CN107483518B

CN107483518B - Media server and media service method

Info

Publication number: CN107483518B
Application number: CN201610399465.8A
Authority: CN
Inventors: 梅君君; 杨勇; 王斌
Original assignee: ZTE Corp
Current assignee: ZTE Corp
Priority date: 2016-06-07
Filing date: 2016-06-07
Publication date: 2022-08-02
Anticipated expiration: 2036-06-07
Also published as: CN107483518A; WO2017211165A1

Abstract

A media server and media service method, the said media server is realized and adopted the distributed architecture based on software that runs on the general hardware platform, including media resource control function MRFC module and media resource processing function MRFP module, connect with one or more MRFP modules under an MRFC module, wherein: the MRFC module is used for selecting an MRFP module for a new call after receiving the new call and informing the selected MRFP module to perform media processing; the MRFP module is used for processing the media of the new call. The media server is realized based on software running on a general hardware platform, does not depend on special hardware, provides basic guarantee for supporting virtualization, and reduces the cost of the media server compared with the traditional special hardware media server.

Description

Media server and media service method

Technical Field

The present invention relates to network technologies, and in particular, to a media server and a media service method.

Background

The media server is an important device in the field of telecommunications, and can be deployed in a Next Generation Network (NGN)/IP Multimedia Subsystem (IMS) to provide audio and video basic capability, as shown in fig. 1. Under the control of an Application Server (AS), through Session Initiation Protocol (SIP) and Media Session Markup Language (MSML)/Media Object Markup Language (MOML) interaction, users are provided with colorful audio and video services, such AS playing voice, playing video, receiving number, recording audio and video, receiving/sending fax, ASR (Automatic Speech Recognition), TTS (Text To Speech ), audio and video conference, and the like. The core processing of the media server is audio and video coding and decoding, and the performance consumption is very high, so that the traditional media server uses special hardware equipment and is configured with a special DSP (Digital Signal processing) chip to perform audio and video coding and decoding. The traditional media server uses special hardware, so the purchase, maintenance and capacity expansion costs are high, and the average utilization rate of resources is not high.

With the development of cloud computing technology, a telecommunication network develops towards features such as virtualization, distribution, flexible resource expansion and the like, so that the resource utilization rate of the network is effectively improved, and the Capital Expenditure)/OPEX (Operating cost) of a telecommunication operator is reduced. As an important device in a telecommunication network, a traditional media server does not support the capabilities of virtualization, resource elastic expansion and the like due to the limitation of dedicated hardware thereof, and thus does not adapt to the requirements of a cloud computing architecture on telecommunication devices.

Disclosure of Invention

In view of this, the present invention provides the following.

A media server, which is implemented based on software running on a general hardware platform and adopts a distributed architecture, and includes a media Resource control Function (MRFC) module and a media Resource processing Function (MRFP) module, one or more MRFP modules being connected to a lower portion of the MRFC module, wherein:

the MRFC module is used for selecting an MRFP module for a new call after receiving the new call and informing the selected MRFP module to perform media processing;

the MRFP module is used for processing the media of the new call.

A media service method is applied to a media server which is deployed on a general hardware platform and is realized based on a distributed architecture, the media server comprises a media resource control function (MRFC) module and a media resource processing function (MRFP) module, one or more MRFP modules are connected under one MRFC module, and the method comprises the following call processing procedures:

after receiving a new call, the MRFC module selects an MRFP module for the new call and informs the selected MRFP module to perform media processing;

and after receiving the notification, the selected MRFP module performs media processing on the new call.

The scheme has at least one of the following technical effects:

the media server is realized based on software running on a general hardware platform, does not depend on special hardware, provides basic guarantee for supporting virtualization, and reduces the cost of the media server compared with the traditional special hardware media server;

by adopting a distributed architecture, the internal load sharing and high reliability of the system can be supported, and the high reliability and load balance of the system are ensured;

a two-stage distributed flat architecture is adopted, so that horizontal capacity expansion is facilitated;

the method supports virtualization deployment, has the characteristic of elastic resource expansion under the cloud computing architecture, can greatly improve the use efficiency of hardware resources, and reduces the operation cost.

Drawings

FIG. 1 is a schematic diagram of the location of media servers in an NGN/IMS network;

FIG. 2 is a schematic structural diagram of a media server according to an embodiment of the present invention;

figure 3 is a schematic diagram of a distribution of a new access call by SIPPROXY according to an embodiment of the present invention;

FIG. 4 is a diagram illustrating a distributed communication mechanism between the MRFC module and the MRFP module according to an embodiment of the present invention;

FIG. 5 is a schematic diagram of resource elastic expansion according to an embodiment of the present invention;

FIG. 6 is a schematic diagram of a MRFC/MRFP combination, direct docking of an AS in an example of the present invention;

FIG. 7 is a schematic diagram of an example of an embodiment of the present invention in which MRFC/MRFP is combined and AS is interfaced via SIPPROXY;

FIG. 8 is a diagram of an example of the present invention in which MRFC is deployed separately from MRFP, and MRFC directly interfaces with AS;

FIG. 9 is a schematic diagram of an example of the present invention in which MRFC is deployed separately from MRFP, and MRFC interfaces with AS through SIPPROXY;

FIG. 10 is a flow chart of a media serving method according to an embodiment of the invention;

FIG. 11 is a flow chart of an exemplary MRF device expansion of the present invention;

FIG. 12 is a flow chart of an exemplary MRFC device expansion of the present invention;

FIG. 13 is a flowchart of an exemplary MRFP device expansion of the present invention;

FIG. 14 is a flow chart of an exemplary MRF device scaling of the present invention;

FIG. 15 is a flow chart of an exemplary MRFC device capacity reduction of the present invention;

FIG. 16 is a flow chart of an exemplary MRFP device scaling of the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be noted that the embodiments and features of the embodiments in the present application may be arbitrarily combined with each other without conflict.

An embodiment of the present invention provides a media server, where the media server is implemented based on software running on a general hardware platform and adopts a distributed architecture, as shown in fig. 2, the media server includes a media resource control function MRFC module and a media resource processing function MRFP module, one or more MRFP modules are connected under one MRFC module, where:

the MRFP module is used for processing the media of the new call.

In this embodiment, for the distribution of the new call, one of the following two ways may be adopted:

the first method is as follows:

and distributing through the AS. In this way, the MRFC module has an interface with the application server AS, and receives a new call distributed by the AS to the MRFC module through the interface with the AS.

The second method comprises the following steps:

distributed by the distribution agent module. In this way, the media server further includes a distribution agent module, and one or more MRFC modules are connected to one distribution agent module. In fig. 2 and 3, distribution proxy (SIPPROXY) of SIP protocol is taken as an example, and fig. 3 shows a schematic diagram of the distribution of SIPPROXY for a new access call.

At this time, the distribution agent module is used for distributing a new call incoming by the application server AS to the MRFC module according to the configured distribution strategy; and the MRFC module is provided with an interface with the distribution agent module, and receives a new call distributed to the MRFC module by the distribution agent module through the interface with the distribution agent module.

In this embodiment, the interface protocol between the modules may adopt the following manner:

the MRFC module and the AS or the distribution agent module can adopt a standard interface and interact through a Session Initiation Protocol (SIP); the MRFC module and the MRFP module can interact through internal instructions; at this time, after receiving a new call, the MRFC module is further configured to parse an SIP signaling of the new call, and send call information in the SIP signaling to the selected MRFP module through an internal instruction.

In this embodiment, the MRFC module and the MRFP module may be deployed in one of the following manners:

the first method is as follows: an MRFC module and an MRFP module connected below the MRFC module are jointly disposed on the same physical or virtual machine, and are referred to as MRF (Media Resource Function) devices herein, as shown in fig. 6 and 7;

the second method comprises the following steps: one MRFC module and one or more MRFP modules connected thereunder are separately configured and deployed on different physical or virtual machines, as can be seen in fig. 8 and 9.

In this embodiment, the MRFC module may monitor the MRFP module in real time, and select the MRFP module based on the state of the MRFP module, specifically:

the MRFC module is further configured to:

performing heartbeat detection on the MRFP module connected below the MRFC module to determine the survival state of the MRFP module connected below the MRFC module; and interacting with the MRFP module connected below the MRFC module to acquire the current resource occupation information of the MRFP module connected below the MRFC module.

The MRFC module selects an MRFP module for the new call, including: and selecting one MRFP module which is in a survival state and has the minimum current resource occupation from the MRFP modules connected under the MRFC module.

In this embodiment, the media server may further include a Distributed Communication Function (DCF) module, configured to implement a Distributed communication mechanism between the MRFC module and the MRFP module, as shown in fig. 4.

The MRFC module is also used for registering to the DCF module when starting, sending a query request for an MRFP module connected under the MRFC module to the DCF module, and establishing a distributed communication link with the searched MRFP module according to a search result;

the MRFP module is also used for registering with the DCF module when starting;

the DCF module is used for recording the registered MRFC module and MRFP module and the connection relation between the MRFC module and the MRFP module; and after receiving the query request, searching the information of the MRFP module connected below the MRFC module initiating the query, and returning the search result to the MRFC module initiating the query.

In this embodiment, the MRFP module is further configured to log out to the DCF module when the MRFP module is offline; the DCF module is further configured to release a connection relationship between the MRFP module initiating logout and the MRFC module; the MRFC module is also used for sending the query request to the DCF module at regular time and updating the information of the MRFP module connected below the MRFC module according to the query result.

In an example, the association relationship between the MRFC module and the MRFP module may be managed in a cluster manner, for example, one MRFC module and all MRFP modules connected thereunder form a cluster; when registering to the DCF module, the MRFC module and the MRFP module both carry cluster identifiers of the clusters to which the MRFC module belongs, and the query request sent by the MRFC module carries the cluster identifiers of the clusters to which the MRFC module belongs. The MRFC modules with the same cluster identifications and the MRFP modules have connection relations. The DCF module is further configured to record cluster identifiers of clusters to which the MRFC module and the MRFP module belong; and when the DCF module searches for the MRFP module connected below the MRFC module initiating the query, recording the MRFP module of which the cluster identifier of the cluster is the same as the cluster identifier carried in the query request as the MRFP module connected below the MRFC module initiating the query.

The DCF module, the MRFC module and the MRFP module can interact with each other through a Representational State Transfer (REST) interface.

In this embodiment, the media server may further include a Virtual Network Function Manager (VNFM) module, as shown in fig. 2. Fig. 5 is a schematic diagram of a resource elastic scaling mechanism.

The MRFC module and the MRFP module are also used for reporting statistical parameters to the VNFM module during operation;

the VNFM module is used for analyzing the reported statistical parameters and carrying out the following processing: and carrying out capacity expansion processing when the set capacity expansion condition is met, and/or carrying out capacity reduction processing when the set capacity reduction condition is met.

Wherein:

the capacity expansion processing performed by the VNFM module may include one of the following processing:

when the media resource function MRF equipment needing capacity expansion is determined, a new virtual machine is created, the newly expanded MRF equipment is deployed on the virtual machine, the newly expanded MRF equipment is started after program installation and parameter configuration are completed, and information of the newly expanded MRF equipment is notified to a distribution agent module or an AS; the MRF device refers to a device where the MRFC module and the MRFP module connected below the MRFC module are arranged in a combined mode;

when determining that the MRFC module which is separately deployed from the MRFP module needs to be expanded, simultaneously expanding the MRFP module connected below the MRFC module, creating a new virtual machine, deploying the newly expanded MRFC module and the MRFP module on the virtual machine, starting the newly expanded MRFC module and the MRFP module after completing program installation and parameter configuration, and notifying a distribution agent module or an AS of information of the newly expanded MRFC module;

when the MRFP module connected with the existing MRFC module needs to be expanded, a new virtual machine is created, the MRFP module needing to be expanded is deployed on the virtual machine, the MRFP module needing to be expanded is started after program installation and parameter configuration are completed, and the MRFP module needing to be expanded is used as the MRFP module connected with the existing MRFC module.

The VNFM module performs a capacity reduction process, which may include one of the following processes:

when determining that the capacity reduction MRF equipment is needed, notifying the capacity reduced MRF equipment to be offline, recovering corresponding virtual machine resources after receiving the notification that the capacity reduction MRF equipment is cleaned, and notifying a distribution agent module or an AS of the information of the capacity reduced MRF equipment;

when determining an MRFC module which needs to be scaled down and is separately deployed from an MRFP module, notifying the scaled-down MRFC module and an MRFP module connected below the scaled-down MRFC module to be offline, after receiving a notification of the completion of cleaning the scaled-down MRFC module and the MRFP module, recovering corresponding virtual machine resources, and notifying a distribution agent module or an AS (application server) of information of the scaled-down MRFC module;

when determining that the MRFP module connected with the existing MRFC module needs to be reduced, informing the reduced MRFP module to be offline, and after receiving the notification that the cleaning of the reduced MRFP module is finished, recycling the corresponding virtual machine resources.

An embodiment of the present invention further provides a media service method, which is applied to a media server, where the media server is implemented based on software running on a general hardware platform and adopts a distributed architecture, and includes an MRFC module and an MRFP module, and one or more MRFP modules are connected under one MRFC module, as shown in fig. 10, the method includes the following call processing procedures:

step 110, after receiving a new call, the MRFC module selects an MRFP module for the new call and notifies the selected MRFP module to perform media processing;

if the call is distributed through the AS, the MRFC module receives a new call distributed to the MRFC module by the AS through the interface between the MRFC module and the AS.

If the media server comprises the distribution agent module, the media server distributes the media through the distribution agent module, and one or more MRFC modules are connected below one distribution agent module; before the MRFC module receives a new call, the distribution agent module distributes the new call incoming by the AS to the MRFC module according to the configured distribution strategy; and the MRFC module receives the new call distributed to the MRFC module by the distribution agent module through the interface between the MRFC module and the distribution agent module.

Optionally, the MRFC module interacts with the AS or the distribution agent module through a session initiation protocol SIP; the MRFC module and the MRFP module interact through internal instructions; after receiving the new call, the MRFC module further includes: and analyzing the SIP signaling of the new call, and sending the call information in the SIP signaling to the selected MRFP module through an internal instruction.

Optionally, one MRFC module and an MRFP module connected thereunder are jointly configured and deployed on the same physical or virtual machine; alternatively, one MRFC module and one or more MRFP modules connected thereunder are separately configured and deployed on different physical or virtual machines.

Step 120, after receiving the notification, the selected MRFP module performs media processing on the new call.

In this embodiment, the following status detection and information acquisition processes are further included:

the MRFC module performs heartbeat detection on the MRFP module connected below the MRFC module to determine the survival state of the MRFP module connected below the MRFC module;

and the MRFC module interacts with the MRFP module connected below the MRFC module to acquire the current resource occupation information of the MRFP module connected below the MRFC module.

Based on the state detection and information acquisition, when the MRFC module selects an MRFP module for the new call, an MRFP module that is in a survival state and has the smallest current resource occupation may be selected from MRFP modules connected to the MRFC module.

In this embodiment, the media server further includes a distributed communication function DCF module, and the method further includes the following registration process:

the MRFC module and the MRFP module are registered with the DCF module when being started, and the DCF module records the registered MRFC module and MRFP module and the connection relation between the MRFC module and the MRFP module; the DCF module may obtain the connection relationship information between the MRFC module and the MRFP module from the registration message of the MRFC module and the MRFP module, or obtain the connection relationship information according to the configuration information or obtain the connection relationship information from other internal or external modules, which is not limited in the present invention.

The method further comprises the following query process:

when the MRFC module is started, sending a query request for an MRFP module connected with the MRFC module to the DCF module; after receiving the query request, the DCF module searches the information of the MRFP module connected below the MRFC module initiating the query, and returns the search result to the MRFC module initiating the query; and the MRFC module establishes a distributed communication link with the searched MRFP module according to the search result.

In this embodiment, the method further includes the following logout procedure: the MRFP module logs out to the DCF module when the MRFP module is offline, and the DCF module releases the connection relation between the MRFP module initiating logging out and the MRFC module;

the query process further comprises: the MRFC module sends the query request to the DCF module at regular time; after receiving the query request, the DCF module searches the information of the MRFP module connected below the MRFC module initiating the query, and returns the search result to the MRFC module initiating the query; and the MRFC module updates the information of the MRFP module connected below the MRFC module according to the query result.

In one example, one MRFC module forms one cluster with all MRFP modules connected thereunder; in the registration process, when the MRFC module and the MRFP module register to the DCF module, the MRFC module and the MRFP module both carry cluster identifiers of the belonged clusters; the DCF module also records cluster identifiers of clusters to which the MRFC module and the MRFP module belong; in the query process, the query request sent by the MRFC module carries the cluster identifier of the cluster to which the MRFC module belongs; the DCF module searches for the MRFP module connected below the MRFC module initiating the query, and the method comprises the following steps: and recording the MRFP module with the same cluster identifier of the cluster to which the cluster belongs and the MRFP module carried in the query request as the MRFP module connected below the MRFC module initiating the query.

In this embodiment, the DCF module, the MRFC module, and the MRFP module communicate REST interface interactions through a representational state transfer.

In this embodiment, the media server further includes a virtual network function management VNFM module; the method further comprises the following capacity adjustment procedure:

the MRFC module and the MRFP module report statistical parameters to the VNFM module during operation;

Wherein the content of the first and second substances,

the VNFM module carries out capacity expansion processing and comprises one of the following processing:

when determining that the MRFP module connected under the existing MRFC module needs to be expanded, creating a new virtual machine, deploying the MRFP module needing to be expanded on the virtual machine, starting the MRFP module newly expanded after program installation and parameter configuration are completed, and taking the MRFP module newly expanded as the MRFP module connected under the existing MRFC module.

The VNFM module performs capacity reduction processing and comprises one of the following processing:

when determining an MRFC module which needs to be separately deployed with an MRFP module, notifying the MRFC module to be reduced and an MRFP module connected below the MRFC module to be reduced to be offline, after receiving the notification that the cleaning of the MRFC module to be reduced and the MRFP module is finished, recovering corresponding virtual machine resources, and notifying a distribution agent module or an AS of the information of the MRFC module to be reduced;

In an example, the VNFM module interacts with the MRFC module and the MRFP module through a representational state transfer REST interface.

The following description is provided in connection with some examples of specific applications.

In this example, each component of the media server is a software program running on a general hardware platform of x86 (an instruction set introduced by Intel) architecture, and does not depend on a dedicated hardware such as a DSP chip, which is also referred to herein as a software media server. The location of the software media server in the network is shown in fig. 1, the software media server is a device providing media services in the NGN/IMS network, provides various media functions under the control of the AS, interacts with the AS through SIP signaling, and interacts with a media gateway in the core network through RTP (Real-time Transport Protocol).

FIG. 2 is a system architecture diagram of the present example software media server. As shown, the system includes a SIPPROXY module, an MRFC module, an MRFP module, a DCF module, and a VNFM module inside. Wherein:

the SIPROXY module is used for distributing a new call accessed to the AS to a specific MRFC according to a configured distribution strategy, and the SIPROXY module and the AS interact through SIP signaling;

the SIPPROXY module is a call access distribution agent module of the media server. Under a large-capacity architecture, a plurality of MRFCs are provided, and SIPROXY can distribute new calls incoming from AS to MRFC modules according to a configured distribution strategy; at small capacity, sip proxy may not be used and AS directly accesses the call onto the MRFC module.

And the MRFC selects a proper MRFP to process the media of the new call according to the configured strategy, and the MRFP interact through an internal interface.

The MRFC module is a signaling processing module of the media server, interacts with the AS/SIPROXY through SIP signaling, maps the call information in the SIP signaling to an internal instruction and sends the internal instruction to the MRFP module. One MRFC module may be followed by one or more MRFP modules.

The MRFP module is a media processing module of the media server and realizes the functions of media packet receiving and sending, audio and video coding and decoding, audio conference mixing, video conference synthesis and the like.

The MRFC module and the MRFP module can be combined or separately deployed.

When the MRFC module is combined with the MRFP module and directly interfaces with the AS, AS shown in fig. 6, the MRFC module and the MRFP module are deployed on the same physical or virtual machine, which is called AS an MRF (Media Resource Function) device, and the AS itself has a distribution Function and can distribute a call to different MRF devices.

When the MRFC module and the MRFP module are combined and are docked with the AS through the SIPPROXY module, AS shown in fig. 7, the MRFC module and the MRFP module are deployed on the same physical or virtual machine, called MRF equipment, the AS is distributed through the SIPPROXY module, and the SIPPROXY module distributes the call to different MRF equipment according to the configured distribution policy.

When the MRFC module and the MRFP module are separately deployed and the MRFC module and the AS are directly docked, AS shown in fig. 8, the MRFC module and the MRFP module are separately deployed on a physical or virtual machine, and the AS itself has a distribution function and can distribute a call to different MRFC devices; and the MRFC selects a proper MRFP device to perform media processing according to a certain strategy.

When the MRFC module and the MRFP module are separately deployed and the MRFC module is docked with the AS through SIPPROXY, AS shown in fig. 9, the MRFC module and the MRFP module are separately deployed on a physical or virtual machine, and are distributed through SIPPROXY, and the SIPPROXY distributes a call to different MRFC devices according to a configured distribution policy; and the MRFC selects proper MRFP equipment to process media according to a certain strategy.

In another example, in the software media server, a part of the MRFC module may be deployed separately from the MRFP module, and a part of the MRFC module is co-located with the MRFP module.

The DCF module is a service node management module of the media server and is used for performing distributed service node management on the MRFC module and the MRFP module, and the DCF module and the MRFC module and the MRFP module are interacted through a Representational State Transfer (REST) interface; MRFC and MRFP are respectively registered on DCF as service, DCF assigns node numbers to the registered MRFC and MRFP, and detects the survival state of each node. The MRFC establishes an internal distributed communication link between the MRFC and the MRFP before the surviving MRFP by inquiring the MRFP from the DCF.

The VNFM module is a resource elastic expansion management module of the media server and is used for dynamically expanding and contracting the MRFC module and/or the MRFP module according to a configured strategy, and the VNFM module, the MRFC module and the MRFP module are interacted through REST interfaces.

Referring to fig. 3, the MRFC/MRFP distributed communication mechanism in this example includes:

the method comprises the following steps: the MRFC module registers with the DCF at startup with a "cluster name" + "MRFC. Different MRFCs carry different cluster names, which represent different clusters;

step two: the MRFP module will register with the DCF at startup with the registration name "cluster name" + "MRFP". All MRFPs connected to the same MRFC carry the same cluster name as the cluster name of the MRFC to which the cluster name belongs during registration;

step three: after MRFC registration is completed, the DCF is timed to query MRFP, and the query name is 'cluster name' + 'MRFP', so that MRFC can query all MRFP addresses connected under the MRFC, and an internal distributed communication link is established with each MRFP. The same MRFC and all MRFPs below the MRFC are called a cluster;

step four: the MRFC performs heartbeat detection on all MRFP in the cluster at regular time through the internal link to judge whether the MRFP is in a survival state; meanwhile, the MRFC sends query requests to all MRFPs under the cluster at regular time through the internal link, and the MRFP returns the current resource occupation condition of the module;

step five: when the new call reaches MRFC, MRFC will select MRFP with minimum current resource occupation and in survival state according to the survival status of each MRFP in the cluster and the resource occupation status of MRFP, and communicate with the selected MRFP through internal link, thus ensuring the reliability and load balance of the system.

Referring to fig. 4, the distribution process of SIPPROXY under high capacity for accessing new calls includes:

the method comprises the following steps: configuring distribution strategies on the SIPPXY (namely SIPPXY module), wherein each strategy represents that the current new incoming call is distributed to which MRFC (namely MRFC module) under a certain condition;

step two: the AS sends the new call to the SIPROXY, and the AS and the SIPROXY interact through SIP signaling;

step three: SIPPROXY distributes the current call to a certain MRFC according to the distribution strategy. SIP proxy and MRFC interact through SIP signaling.

The heartbeat detection is carried out between the SIPPXY and each MRFC, so that the SIPPXY can acquire the survival state of each MRFC, when a new call enters, the SIPPXY selects one MRFC in the survival state to distribute according to a distribution strategy, and the reliability and the load balance of the system are ensured.

When the system capacity is small or the AS has a distribution mechanism, the AS is directly connected with the MRFC without using the SIPPXY and interacts through SIP signaling.

Through the above technology, the software media server system has two layers of distributed architecture: the MRFC can be multiple, is horizontally expanded, and distributes the newly accessed call through SIPPROXY or directly through AS; there may be several MRFPs (i.e. MRFP modules) under the same MRFC, and the MRFC may horizontally expand, and select a suitable MRFP to process the newly accessed call according to a certain policy. The system considers high reliability and load balance when selecting MRFC and MRFP; meanwhile, the distributed and flat system architecture is very suitable for elastic resource expansion and contraction under the cloud computing architecture.

Referring to fig. 5, the resource elastic scaling process of the present example includes:

the method comprises the following steps: MRFC and MRFP programs are deployed in a Virtual machine environment and used as Network Function Virtualization (NFV) Network elements, and some statistical parameters such as CPU occupancy rate, memory occupancy rate, current online telephone traffic number and the like are reported to VNFM (Virtual Network Function Management) Network elements at regular time in an operation session;

step two: and the VNFM analyzes the reported statistical parameters and judges according to a resource expansion strategy. And if the capacity expansion conditions are met, for example, the parameters exceed the configured threshold values, performing capacity expansion according to a capacity expansion strategy. The VNFM will create a new virtual machine, automatically deploy the MRFC or MRFP program to be expanded on the newly expanded virtual machine, and start the MRFC or MRFP program after automatically configuring the relevant parameters. When the MRFC is expanded, the MRFP is expanded simultaneously, and a new cluster is formed on the newly expanded MRFP and the newly expanded MRFC; if only MRFP is expanded, the expanded MRFP is added to the existing cluster;

step three: and the VNFM analyzes the reported statistical parameters and judges according to a resource expansion strategy. And if the capacity reduction condition is met, for example, the parameters are lower than the configured threshold value, carrying out capacity reduction according to a capacity reduction strategy. The VNFM sends an indication to the MRFC or MRFP on the virtual machine needing capacity reduction, the MRFC or MRFP cleans up related resources, then the MRFC or MRFP program is stopped, and finally the virtual machine resources are recycled. When the MRFC is reduced, the MRFP of the cluster is also reduced.

Through the elastic expansion of resources, the number of MRFC and MRFP processing modules of the software media server can dynamically expand and contract the resources according to the current real-time telephone traffic requirement, thereby greatly improving the service efficiency of the resources and reducing the overall cost of the software media server.

Referring to fig. 6, the call processing procedure when the MRF directly interfaces with the AS includes:

the method comprises the following steps: the AS service logic judges that a software media server needs to be called;

step two: and when the AS discovers a plurality of MRFs, selecting a proper MRF device according to the distribution strategy of the AS. Because the heartbeat detection exists between the AS and the MRF, the current broken MRF equipment can be found, so the MRF selected by the AS is the MRF which can normally work; if all MRF devices are broken, the AS can not find the proper MRF device, can not initiate a new call of the MRF, and the process is ended;

step three: after the AS sends the new call to the MRFC module on the selected MRF equipment, the MRFC checks the MRFP under the cluster, and the next cluster in the networking has only one MRFP;

step four: the MRFC determines whether the MRFP is able to provide services, such as whether the MRFP is alive or not, and whether the processing capability of the MRFP is sufficient. If MRFP meets the requirement, MRFC informs MRFP to process through internal interface, otherwise MRFC returns failure to AS.

Referring to fig. 7, the call processing procedure when the MRF interfaces the AS through sip proxy includes:

step two: the AS is butted with the SIPPXY of the software media server and sends the new call to the SIPPXY;

step three: SIPPROXY distributes the new call to an appropriate MRF device according to the configured distribution strategy. Because the heartbeat detection exists between the SIPPROXY and the MRF, the MRF selected by the SIPPROXY can be the MRF which can normally work; if all MRF devices are broken, the SIPROXY can not find the proper MRF device, discards the current new call message, and the flow is ended;

step four: after SIPROXY sends the new call to MRFC module on the selected MRF device, MRFC looks up MRFP under the cluster, and the next cluster in the networking has only one MRFP;

step five: the MRFC determines whether the MRFP is able to provide services, such as whether the MRFP is alive or not, and whether the processing capability of the MRFP is sufficient. If MRFP meets the requirement, MRFC informs MRFP to process through internal interface, otherwise MRFC returns failure to SIPROXY, SIPROXY sends the failure response to AS.

Referring to fig. 8, MRFC/MRFP are separately deployed, and the call processing procedure when directly interfacing AS includes:

step two: and when the AS discovers a plurality of MRFC, selecting a proper MRFC device according to the distribution strategy of the AS. Because the heartbeat detection exists between the AS and the MRCF, the MRFC equipment which is currently broken can be found, so the MRFC selected by the AS is the MRFC which can normally work; if all MRFC devices are broken, the AS cannot find the proper MRFC device, cannot initiate a new call to the MRFC, and the process is ended;

step three: the AS sends the new call to the selected MRFC equipment, the MRFC checks the MRFP under the cluster, and the next cluster of the network has a plurality of MRFP equipment;

step four: the MRFC selects a surviving MRFP from under the cluster with the smallest current load. If the MRFP is found, the MRFC informs the MRFP to process through an internal interface, otherwise, the MRFC returns failure to the AS.

Referring to fig. 9, MRFC/MRFP are separately deployed, and the call processing procedure when the AS is docked by SIPPROXY includes:

step three: SIPPROXY distributes the new call to an appropriate MRFC device according to the configured distribution strategy. Because there is heartbeat detection between SIPROXY and MRFC, can find the MRFC equipment that is broken at present, so the MRFC that SIPROXY chooses is MRFC that can work normally; if all MRFC devices are broken, the SIP proxy can not find the proper MRFC device, discards the current new call message and ends the process;

step four: SIPROXY sends the new call to the selected MRFC equipment, MRFC looks up MRFP under the cluster, and the next cluster in the network has a plurality of MRFP equipment;

step five: the MRFC selects a surviving MRFP from under the cluster with the smallest current load. If the MRFP is found, the MRFC informs the MRFP to process through an internal interface, otherwise, the MRFC returns failure to the SIPROXY, and the SIPROXY sends the failure response to the AS.

Under the condition, MRFC and MRFP programs are separately deployed on a physical or virtual machine and distributed through SIPROXY, and the SIPROXY distributes the call to different MRFC equipment according to the configured distribution strategy; and the MRFC selects proper MRFP equipment to perform media processing according to the configured strategy.

Referring to fig. 11, the capacity expansion process of the MRF device of this example includes:

the MRF device actively reports attribute parameters (such as cluster names and the like) and statistical parameters (such as CPU occupancy rate, memory occupancy rate, current session number and the like) to the VNFM device;

the VNFM equipment analyzes the statistical parameters reported by the MRF equipment, judges that the system load exceeds a configured threshold value according to a telescopic strategy, and needs to expand the MRF equipment to construct a new cluster;

the VNFM equipment creates a new virtual machine;

the VNFM equipment installs an MRF program (MRFC + MRFP) on the newly created virtual machine, completes configuration and starts the MRF program;

if the system has SIPROXY, the VNFM informs the SIPROXY of the newly expanded MRF equipment information, and the SIPROXY adds the SIPROXY into the distribution object; if the system does not have SIPROXY, the VNFM informs the AS of the newly expanded MRF equipment information, and the AS adds the newly expanded MRF equipment information into a distribution object;

after starting MRFC/MRFP program on MRF, MRFC registers to DCF, MRFC inquires MRFP from DCF, and establishes internal communication link between MRFC and MRFP.

Referring to fig. 12, the capacity expansion flow packet of the MRFC device of this example includes:

the MRFC device (the device where the MRFC is located when the MRFC and MRFP are separately deployed) will actively report attribute parameters (such as cluster name, etc.), statistical parameters (such as CPU occupancy, memory occupancy, current session number, etc.) to the VNFM device;

and the VNFM equipment analyzes the statistical parameters reported by the MRFC equipment, judges that the system load exceeds a configured threshold value according to a telescopic strategy and needs to expand the MRFC equipment. When the MRFC equipment is expanded, the MRFP equipment under the MRFC equipment is newly built to form a new cluster;

the VNFM equipment creates a new virtual machine for the expanded MRFC and MRFP equipment;

and the VNFM equipment installs the MRFC program on the newly expanded MRFC equipment and installs the MRFP program on the newly expanded MRFP equipment. Completing the relevant configuration, starting MRFC and MRFP programs and constructing a new cluster;

if the system has SIPROXY, the VNFM informs the SIPROXY of the newly expanded MRFC equipment information, and the SIPROXY adds the SIPROXY into the distribution object; if the system does not have SIPROXY, the VNFM informs the AS of the newly expanded MRFC equipment information, and the AS adds the newly expanded MRFC equipment information into a distribution object;

after starting the new expanded MRFC and MRFP programs, the MRFC registers to the DCF, and the MRFC inquires the MRFP from the DCF to establish an internal communication link between the MRFC equipment and the MRFP equipment.

Referring to fig. 13, the capacity expansion process of the MRFP device of this example includes:

the MRFP equipment (equipment where the MRFP is located when the MRFC and the MRFP are separately deployed) actively reports attribute parameters (such as cluster names and the like) and statistical parameters (such as CPU occupancy rate, memory occupancy rate, current session number and the like) to the VNFM equipment;

the VNFM equipment judges that MRFP load under a certain cluster exceeds a configured threshold value according to a telescopic strategy, and MRFP equipment under the cluster needs to be expanded;

the VNFM equipment creates a new virtual machine for the newly expanded MRFP equipment;

and the VNFM equipment installs MRFP programs on the newly expanded MRFP equipment. Completing the relevant configuration, starting an MRFP program, and attributing the MRFP equipment to the current cluster;

registering the newly expanded MRFP program to the DCF after starting;

the MRFC of the cluster is set to inquire the DCF at regular time to acquire the newly expanded MRFP equipment and establish an internal communication link between the MRFC and the MRFP equipment.

Referring to fig. 14, the capacity reduction process of the MRF apparatus of this example includes:

the MRF device can actively report attribute parameters (such as cluster names and the like) and statistical parameters (such as CPU occupancy rate, memory occupancy rate, current session number and the like) to the VNFM device;

the VNFM equipment judges that the MRF load is lower than a configured threshold value according to a telescopic strategy and needs to be reduced and held;

the VNFM informs the contracted MRF equipment of off-line;

the offline MRF performs resource cleaning work, notifies the VNFM of the completion of cleaning, and then stops the MRFC and MRFP programs thereon;

the VNFM recovers virtual machine resources of the off-line MRF equipment;

if the system has the SIPROXY, the VNFM informs the MRF equipment information of the SIPROXY which is off-line, and the SIPROXY removes the SIPROXY from the distribution object; if there is no SIPPROXY in the system, the VNFM informs the AS of the offline MRF device information, and the AS removes the information from the distribution object.

Referring to fig. 15, the capacity reduction process of the MRFC apparatus of this example includes:

the MRFC device will actively report attribute parameters (such as cluster name, etc.) and statistical parameters (such as CPU occupancy, memory occupancy, current session number, etc.) to the VNFM device;

and the VNFM equipment judges that the MRFC load is lower than a configured threshold value according to a telescopic strategy and needs to reduce the MRFC equipment. When the MRFC equipment is reduced, the MRFP equipment under the MRFC equipment is dropped, namely the cluster is removed;

the VNFM informs the MRFC and MRFP devices which are subjected to capacity reduction to be offline;

the offline MRFC performs resource cleaning work, notifies the VNFM of the completion of cleaning, and then stops the MRFC program on the VNFM; the offline MRFP does resource cleaning work, informs the VNFM of the completion of cleaning, and then stops the MRFP program on the VNFM;

the VNFM recovers virtual machine resources of the off-line MRFC and MRFP equipment;

if the system has SIPROXY, the VNFM informs MRFC equipment information of the SIPROXY which is off-line, and the SIPROXY removes the SIPROXY from the distribution object; if there is no SIPPROXY in the system, the VNFM informs the AS of the offline MRFC device information, and the AS removes the information from the distribution object.

Referring to fig. 16, the capacity reduction process of the MRFP device of this example includes:

the MRFP equipment can actively report attribute parameters (such as cluster names and the like) and statistical parameters (such as CPU occupancy rate, memory occupancy rate, current session number and the like) to the VNFM equipment;

the VNFM equipment judges that MRFP load under a certain cluster is lower than a configuration threshold value according to a telescopic strategy and needs to reduce capacity of the MRFP equipment;

the VNFM informs the contracted MRFP equipment of off-line;

the offline MRFP does resource cleaning work, informs the VNFM of the completion of cleaning, and then stops the MRFP program on the VNFM;

the VNFM recovers virtual machine resources of the off-line MRFP equipment;

the MRFP which is offline also logs out to the DCF module, the DCF module releases the connection relation between the MRFP module initiating logging out and the MRFC module, and the MRFC in the cluster can know that the existing MRFP equipment in the cluster is offline by inquiring the MRFP from the DCF at regular time, so that the MRFP equipment is removed from a distribution object (the MRFP which can distribute new calls).

The above-mentioned serial numbers of the embodiments of the present invention are merely for description and do not represent the merits of the embodiments. Through the above description of the embodiments, those skilled in the art will clearly understand that the method of the above embodiments can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware, but in many cases, the former is a better implementation manner. Based on such understanding, the technical solutions of the embodiments of the present invention may be embodied in the form of a software product, which is stored in a storage medium (e.g., ROM/RAM, magnetic disk, optical disk) and includes instructions for enabling a terminal device (e.g., a mobile phone, a computer, a server, or a network device) to execute the method according to the embodiments of the present invention.

The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A media server is realized based on software running on a general hardware platform and adopts a distributed architecture, and comprises a media resource control function (MRFC) module, a media resource processing function (MRFP) module and a Virtual Network Function Management (VNFM) module, wherein one or more MRFP modules are connected below one MRFC module, and the media server is characterized in that:

the MRFP module is used for carrying out media processing on the new call;

the VNFM module is used for analyzing the reported statistical parameters and carrying out the following processing: carrying out capacity expansion processing on the media server when the set capacity expansion condition is met, and/or carrying out capacity reduction processing on the media server when the set capacity reduction condition is met;

wherein, the VNFM module carries out dilatation processing, including:

when determining that the MRFP module connected with the existing MRFC module needs to be expanded, creating a new virtual machine, deploying the MRFP module needing to be expanded on the virtual machine, starting the MRFP module newly expanded after program installation and parameter configuration are completed, and taking the MRFP module newly expanded as the MRFP module connected with the existing MRFC module;

2. The media server of claim 1, wherein:

the MRFC module is provided with an interface with an application server AS;

the MRFC module receives a new call, including: and receiving a new call distributed to the MRFC module by the AS through an interface between the MRFC module and the AS.

3. The media server of claim 1, wherein:

the media server also comprises a distribution agent module, and one or more MRFC modules are connected below one distribution agent module;

the distribution agent module is used for distributing a new call incoming by the application server AS to the MRFC module according to the configured distribution strategy;

the MRFC module is provided with an interface with the distribution agent module; the MRFC module receives a new call, including: and receiving a new call distributed to the MRFC module by the distribution agent module through an interface between the MRFC module and the distribution agent module.

4. A media server according to claim 2 or 3, wherein:

the MRFC module and the AS or the distribution agent module interact through a Session Initiation Protocol (SIP);

the MRFC module and the MRFP module interact through internal instructions;

after receiving the new call, the MRFC module is further configured to parse an SIP signaling of the new call, and send call information in the SIP signaling to the selected MRFP module through an internal instruction.

5. A media server according to claim 1, 2 or 3, wherein:

one MRFC module and an MRFP module connected below the MRFC module are jointly arranged and deployed on the same physical or virtual machine; or

One MRFC module and one or more MRFP modules connected below the MRFC module are arranged separately and are deployed on different physical or virtual machines.

6. A media server according to claim 1, 2 or 3, wherein:

the MRFC module is further configured to:

performing heartbeat detection on the MRFP module connected below the MRFC module to determine the survival state of the MRFP module connected below the MRFC module;

and interacting with the MRFP module connected below the MRFC module to acquire the current resource occupation information of the MRFP module connected below the MRFC module.

7. The media server of claim 6, wherein:

the MRFC module selects an MRFP module for the new call, including:

and selecting one MRFP module which is in a survival state and has the minimum current resource occupation from the MRFP modules connected under the MRFC module.

8. The media server of claim 1, wherein:

the media server also comprises a Distributed Communication Function (DCF) module;

the MRFP module is also used for registering with the DCF module when starting;

9. The media server of claim 8, wherein:

the MRFP module is also used for logging out to the DCF module when the DCF module is offline;

the DCF module is further configured to release a connection relationship between the MRFP module initiating logout and the MRFC module;

the MRFC module is also used for sending the query request to the DCF module at regular time and updating the information of the MRFP module connected below the MRFC module according to the query result.

10. The media server of claim 8, wherein:

one MRFC module and all the MRFP modules connected below the MRFC module form a cluster;

when the MRFC module and the MRFP module register to the DCF module, the MRFC module and the MRFP module both carry cluster identifications of the belonged clusters; the query request sent by the MRFC module carries the cluster identifier of the cluster to which the MRFC module belongs;

the DCF module is further configured to record cluster identifiers of clusters to which the MRFC module and the MRFP module belong; the DCF module searches for the MRFP module connected below the MRFC module initiating the query, and the method comprises the following steps: and recording the MRFP module with the same cluster identifier of the cluster to which the cluster belongs and the MRFP module carried in the query request as the MRFP module connected below the MRFC module initiating the query.

11. The media server of claim 8, wherein:

and the DCF module, the MRFC module and the MRFP module interact through a representational state transfer REST interface.

12. The media server of claim 1, wherein:

and the VNFM module, the MRFC module and the MRFP module interact through a representational state transfer REST interface.

13. A media service method is applied to a media server, the media server is realized based on software running on a general hardware platform and adopts a distributed architecture, the media server comprises a media resource control function (MRFC) module, a media resource processing function (MRFP) module and a Virtual Network Function Management (VNFM) module, one MRFC module is connected with one or more MRFP modules, and the method comprises the following call processing procedures:

after receiving the notification, the selected MRFP module performs media processing on the new call;

the method further comprises the following capacity adjustment procedure:

the VNFM module is used for analyzing the reported statistical parameters and carrying out the following processing: carrying out capacity expansion processing on the media server when set capacity expansion conditions are met, and/or carrying out capacity reduction processing on the media server when set capacity reduction conditions are met;

the VNFM module performs capacity expansion processing and comprises one of the following processing:

14. The method of claim 13, wherein:

the MRFC module is provided with an interface with an application server AS;

the MRFC module receives a new call, including: and receiving a new call distributed to the MRFC module by the AS through the interface between the MRFC module and the AS.

15. The method of claim 13, wherein:

before the MRFC module receives the new call, the method further includes: the distribution agent module distributes the new call incoming by the AS to the MRFC module according to the configured distribution strategy;

the MRFC module receives a new call, including: and receiving a new call distributed to the MRFC module by the distribution agent module through an interface between the MRFC module and the distribution agent module.

16. The method of claim 14 or 15, wherein:

the MRFC module and the MRFP module interact through internal instructions;

after receiving the new call, the MRFC module further includes: and analyzing the SIP signaling of the new call, and sending the call information in the SIP signaling to the selected MRFP module through an internal instruction.

17. The method of claim 13, 14 or 15, wherein:

18. The method of claim 13, 14 or 15, wherein:

the method further comprises the following status detection and information acquisition processes:

19. The method of claim 18, wherein:

20. The method of claim 13, wherein:

the method further comprises the following registration procedure:

the MRFC module and the MRFP module are registered with the DCF module when being started, and the DCF module records the registered MRFC module and MRFP module and the connection relation between the MRFC module and the MRFP module;

the method further comprises the following query process:

21. The method of claim 20, wherein:

the method further comprises the following deregistration procedure:

the MRFP module logs out to the DCF module when the MRFP module is offline, and the DCF module releases the connection relation between the MRFP module initiating logging out and the MRFC module;

the query process further comprises:

the MRFC module sends the query request to the DCF module at regular time; after receiving the query request, the DCF module searches the information of the MRFP module connected below the MRFC module initiating the query, and returns the search result to the MRFC module initiating the query; and the MRFC module updates the information of the MRFP module connected below the MRFC module according to the query result.

22. The method of claim 20, wherein:

in the registration process, when the MRFC module and the MRFP module register to the DCF module, the MRFC module and the MRFP module both carry cluster identifiers of the belonged clusters; the DCF module also records cluster identifiers of clusters to which the MRFC module and the MRFP module belong;

in the query process, the query request sent by the MRFC module carries the cluster identifier of the cluster to which the MRFC module belongs; the DCF module searches for the MRFP module connected below the MRFC module initiating the query, and the method comprises the following steps: and recording the MRFP module with the same cluster identifier of the cluster to which the cluster belongs and the MRFP module carried in the query request as the MRFP module connected below the MRFC module initiating the query.

23. The method of claim 20, wherein:

24. The method of claim 13, wherein: