JP4459480B2 - Network service operation system - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a network service operation system, and more particularly, to a network service operation system that guarantees the quality of service (QoS: Quality of Service).
[0002]
[Prior art]
In recent years, the Internet and a shared use network in which a plurality of companies / organizations are connected has become widespread. In this shared use network, a QoS guaranteed network is used based on a service level agreement (hereinafter referred to as SLA). Further, when a service using digital contents is performed on this network, the necessity and effectiveness of QoS guarantee are described in, for example, the paper “Design and Development of QoS enabled Content for Business Business” (Akira Shirase's IC: Etc. are described.
[0003]
Also, in the so-called IT era and e-commerce era that have come due to the rapid development of the information industry, it is used by a plurality of service providers (for example, Internet service provider systems) included in an Internet data center or a shared data center. Of the servers, there is a case where the access is suddenly concentrated on a specific server and the server or the network is congested. In order to avoid this congestion state, conventionally, the network and server hardware have been increased along with the renewal of the SLA contract.
[0004]
[Problems to be solved by the invention]
However, in response to this increase in network and server hardware, the user cannot fully access the server until the network or server is increased. First, with regard to servers, physically different devices are installed for each service provider who is a data center user, and if each service provider has server stock in advance, the server will be strengthened. It is assumed that the time until can be made relatively short.
[0005]
Next, regarding the network, even if each server provider applies for an enhancement to the data center operator (system) and a contract is established, the data center operator further communicates with the communication operator (system). Therefore, it is assumed that it is necessary to apply for an enhancement of a line (Wide Area Network: hereinafter referred to as WAN) for connecting to the outside. Therefore, when access to a specific server increases rapidly, it becomes difficult for a service provider who operates the specific server to ensure sufficient quality without congestion during the period until the network is augmented. In addition, the service provider may miss a business opportunity during this period.
[0006]
In view of the above, one of the objects of the present invention is a network service operation in which the WAN connection bandwidth in a data center network is flexibly increased in a short period of time, and the service is continuously provided even when access is rapidly increased. To provide a system. Another object of the present invention is to provide a network service operation system that avoids a server or network congestion state and suppresses loss of business opportunities.
[0007]
[Means for Solving the Problems]
According to the solution of the present invention,
A network service operation system for connecting an external network and a service providing server,
A router that determines whether access to the service providing server exceeds a predetermined service band, transmits data in the service band to the service providing server, and discards data exceeding the service band ,
A service bandwidth information is stored for each service providing server , and for each service providing server, there is a spare bandwidth to be added, or a database storing a bandwidth amount;
When a predetermined number or more of data discards by the router are detected, the database is referenced to determine whether a spare band is allocated to the service providing server, and when a spare band is allocated, A bandwidth management server for performing notification for changing the service bandwidth of the service providing server;
Referring to the database, the service band for the service providing server is set in the router, and the router is connected to the service providing server according to a notification for changing the service band from the band management server. A policy server that changes a service band and changes service band information of the service providing server with respect to the database; and
A network service operation system provided with
[0008]
As one of the features of the present invention, QoS control is performed for each SLA contract unit between a data center operator system connected to an external network and a WAN and a service providing system that performs a desired service using the data center operator system. , A QoS information setting means for instructing the network connection apparatus to set QoS control, a measurement means for measuring a packet loss generated in the network connection apparatus for each SLA contract unit, and the SLA contract SLA management means for managing the database. Here, the SLA contract includes information on the normal service band and the emergency service band in the WAN connection.
[0009]
As another feature of the present invention, for example, based on a measurement result of the measuring unit, a packet loss caused by a lack of contract bandwidth in the network connection device is an emergency state based on information included in the SLA contract. When determined by the SLA management unit, the QoS information setting unit may further include an emergency state instruction unit that instructs the QoS service setting unit to divert the emergency service band from another band as a band for the service providing system. .
[0010]
As another feature of the present invention, for example, the SLA contract may include information for the service providing system to pay a fee to the data center provider system in advance according to time using the emergency service band. . As another feature of the present invention, for example, the SLA contract may include information related to an emergency treatment when a communication band with the outside is insufficient in a data center provider system using a QoS guaranteed network.
[0011]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
FIG. 1 is a schematic configuration diagram of a network service operation system according to the present embodiment.
The network service operation system (here, the data center 1) includes, for example, a line (WAN) 2, a router 3, a policy server 4, a measurement server 5, a bandwidth management server 6, an SLA database 7, and a WWW (World Wide Web) server 11. , 12 and 13. A line (WAN) 2 is a line (WAN) connecting the data center 1 and an external network. The router 3 performs QoS control for each SLA contract unit between the data center 1 that is WAN-connected to the external network and a service providing system that performs a desired service using the data center 1. The policy server 4 instructs the router 3 to set QoS control. The measurement server 5 measures the packet loss (discard) generated in the router 3 for each SLA contract unit. The bandwidth management server 6 manages the SLA database 7 of the SLA contract. The SLA database 7 is a database that stores SLA information.
[0012]
The WWW servers 11, 12, and 13 are servers that provide the web. For example, the data center 1 provides information such as products to external customers. Here, the WWW servers 11, 12, and 13 are WWW servers for company A, company B, and company C, which are service providers, respectively.
[0013]
FIG. 2 is a schematic diagram of a contract bandwidth in an SLA contract.
The contract bandwidth configuration here is stored in, for example, the first memory of the SLA database 7, and the breakdown of the bandwidth of the WAN line 2 is divided into the data center 1 and each service provider system (here, the WWW servers 11, 12,. 13) is set in the SLA contract. In the present embodiment, the bandwidth of the WAN line 2 will be described below assuming that the bandwidth is, for example, 100 megabits / second.
[0014]
The contracted bandwidth includes, for example, a contracted bandwidth, a control bandwidth, and a backup bandwidth that are held by the WWW servers 11, 12, and 13, respectively. Specifically, the bandwidth contracted as the normal service bandwidth of the service provider A is 20 megabits / second. This indicates that the communication bandwidth between the company A WWW server 11 and the outside (that is, the company A customer on the external network) is a maximum of 20 megabits / second. Similarly, the contract bandwidth of service provider B is 30 megabits / second, and the contract bandwidth of service provider C is 15 megabits / second. Further, the control bandwidth used by the data center 1 for operation management or the like is 3 megabits / second, and the remaining 32 megabits / second is used as a spare bandwidth.
[0015]
Here, the operation of the network service operation system will be described.
FIG. 3 is a sequence diagram for the operation of the data center 1 according to the present embodiment. Here, as an example, packet communication between a customer of company A and the WWW server 11 for company A will be described.
First, the router 3 determines whether or not the amount of packet data between the customer of the company A and the WWW server 11 for the company A exceeds the normal service bandwidth (that is, whether or not the bandwidth is sufficient). (S101). Here, the initial setting of the data center 1 (for example, setting of the normal service band) will be described. The policy server 4 refers to the contents of the SLA database 7 and refers to the contracted band (20, 20) set in the SLA contract. 30 and 15 megabits / second), a control bandwidth (3 megabits / second) and a spare bandwidth (32 megabits / second) are set in the router 3 as QoS control information. In addition, as for the bandwidth allocation of each company in the WAN line 2, that is, the QoS control method in the router 3, a method using an IP address is known, and the description thereof is omitted.
[0016]
Next, when the amount of packet data between the customer of company A and the WWW server 11 for company A does not exceed the normal service bandwidth (here, 20 megabits / second), the router 3 performs the process of step A113 to the company A. Connection processing to the WWW server 11 is performed. That is, here, if the access to the company-wide WWW server 11 is performed within a range not exceeding 20 megabits / second, which is the contract bandwidth of the service provider company A, the router 3 uses the company-wide WWW server 11. And packet data between the customer of company A are not discarded.
[0017]
On the other hand, if it is determined in step S101 that there is not enough bandwidth (when communication between the customer of company A and the WWW server 11 for company A continues to exceed 20 megabits / second), it exceeds the normal service bandwidth. A part of the packet data is discarded by the router 3 (that is, packet loss). That is, when the access to the company A WWW server 11 exceeds 20 megabits / second, which is the contract bandwidth of the service provider A, the access from the company A customer to the company A WWW server 11 3 is not possible due to the packet discard process. If this state continues, the customer of company A will not be able to access company A's WWW server 11 (that is, a loss of business opportunity).
[0018]
Therefore, the measurement server 5 measures the amount of packet data discarded by the router 3 due to a lack of line bandwidth, and detects that the number of discarded packets exceeds a predetermined number (predetermined threshold). Then, a message is transmitted to the bandwidth management server 6 (S103). When the bandwidth management server 6 receives the message, the bandwidth management server 6 refers to the second memory of the SLA database 7 storing the setting state of the spare bandwidth for each service provider (S105).
[0019]
FIG. 4 is an explanatory diagram of the second memory 20.
The second memory 20 stores the contents of the SLA database 7, and includes, for example, a service provider name (for example, company A, company B, company C, etc.) 21, spare band information (presence / absence or bandwidth) 22 is included. In addition, you may comprise the 1st and 2nd memory as one memory which memorize | stored each data corresponding to the service provider.
[0020]
Next, based on the second memory, the bandwidth management server 6 contracts the WWW server 11 for Company A for a reserve bandwidth (emergency service bandwidth: for example, the service provider A company provides emergency service bandwidth as SLA information). It is determined whether or not (S107). When the bandwidth management server 6 confirms that the emergency service bandwidth is contracted in step S107 and the bandwidth to be added is allocated from the backup bandwidth, the bandwidth management server 6 sends a notification to the policy server 4 to change the bandwidth of the company A. To do.
[0021]
The policy server 4 that has received the notification of changing the band of the company A instructs the router 3 and the SLA database 7 to change the band to be connected to the company-wide WWW server 11 (S111). In step S111, the router 3 changes the maximum bandwidth of the company A in the WAN line 2 and connects to the company A WWW server 11 (S113). Thereby, packet discarding between the WWW server 11 for company A and the customer of company A is reduced, and loss of business opportunities can be suppressed.
[0022]
If the measurement server 5 determines that there are no packets to be discarded for a predetermined time, it outputs an instruction for returning to the original state in which the spare bandwidth is not used to the bandwidth management server 6, thereby passing through the policy server 4. The router 3 restores the bandwidth. Further, when “bandwidth” is used as the spare bandwidth information in the second memory 20 of the SLA database 7, it is determined that if the remainder of the spare bandwidth does not exceed this “bandwidth”, it can be allocated. The The predetermined number of discarded packets may be one time or an appropriate value. In addition, the measurement server 5 measures the number of packets that exceed the predetermined bandwidth, not the discarded packets, and outputs a bandwidth change instruction to the bandwidth management server 6 when the number exceeds a predetermined value. The bandwidth of the router 3 may be changed via the server 4.
[0023]
Further, in the network service operation system, for example, in the SLA contract, it is possible to include information for the service providing system to pay the data center operator system in advance according to the time using the emergency service band in advance. At this time, in order to conclude a contract for the normal service band and the emergency service band as an SLA contract between the data center provider system and the service provider system, the service band in the emergency is more in the conditions such as the charge per unit band. Suppose that it is expensive. With this SLA contract, the service provider system only needs to make a contract for a relatively inexpensive band (normal service band) within the expected traffic volume. The advantage of deterring losses can be obtained. The management of these charges can be performed by the bandwidth management server 6 or the like by providing a memory storing the usage time of the spare bandwidth for each service provider. On the other hand, for the data center operator system, the bandwidth unit price is lowered by subscribing the WAN line 2 contract from the telecommunications carrier to a broadband, and a relatively expensive bandwidth usage fee is obtained from the service provider system with a rapid increase in traffic. There is a merit that
[0024]
【The invention's effect】
According to the present invention, as described above, the WAN connection band in the data center network can be flexibly increased in a short period of time, and the service can be continuously performed even if the access increases rapidly. Further, according to the present invention, it is possible to avoid the congestion state of the server or the network and suppress the loss of business opportunities.
[0025]
Further, according to the present invention, it is possible to urgently increase the bandwidth of the WAN line in response to a sudden increase in traffic. For the data center operator system, a part of the WAN bandwidth is urgently provided by the service provider. By assigning to the system, you can get a higher usage fee than a normal contract.
[Brief description of the drawings]
FIG. 1 is a schematic configuration diagram of a network service operation system according to an embodiment.
FIG. 2 is a schematic diagram of a contract bandwidth in an SLA contract.
FIG. 3 is a sequence diagram for the operation of the data center 1 according to the present embodiment.
4 is an explanatory diagram of a second memory 20. FIG.
[Explanation of symbols]
1 Data center 2 Data center and external network (WAN)
3 Router 4 Policy Server 5 Measurement Server 6 Bandwidth Management Server 7 SLA Database 11 WWW Server for Company A as a Service Provider 12 WWW Server for Company B as a Service Provider 13 WWW Server for Company C as a Service Provider

Claims (3)

A network service operation system for connecting an external network and a service providing server,
A router that determines whether access to the service providing server exceeds a predetermined service band, transmits data in the service band to the service providing server, and discards data exceeding the service band ,
A service bandwidth information is stored for each service providing server , and for each service providing server, there is a spare bandwidth to be added, or a database storing the bandwidth amount;
When a predetermined number or more of data discards by the router are detected, the database is referenced to determine whether a spare band is allocated to the service providing server, and when a spare band is allocated, A bandwidth management server for performing notification for changing the service bandwidth of the service providing server;
Referring to the database, the service band for the service providing server is set in the router, and the router is connected to the service providing server according to a notification for changing the service band from the band management server. A policy server that changes a service band and changes service band information of the service providing server with respect to the database; and
Network service operation system with 2. The network service operation system according to claim 1, wherein the router performs connection control between an external network and the service providing server according to information on a service band for each service providing server stored in the database. 3. The network according to claim 1, wherein the bandwidth management server stores, for each service providing server, information for the service provider server to charge according to a time using the spare bandwidth. Service operation system.

Images