JPH10198642A - Server device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve the performance of processing by processing one sort of service by load distribution using plural back end servers. SOLUTION: The server device is provided with plural back end servers 12 each of which includes a sever process 21 for processing a client request, a front end server 11 for receiving a request from a client and a request transfer control mechanism 113 built in the server 11 and capable of transferring the received request to a suitable back end server 12. The mechanism 113 determines a back end server 12 for transferring requests from plural clients to the same service by using client identification information and transfers the requests to the determined server 12.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a server configuration of a client / server system in a computer environment, and more particularly, to a configuration of a server in a cluster system (a system in which a plurality of nodes are connected). The present invention relates to a server device that performs load distribution.

[0002] In recent years, with the spread of open systems such as the Internet, interest in client / server systems has increased, and in particular, server systems in which requests from an unspecified large number of clients are expected to be concentrated have high performance / high performance. As a technology for making it usable (availability is increased), it is desired to develop a technology relating to load balancing of a cluster system.

[0003]

2. Description of the Related Art One of load balancing techniques in a conventional cluster system is redirection (allocation of work) by a router or a packet filter. In this method, the server includes a front-end node (FEP node) that receives a request from a client, and a back-end node (BEP node) group in which a function of processing a request from the client is distributed and arranged in a plurality of nodes. Was. All clients send requests to FEP nodes, which forward requests to the appropriate BEP nodes for the requested service and perform load balancing.

FIG. 7 is an explanatory view (1) of a conventional example, and FIG. 8 is an explanatory view (2) of a conventional example. Hereinafter, a conventional example will be described with reference to FIGS. FIG. 7 shows an example of a conventional load distribution using redirection, in which server domain 1 is used.
And a plurality of client domains are connected via a WAN (Wide Area Network).

The server domain 1 has an FEP server 1
1, a plurality of servers 12 (servers A to C), a domain name server (DNS) 13, and a router 14 are provided. The client domain contains multiple clients (1)
-(3), a DNS 43 and a router 44 are provided. The client domain includes multiple clients (4) to
(6) The DNS 43 and the router 44 are provided.

[0006] The FEP server 11 has a transfer control process for receiving a request from a client via a LAN (local area network) and transferring the request to an appropriate BEP node through high-speed node-to-node coupling.
Node. The servers 12 (servers A to C) are BEP nodes having server processes X to Z for processing different types of services X to Z, respectively.

The DNS 13 is connected to the LAN and responds with an IP address (address unique to each computer) corresponding to a service name inquired from a client or the like. For example, FIG. 8B illustrates an example of an IP address. As illustrated in FIG. 8B, the client acquires the IP address “133.160.12.5” of the service name “ftp-server” from the DNS 13 and Obtained IP
It makes a request to an address. The router 14 performs a relay process between the WAN and the LAN.

The plurality of clients (1) to (3) and the plurality of clients (4) to (6) are connected by respective LANs and request services (data processing). The DNS 43 is connected to the LAN and responds with an IP address corresponding to the service name for which an inquiry has been made. The router 44 performs a relay process between the LAN and the WAN.

FIG. 8A is an explanation of a work allocation table, and shows the arrangement of each server process managed by a process on the FEP node. FIG. 8A shows a transfer destination server identifier corresponding to the service identifier. For example, service X is transferred to server A, service Y is transferred to server B, and service Z is transferred to server C.

[0010] Requests from all clients are:
Client domain LAN, router 44, WAN,
FEP via router 14, server domain 1 LAN
Sent to the node, the FEP node determines the destination node of the request according to the type of service requested, and forwards the request to the appropriate BEP node via high-speed inter-node coupling. The server processes X to Z on the BEP nodes A to C process the request transferred from the FEP node and reply to the client.

[0011]

In the above-mentioned conventional redirection method, a server process for processing a different service is arranged for each BEP node, and the service requested by the client is determined using a table as shown in FIG. Is used as a key to determine the transfer destination BEP node. For this reason, even if the processing of one type of service increases, a plurality of B
There is a problem that the load cannot be distributed by the EP node.

[0012] The present invention solves such a conventional problem, and enables requests from a plurality of clients for the same service to be load-balanced by a plurality of server nodes. And
An object of the present invention is to reduce system management costs by having a load adjustment function.

[0013]

FIG. 1 is a diagram illustrating the principle of the present invention. In FIG. 1, 11 is a front-end server, 12
Is a back-end server, 21 is a server process, 22 is a communication control unit between nodes, 23 is a connection state monitoring mechanism, 111 is a list of allocated transfer paths, 112 is a request transfer control table, 113 is a request transfer control mechanism, 11
4 is a client communication control unit, 115 is a communication control unit between nodes in the cluster, 121 is a load adjustment unit, 122 is a node failure detection mechanism, 123 is a node load monitoring mechanism, and 124
Is a server process control mechanism, and 125 is a user interface mechanism.

The present invention is configured as follows to solve the above-mentioned conventional problems. (1): a plurality of back-end servers 12 in which server processes 21 for processing client requests are arranged
And a front-end server 11 for receiving a request from a client, and a request transfer control mechanism 113 arranged on the front-end server 11 for transferring the received request to an appropriate back-end server 12
The request transfer control mechanism 113 determines the backend server 12 to which requests from a plurality of clients for the same service are transferred using the identification information of the clients, and transfers the requests.

(2): In the server device of (1), the request transfer control mechanism 113 controls the ratio of requests to be transferred for each back-end server 12 using the request transfer control table 112.

(3): In the server device of (1) or (2), a node failure detecting mechanism 122 for detecting a server failure is provided, and when the node failure detecting mechanism 122 detects a server failure, the request transfer control is performed. The mechanism 113 is
The request transfer to the failed back-end server 12 is stopped, and the request is transferred to another normally operating back-end server 12 at the time of re-request from the client.

(4) In the server device of (1) to (3), a node load monitoring mechanism 123 for monitoring the load of the back-end server 12 and changing the request transfer ratio is provided. When the mechanism 123 finds the backend server 12 with a high load, the request transfer control mechanism 113 lowers the transfer ratio to the backend server 12 with a high load, and averages the load between the backend servers 12.

(5) In the server device of (1) to (4), the system administrator operates the back-end server 1
There is provided a user interface mechanism 125 for controlling the transfer ratio to the second.

(6) In the server apparatus of (1) to (5), a connection status monitoring mechanism 23 for monitoring a connection status between the server process 21 and the client process is provided, and the request transfer control mechanism 11 is provided.
3 inquires of the connection state monitoring mechanism 23 about the connection state, and changes the transfer route for the connection in service after the end of the service.

(7): In the server device of (1) to (6), a server process control mechanism 124 for controlling the server process 21 is provided, and each back-end server in which the server process 21 of a certain service is arranged. 12 when the load on the server process control mechanism 1 is increased.
24 starts the server process 21 on the back-end server 12 where the server process 21 is not arranged.

(8) In the server device of (1) to (7), the request transfer control mechanism 113 is provided between the network driver for controlling network communication and the packet processing unit of the operating system for processing packets. It consists of a packet filter located in

(9) In the server device of (1) to (8), a pair of a source address and a source port number of a packet of a request received from the client is used as the client identification information.

(Operation) The operation based on the above configuration will be described. The server process 21 arranged in the plurality of back-end servers 12 processes the client request, and the request from the client received by the request transfer control mechanism 113 arranged in the front-end server 11 is identified by using the client identification information. The back-end server 12 to which requests from a plurality of clients for a service are transferred is determined and the requests are transferred. Therefore, the processing of one type of service can be load-balanced among a plurality of back-end servers, and the processing performance can be improved.

The request transfer control mechanism 113
Then, the ratio of requests to be transferred is controlled for each back-end server 12 using the request transfer control table 112.
Therefore, the load can be distributed according to the processing capacity of the back-end server, and the processing performance can be further improved.

Further, when the node failure detection mechanism 122 detects a failure of the server, the request transfer control mechanism 113
Then, the transfer of the request to the failed back-end server 12 is stopped, and the request is transferred to another normally operating back-end server 12 at the time of re-request from the client. Therefore, the failure of the back-end server can be concealed from the client by the re-request from the client.

When the node load monitoring mechanism 123 finds the backend server 12 with a high load, the request transfer control mechanism 113 lowers the transfer ratio to the backend server 12 with a high load, and The load between them. Therefore, the load can be constantly averaged, and the processing performance can be improved.

Further, the user interface mechanism 125
Then, the system administrator controls the transfer ratio to the back-end server 12. For this reason, for example, the transfer ratio of the back-end server for active maintenance can be set to zero, and the node can be temporarily disconnected or re-installed.

The request transfer control mechanism 113
Inquires the connection state monitoring mechanism 23 for the connection state, and for the connection in service,
After the service ends, the transfer path is changed. For this reason, it is not possible to switch to a new back-end server during the service, and interruption of the processing can be prevented.

Further, a server process 2 of a certain service
When the load on each back-end server 12 on which the server process 1 is increased, the server process control mechanism 124 starts the server process 21 on the back-end server 12 on which the server process 21 is not located. For this reason, it is possible to adjust the load of the plurality of back-end servers as a whole, and to improve the processing performance.

The request transfer control mechanism 113
Is composed of a network driver that controls network communication and a packet filter that is arranged between a packet processing unit of an operating system that processes packets. Therefore, the processing of one type of service can be load-balanced by a plurality of back-end servers by the packet filter, and the processing performance is improved.

Further, a pair of a source address and a source port number of a packet of a request received from the client is used as the identification information of the client.
Therefore, it is possible to control the transfer destination node of the request for each client process.

[0032]

FIG. 2 to FIG. 6 are views showing an embodiment of the present invention. According to the present invention, the request transfer control in the FEP node uses, in addition to the type of service requested by the client, the identifier of the client that sent the request, the BEP node to which the request for the same service is transferred is controlled for each client. .
Hereinafter, embodiments of the present invention will be described with reference to the drawings.

(1): Description of server device: Description of device configuration FIG. 2 is a device configuration diagram (1). Hereinafter, the server device will be described with reference to FIG.

In FIG. 2, the server device includes an FEP server (FEP node) 11 and a plurality of BEP servers (BEP servers).
The nodes 12a, 12b, 12c,... Client 4 is coupled to FEP node 11 via LAN or WAN 5.

The FEP server 11 includes an assigned transfer path list 111, a transfer control table 112, a request transfer processing unit 113, a client communication control unit 114, an inter-cluster communication control unit 115, a load adjustment unit 121, and a node. Failure detection unit 122, node load monitoring unit 123, server process control unit 124, system administrator interface 1
25 are provided.

The BEP server 12a is provided with a server process 21a, an inter-node communication control unit 22a, and a connection status monitoring unit 23a.
A server process 21b, an inter-node communication control unit 22b, and a connection state monitoring unit 23b are provided. The BEP server 12c is also provided with a server process 21c, an inter-node communication control unit 22c, and a connection state monitoring unit 23c.

The assigned transfer path list 111 records the transfer paths of requests assigned to each client. The transfer control table 112 records the arrangement of server processes that process requests and the distribution ratio of requests. The request transfer processing unit 113 transfers a request from a client to a BEP node. The communication-to-client control unit 114 controls communication with the client. The intra-cluster communication control unit 115 controls communication with the BEP node. The load adjustment unit 121 dynamically adjusts the transfer control according to a request from a system administrator or an event that has occurred. The node failure detection unit 122
This is to detect a failure of the P node. The node load monitoring unit 123 monitors the load state of the BEP node. The server process control unit 124 starts a server process required for the BEP node. The system administrator interface 125 provides the system administrator 6 with a user interface.

Server processes 21a, 21b, 21c
Is a process that processes requests from clients. Inter-node communication control units 22a, 22b, 22c
Controls communication with the FEP node. The connection state monitoring units 23a, 23b, and 23c monitor the connection state between the client and the server process.

Description of Transfer Control Table and Assigned Transfer Route List Table FIG. 3 is an explanatory diagram of the transfer control table and the assigned transfer route list table. FIG. b) is an explanation of the assigned transfer route list.

In FIG. 3A, in the transfer control table 112, a service identifier using a port number or the like as a destination of a request, an IP address or the like of an available BEP node as a transfer destination can be used. A ratio is provided as a BEP node identifier and a request distribution ratio (transfer ratio) for the same service.

In FIG. 3B, the assigned transfer route list 111 includes a service identifier using a port number or the like as a destination of a request, and a client identifier using a source IP address and a port number of a source port. , A destination BEP node identifier using the IP address or the like of the destination BEP node.

(2) Description of the Operation of the Server The operation of the server will be described below with reference to FIG. : Explanation when there is a request from client Client 4 sends request to LAN or WAN 5
The request received by the client communication control unit 114 is sent to the request transfer processing unit 113. First, the request transfer processing unit 113 uses the assigned transfer path list 111 to check whether the request is from a client to which a transfer path has already been assigned. The assigned transfer path list 111 records the identifier of the assigned transfer destination BEP node using a pair of the identifier of the requesting client and the identifier of the requested service as a key.

If there is no client entry in the assigned transfer route list 111, the request transfer processing unit 1
13 determines the transfer destination BEP node with reference to the transfer control table 112. In the transfer control table 112, available BEP node identifiers and request distribution ratios are recorded for each service.

The transfer destination B of the request in the transfer control table 112
When the EP node is determined, the request transfer processing unit 113
Creates an entry of the client to which the transfer destination is assigned in the allocated transfer route list 111, and transfers the request to the BEP node determined as the transfer destination via the inter-node communication control unit 115 and the inter-node connection 3. I do.

In the BEP node to which the request has been transferred (for example, the BEP node 12a), the inter-node communication control unit 22a receives the request and sends the request to the server process 21a. Server process 21a
Processes the request, the inter-node communication control unit 22
a. A reply is sent to the FEP node 11 via the inter-node connection 3. The FEP node 11 receives a reply at the inter-cluster communication control unit 115, and the request transfer processing unit 113 transmits the reply to the client communication control unit 114 and the LA.
A reply is returned to the client 4 via N or WAN5.

In the case of a request from a client whose entry exists in the assigned transfer route list 111, the request transfer processing unit 113 sets the assigned transfer route list 11
The request is sent to the BEP node assigned in step 1. By performing such processing, a request from a client identified by the same client identifier is transferred to the same BEP server.

Description in Case of Node Failure The node failure detection unit 122 periodically confirms that all BEP nodes are operating normally. Upon detecting a BEP node failure, the node failure detection unit 122 rewrites the transfer control table 112 via the load adjustment unit 121, and
Stop distribution of requests to the failed BEP node. The request transfer processing unit 113 deletes all entries in the assigned transfer route list 111 in which the transfer destination BEP node is a failed BEP node.

Although the connection between the client using the failed BEP node and the server process on the BEP node is disconnected by the failure of the BEP node, the failure is hidden by the client retrying the request.

When the re-request from the client is sent to the request transfer processing unit 113, the request transfer processing unit 113 checks the assigned transfer path list 111, but since the entry of the re-requested client has already been deleted. The request transfer processing unit 113 determines a new transfer destination based on the transfer control table 112 and transfers the request. The BEP node to which the request has been transferred processes the request and replies in the same manner as a normal request.

Description of Node Load Monitoring The node load monitoring unit 123 periodically monitors the load status of all BEP nodes. The node load monitoring unit 123
If a BEP node with a high load is found, the load adjustment unit 12
1, the transfer control table 112 is rewritten, and the distribution ratio of requests to BEP nodes with a high load is reduced. The request transfer processing unit 113 inquires of the connection status monitoring units 23a, 23b, 23c,... Of each BEP server about the status of the assigned transfer route when the distribution ratio is changed, and from the assigned transfer route list 111, the inactive connection. , The entry of the transfer route is deleted, and the request distribution ratio at that time is calculated.

The request transfer processing unit 113 adjusts a request transfer destination from a new client and updates the assigned transfer route list 111 by inquiring the connection state monitoring units 23a, 23b, 23c,. The actual distribution ratio is made closer to the distribution ratio specified in the rewritten transfer control table 112.

Description of Start / Stop of Server Process The server process control unit 124 controls the server process on each BEP node via the inter-node communication control unit 115 and the inter-node connection 3.

All BEs assigned to a service
When the load of the P node is high, the node load monitoring unit 12
No. 3 requests the load adjustment unit 121 to increase the allocation of BEP nodes. When the load adjustment unit 121 that manages the assignment of the BEP node to each service decides to increase the BEP node assignment to the service having a high load, the load adjustment unit 121 starts the server process to the newly assigned BEP node. Ask 124. Therefore, the server process control unit 124 includes the communication control unit 115 between the nodes in the cluster.
The server process on the designated BEP node is started via the inter-node connection 3. When the activation of the server process is completed, the load adjusting unit 121 rewrites the transfer control table 112 so as to transfer the request to the newly assigned BEP node.

When it is necessary to stop the server process that has been running on the newly assigned BEP node, the load adjusting unit 121 rewrites the transfer control table 112 and stops the distribution for the request (distribution). The ratio is 0). The request transfer processing unit 113 inquires of the connection state monitoring unit of the BEP node instructed to stop the transfer of the request about the connection state, and deletes the unused state entry in the allocated transfer path list 111. The load adjustment unit 121 performs the assignment transfer list 1
Referring to FIG. 11, when there is no more entry for a connection to the designated BEP node, the server process controller 124 is requested to stop the server process. The server process control unit 124 sends the designated BEP via the intra-cluster communication control unit 115 and the inter-node connection 3.
Stop the server process on the node.

Description of System Administrator Interface The system administrator interface 125 provides the system administrator 6 with an interface for system management.

The functions provided by the system administrator interface 125 include a dynamic (operating) operation for a cluster.
There are functions such as addition / disconnection of nodes, designation / change of assignment of BEP nodes to services, and reference to other system states.

(3): Description using a UNIX (General-purpose Operating System of AT & T) Machine A description will be given below using a UNIX machine with reference to FIGS.

FIG. 4 is an explanatory diagram of a client / server system.
Hereinafter, description will be made with reference to FIG.

FIG. 4 shows an example of load distribution using redirection, in which a server domain 1 and a plurality of client domains are connected via a WAN. The server domain 1 includes an FEP server 11, a plurality of servers 12 (servers A to F), a domain name server (DNS) 13, and a router 14. The client domain includes multiple clients (1) to
(5) A DNS 43 and a router 44 are provided. The client domain includes a plurality of clients (6) to
(10) A DNS 43 and a router 44 are provided.

The FEP server 11 receives a request from a client via a LAN (local area network), and performs appropriate BEP node 1
2 is a FEP node having a packet filter 113 for forwarding a request to the FEP node. Server 12 (servers A to F)
Is a server process 21 that processes services X, Y, and Z
(Server process X, Y, Z). The DNS 13 is connected to the LAN and has an I corresponding to a service name inquired from a client or the like.
It responds with a P address (an address unique to each computer).

A plurality of clients (1) to (5) and a plurality of clients (6) to (10)
And is a side that requests a service (data processing). The DNS 43 is connected to the LAN and responds with an IP address corresponding to the service name for which an inquiry has been made. The router 44 performs a relay process between the LAN and the WAN.

Description of the server device Regarding the load distribution in the server device (cluster system) of the UNIX machine, the request transfer processing unit 1 in FIG.
Reference numeral 13 is realized as a packet filter located at the lowest layer of the stream processing unit that performs data processing on the network. The assigned transfer route list 111 and the transfer control table 112 are located in the packet filter. Also,
The load adjustment unit 121 in FIG. 2 is realized as a load adjustment process that is a UNIX process in the user space. This load adjustment process rewrites the transfer control table in the packet filter using an IOCTL system call provided by a UNIX OS (operating system).

FIG. 5 is a device configuration diagram (2). Hereinafter, the server device will be described with reference to FIG. In FIG.
The server device includes an FEP node 11 and a BEP node 12
(Actually, a plurality of BEP nodes are provided) are connected by an inter-node coupling network 3. The client is an FEP node 1 via a network 5 such as a LAN.
It is connected to 1.

The FEP node 11 has a kernel (central part) space and a user space. In the kernel space, a stream processing unit, an external network driver 114, and an internal network driver 115 are provided. The stream processing unit includes a packet filter 113, an IP packet processing layer 131, and a TCP packet processing layer 132.
The packet filter 113 includes an assigned transfer route list 1
11 and a transfer control table 112 are provided. In the user space, a load adjustment process 121, a node failure detection process 122, a node load monitoring process 123, a server control process 124, and a user I / F processing process 125 are provided.

The BEP node 12 also has a kernel space and a user space. The kernel space has a stream processing unit,
An internal network driver 22 is provided. The stream processing unit includes an IP packet processing layer 241 and a TCP packet processing layer 242. In the user space, a connection monitoring process 23 and a server process 21 are provided.

The assigned transfer path list 111 records the transfer paths of requests assigned to each client. The transfer control table 112 records the arrangement of server processes that process requests and the distribution ratio of requests. The packet filter 113 transfers a request from a client to a BEP node. The external network driver 114 controls communication with the client. The internal network driver 115 controls communication with the BEP node. The load adjustment process 121 dynamically adjusts transfer control according to a request from a system administrator or an event that has occurred. The node failure detection process 122
This is to detect a failure of the BEP node. The node load monitoring process 123 monitors the load state of the BEP node. The server control process 124 performs BEP
Starts server processes required for the node.
The user I / F processing process 125 provides a user interface to a system administrator. The IP packet processing layer 131 processes Internet Protocol (IP) packets, which is one of the communication protocols. The TCP packet processing layer 132 is a TCP (transmission control protocol) which is one of the communication protocols.
tocol) packet processing.

The server process 21 is a process for processing a request from a client. The internal network driver 22 controls communication with the FEP node 11. The connection monitoring process 23 monitors the connection state between the client and the server process. The IP packet processing layer 241 processes Internet Protocol packets. TC
The P packet processing layer 242 processes a TCP packet.

Description of Transfer Control Table and List of Assigned Transfer Routes FIG. 6 is an explanatory diagram of the transfer control table and the list of assigned transfer routes. FIG. 6A shows the description of the transfer control table, and FIG. b) is an explanation of the assigned transfer route list.

In FIG. 6A, the transfer control table 112 has a service identifier, a transfer destination BEP node identifier, and a ratio which is a request distribution ratio for the same service.

The port number of the destination (destination) port is used as the service identifier. In this example, “8080,8081,8082” is recorded.
A destination IP address is used as the transfer destination BEP node identifier. In this example, for example, “IP-A, IP-B” is recorded in the destination IP address of the service “8080”, and the ratio of these is , IP-A and the same ratio as “1” are recorded in IP-B and IP-B, respectively.

In FIG. 6B, the assigned transfer route list 111 includes a service identifier, a client identifier,
A destination BEP node identifier is provided. The port number of the destination port is used as the service identifier, and in this example, “8080,8081,8082” is recorded.

A pair of a source IP address and a source port number is used as a client identifier. The source IP address is the IP address of the client, and the source port number is provided for each process. In this example,
(For example, client (1) in FIG. 4) The source IP address “IP-CL1” has three source port numbers “P
ORT-CL1-xx1 "," PORT-CL1-xx1 "
2 "and" PORT-CL1-xx3 ".

A destination IP address is used as the transfer destination BEP node identifier. In this example, for example, “IP-A” and “IP-B” are recorded in the destination IP address of the service “8080”. .

The correspondence between the service and the port number is UN
Each node in the cluster is defined in a file on the IX OS and uses a common service port. : Description of the operation of the server device a: Description when there is a request from the client External network driver 114 of the FEP node 11
Receives the request from the client and sends it to the packet filter 113. The packet filter 113
The BEP node to which the packet is to be transferred is determined using the assigned transfer path list 111 and the transfer control table 112, and the packet is transferred to the BEP node 1 via the inter-node connection network 3.
Send to 2.

BEP node 1 to which this packet was transferred
The second internal network driver 22 sends the received packet to the stream processing unit. The server process 21 reads the packet that has reached the TCP packet processing layer 242 using a system call interface provided by UNIX, and processes the request.

When the processing of the request is completed, the server process 21
The packet is passed to the stream processing unit using the IX system call. The reply packet is transferred to the FEP node 11 via the internal network driver 22 and the inter-node coupling network 3 and then transmitted to the FEP node 11.
Is sent to the client.

B: Description of Case of Node Failure The node failure detection process 122 is implemented as a user process. The node failure detection process 122
Periodically, it is confirmed that all the BEP nodes are operating normally, and when a failure of the BEP node is detected, the transfer control table 112 of the packet filter 113 is rewritten via the load adjustment process 121 and the failed BEP node is rewritten. Stop distributing requests to. Packet filter 113
Deletes all entries in the assigned transfer route list 111 in which the transfer destination BEP node is a failed BEP node.

Although the connection between the client using the failed BEP node and the server process on the BEP node is disconnected by the failure of the BEP node, the failure is hidden by the client retrying the request.

When the re-request from the client is sent to the packet filter 113, the packet filter 11
3 examines the assigned transfer route list 111, but since the entry of the client that has re-requested has already been deleted, the packet filter 113 checks the transfer control table 112.
A new transfer destination is determined based on the request, and the request is transferred. The BEP node to which the request has been transferred processes the request and replies in the same manner as a normal request.

C: Description of Node Load Monitoring The node load monitoring process 123 periodically communicates with the BEP node 12 to monitor the load state of the BEP node 12. The node load monitoring process 123 checks for a high load BE.
When the P node 12 is found, the P node 12 is notified to the load adjustment process 121.
Packet filter 113 using CTL system call
Is rewritten in the transfer control table 112.

The packet filter 113 inquires the connection monitoring process 23 on the BEP node 12 about the connection status of the client. The connection monitoring process 23
It inquires of the stream processing unit of its own node about the connection state using TL, and notifies the packet filter 113 of the connection state. The packet filter 113 adjusts the transfer ratio by changing only the transfer destination of the unused connection.

The load adjustment process 121 requests the server control process 124 to start or stop the server process 21 to the BEP node 12, if necessary. As described above, according to the present invention, it is possible to load-balance requests from a plurality of clients for the same service by a plurality of server nodes. Further, the system management cost can be reduced by providing an automatic server node failure concealment function and a load adjustment function.

[0083]

As described above, the present invention has the following effects. (1): A request from a client received by a request transfer control mechanism arranged in a front-end server is determined by using a client's identification information to determine a back-end server that transfers requests from a plurality of clients for the same service. To transfer
The processing of one type of service can be load-balanced by a plurality of back-end servers, and the processing performance can be improved.

(2): The request transfer control mechanism controls the ratio of requests to be transferred using the request transfer control table for each back-end server, so that load can be distributed according to the processing capacity of the back-end server. Processing performance can be improved.

(3): When the node failure detection mechanism detects the failure of the server, the request transfer control mechanism stops the request transfer to the failed back-end server and operates normally when the client re-requests. Since the request is forwarded to another back-end server, the re-request from the client can hide the failure of the back-end server from the client.

(4): When the node load monitoring mechanism finds a backend server with a high load, the request transfer control mechanism lowers the transfer ratio to the backend server with a high load to reduce the load between the backend servers. Since the averaging is performed, the load can be constantly averaged, and the processing performance can be improved.

(5) In the user interface mechanism, the system administrator controls the transfer ratio to the back-end server. For example, the transfer ratio of the back-end server for active maintenance is set to zero, and the node is temporarily disconnected. And re-incorporation becomes possible.

(6): The request transfer control mechanism inquires of the connection state monitoring mechanism about the connection state, and changes the transfer route for the connection in service after the end of the service. It is possible to prevent interruption of processing without switching to the back-end server.

(7): When the load on each back-end server on which a server process of a certain service is located becomes high, the server process control mechanism starts the server process on the back-end server on which the server process is not located. Thus, the load on the entire plurality of back-end servers can be adjusted, and the processing performance can be improved.

(8): Since the request transfer control mechanism is composed of a network driver for controlling network communication and a packet filter arranged between the packet processing unit of the operating system for processing packets, one type of packet filter is used. The processing of this service can be load-balanced among multiple back-end servers, and the processing performance improves.

(9): As the identification information of the client,
To use the source address and source port number pair in the packet of the request received from the client,
The request transfer destination node can be controlled for each client process.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating the principle of the present invention.

FIG. 2 is a device configuration diagram (1) according to an embodiment.

FIG. 3 is an explanatory diagram of a transfer control table and an assigned transfer route list according to the embodiment;

FIG. 4 is an explanatory diagram of a client / server system in the embodiment.

FIG. 5 is a device configuration diagram (2) according to the embodiment.

FIG. 6 is an explanatory diagram of a transfer control table and an assigned transfer route list according to the embodiment.

FIG. 7 is an explanatory view (1) of a conventional example.

FIG. 8 is an explanatory view (2) of a conventional example.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 11 Front-end server 12 Back-end server 21 Server process 22 Inter-node communication control unit 23 Connection status monitoring mechanism 111 Assigned transfer route list table 112 Request transfer control table 113 Request transfer control mechanism 114 Client communication control unit 115 Between nodes in cluster Communication control unit 121 Load adjustment unit 122 Node failure detection mechanism 123 Node load monitoring mechanism 124 Server process control mechanism 125 User interface mechanism

Claims (9)

[Claims] A plurality of back-end servers in which server processes for processing client requests are arranged; a front-end server for receiving requests from clients; and a plurality of back-end servers arranged in the front-end server. Request transfer control mechanism for transferring to the back-end server, the request transfer control mechanism determines, using the identification information of the client, the back-end server that transfers requests from a plurality of clients for the same service. A server device characterized in that a request is forwarded through the server. 2. The method according to claim 1, wherein the request transfer control mechanism controls a ratio of requests to be transferred for each back-end server using a request transfer control table.
The server device according to the above. 3. A node failure detection mechanism for detecting a failure of a server. When the node failure detection mechanism detects a failure of the server, the request transfer control mechanism stops request transfer to the failed back-end server. 3. The server device according to claim 1, wherein the request is transferred to another back-end server that is operating normally when a request is made again from a client. 4. A node load monitoring mechanism for monitoring a load on the back-end server and changing a request transfer ratio, and when the node load monitoring mechanism finds the back-end server with a high load, a request transfer control mechanism. The server device according to any one of claims 1 to 3, wherein a load ratio between the back-end servers is averaged by lowering a transfer ratio to a back-end server having a high load. 5. The server device according to claim 1, wherein a user interface mechanism for controlling a transfer ratio to said back-end server by a system administrator is provided. 6. A connection state monitoring mechanism for monitoring a connection state between the server process and the client process, wherein the request transfer control mechanism inquires of the connection state monitoring mechanism about a connection state, and 6. The server device according to claim 1, wherein a transfer path is changed after the service is completed. 7. A server process control mechanism for controlling the server process, wherein when a load on each back-end server in which a server process of a certain service is increased, the server process control mechanism controls the server process. 7. The server device according to claim 1, wherein a server process is started on an unarranged back-end server. 8. The apparatus according to claim 1, wherein said request transfer control mechanism comprises a packet filter disposed between a network driver for controlling network communication and a packet processing unit of an operating system for processing packets. The server device according to any one of claims 7 to 7. 9. The server device according to claim 1, wherein a pair of a source address and a source port number of a packet of a request received from the client is used as the identification information of the client.