multicast gateway member MCGW2 to the network card E and the network card G, thereby realizing continuous reception of market information of member units 1 and 2. 7. A method for distributing multicast high-availability market information according to claim 5, characterized in that: the VXLAN is the abbreviation of Virtual eXtensible LAN, that is, the extensible virtual local area network, and the IGMP is the abbreviation of Internet Group Management Protocol, that is, Internet Group Management Protocol snooping. 8. A method for distributing multicast high-availability market information according to claim 7, characterized in that: the VTEP is the abbreviation of VXLAN Tunnel End Point, that is, the VXLAN tunnel endpoint, the VSI is the abbreviation of Virtual Switch Instance, that is, the virtual switching instance, and the VTEP is a virtual switching instance that provides a Layer 2 switching service for a VXLAN. 9. A method for multicast high-availability market distribution according to claim 5, characterized in that: the method uses a device for multicast high-availability market distribution, including multiple switches and routers, and when the server composed of the switches and routers is running, a method for multicast high-availability market distribution according to any one of claims 1-8. 10. A device, characterized in that it comprises a multicast high-availability market information distribution device as claimed in claim 9, wherein the device is mainly composed of a plurality of switches and routers, and a server housing is assembled on the outside.