JP2000322350A - Server switching method for client / server systems - Google Patents

Abstract

(57) [Summary] (Modified) [Problem] In one or a plurality of LANs composed of a plurality of client server systems (C / S) in a parallel relationship, when a failure occurs in a connection destination server, the client is aware of the failure. Instead, the purpose is to dynamically switch the connection destination server to another server by the management device. When a server switching factor occurs, an address of a failed server is described in a primary table, an address of a client to be switched is described in a secondary table, and a server address of a switching destination is described in a tertiary table. Shall be.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a client / server system (C / S) which is used when a server failure occurs.
The present invention relates to a network method for switching a connection destination server of a client in real time.

[0002]

2. Description of the Related Art A backup of a server function in a C / S is mainly performed by changing the configuration of the server itself so that the server can respond to a signal from a client, and a server connection destination is determined by the client. There is something to switch.

According to the prior application, a client has an address table of a switching destination server in advance, and when each client detects a server failure, the server connection destination is changed according to the judgment of each client side.
No. 5498, the backup server is connected online when the regular server and the backup server have dual LANs and addresses for control and connection, and when the regular server fails and the client is switched to the backup server. JP-A-10-135993 which rewrites the LAN address for use to the same address as the address of the regular server and implements backup of the server function, and when a server failure occurs in a LAN to which a plurality of servers are connected, JP-A-6-3430, in which a signal addressed to the server is relayed from a control device to each repeater to a route preset in each repeater, thereby distributing the signal to a backup server.
No. 76 publication.

[0004]

With respect to the backup of the server function in the C / S, according to the existing invention, a time delay due to the restart of the server or the like and uncertainty due to the judgment on the client side are considered. In addition, it is considered that a problem is that backup measures against the system configuration have a strong peculiarity and the system lacks flexibility and scalability.

An object of the present invention is to realize a server backup system which does not require a client to make a server switching decision and does not affect the versatility and expandability of an existing system.

[0006]

According to the present invention, in order to achieve the above object, according to the present invention, when a server switching factor occurs, a management device changes to a connection destination switching device, and a client belonging to the factor generating server changes to the factor generating server. Is set so that the destination address of the signal is rewritten so that the signal is delivered to the server to be switched. Conversely, the transmission from the switched server to the client is set so that the source address of the signal is rewritten to the address of the cause server as if it were sent from the server where the cause occurred.

The operation of the connection destination switching device does not convert signals for transmission and reception between the client and the client and between the server and server, but converts signals only for transmission and reception between the server and the client.

In the signal conversion method of the connection destination switching device, a first table, a second table, and a third table are prepared, and when a server switching factor occurs, the server which has caused the switching factor to the first table is prepared. Address to the second
Table describes the client address for switching the server, and the third table describes the server address of the switching destination. For sending from client to server,
Only for the signal from the client described in the second table, the destination address of the signal is converted from that of the address described in the first table to that of the table described in the third table. For transmission from the server to the client, the source address of the signal is converted from the third table to the server address of the first table only for the signal to the client address described in the second table.

[0009]

DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 shows a C / S network configuration. In the figure, reference numerals 1-1 and 1-2 denote servers, reference numeral 2 denotes a management device,
Reference numerals 3-1 and 3-2 denote connection destination switching devices, and reference numerals 4-1 to 4-6 denote clients. When a client switching factor has occurred in the server 1-1, the switching control is performed from the management device 2, and the clients 4-1 to 4-6 are connected to the server 1-2 by the connection switching device 3-1. Will be changed.

FIG. 2 shows the internal configuration of the connection destination switching device.
In the figure, control information from the management device 2 is stored in a data holding unit 11, and for transmission from a client to a server, a signal received by a client-side interface unit 5 is analyzed by an address conversion unit 7, and converted or not converted. To the server side. As a specific operation of the address conversion mechanism 7, the address analysis unit 8 analyzes a signal, and the address search unit 9 stores the server / client switching information described in the data holding unit 11, the destination address of the signal, and the like. Check the source address. As a result, when the received signal matches the held data, the destination address conversion unit 10 converts the signal.

On the other hand, when the collation is not performed, the data is transmitted without conversion. Similarly, for transmission from the server to the client, the signal received by the server-side interface unit 6 is analyzed by the address conversion mechanism 15 and converted or non-converted and transmitted to the client. As a specific operation in the address conversion mechanism 15, the address analysis unit 18 analyzes a signal, and the address search unit 17 converts the switching information described in the data holding unit 11 and the destination and source addresses of the signal. Collate. As a result, when the received signal is compared with the held data, the source address converter 16 converts the signal.
On the other hand, if the collation is not performed, it is transmitted without conversion.

In the present invention, in order to realize a server backup method by converting addresses in the connection destination switching device, addresses are assigned to the above-described server, client and connection destination switching device, and the following description will be given.
Note that the TCP / IP protocol is assumed for specific signal contents. In the TCP / IP protocol, one signal includes a MAC address and an IP address. However, for convenience, the address indicates an IP address and is used as an identifier for final determination. In the case of the same protocol, a MAC address is used for transmission between devices, and an appropriate MAC address is derived for the MAC address when the device transmits from an IP address according to an existing invention.

As for the address, as shown in FIG. 3, an address S100 is stored in the server 1-1, and the subordinate clients 4-1 through 4-1.
4-3 shows CL101, CL102, and CL103, respectively.
Is given. Similarly, S200 is assigned to the server 1-2, and CL201 and CL20 are assigned to the subordinate clients 4-4 to 4-6, respectively.
2, CL 203 is given. Also, connection destination switching device
K10 and K20 are assigned to 3-1 and 3-2, respectively. Here, something happens to the server 1-1, and the client 4
Consider a case where -1 to 4-3 are switched to server 1-2.
At this time, the administrator sends the connection destination switching device 3-
Perform the following operations on 1.

The connection destination switching device 3-1 (address K1
For 0), the server address where the switching factor has occurred is set in the first table as shown in FIG. Here, the address S100 of the server 1-1 is set. Next, as shown in FIG. 5, a client address for changing the connection destination is set in the second table. Here, the addresses CL101, CL102, CL103 of the clients 4-1 to 4-3 are set. Next, as shown in FIG. 6, the server address of the switching destination is set in the third table. Here, the address S200 of the server 1-2 is set.

The principle of actually switching the connection destination will be described with reference to the flowchart of FIG. When any one of the clients 4-1 to 4-3 transmits a signal to the server 1-1, the signal is first transmitted to the connection destination switching device 3-1 as routing for accessing the server 1-1. Connection destination switching device 3-1
Receives the signal from the client, determines whether it is addressed to itself or routing based on the IP address, and outputs it to the server if necessary.

When sending a call to the server, the connection destination switching device 3-1 performs address conversion or non-conversion by the address conversion mechanism 7 shown in FIG. The address translation mechanism 7 analyzes the address of the signal by the address analysis unit in FIG. 7 and, when the client 4-1 has transmitted to the server 1-1,
S100 is obtained as the destination address, and CL101 is obtained as the source address (step S1).

Next, the first table of data held by the connection destination switching device 3-1 is searched, and the destination address S100
Is obtained (step S2). Next, the second table is searched to find that the source address CL101 is on the second table (step S3).

As a result, the address conversion unit rewrites the destination address S100 into the contents S200 of the third table and transmits the signal to the server (step S4). Here, the third table is the address of the server 1-2 (S200).
Therefore, it is transmitted to the server 1-2. In addition, if the result of the search on the first and second tables does not correspond, the address of the signal is not converted and transmitted as it is in the server-side interface unit as shown in the flowchart of FIG.

FIG. 9 schematically shows the contents of a signal transmitted on the LAN. The left end of the figure corresponds to the head of the signal, and the right end of the figure corresponds to the end of the signal. Here, since only the parts related to the present invention are extracted and represented, the TCP / IP
In the case of the protocol, the destination MAC address 19,
A source MAC address 20, a source IP address 21, a destination IP address 22, and a data part are provided.

The above-mentioned conversion is traced again in terms of the signal content. Client 4-1 (address CL10
FIG. 10 shows the contents before and after the conversion of the signal transmitted from 1) to the server 1-1 (address S100) and switched to the server 1-2 (address S200) by the connection destination switching device 3-1. The client sets S100 (reference numeral 26) as the destination IP address and CL1 as the source IP address.
01 (symbol 25) is set, the MAC address 24 of the client 4-1 is set as the source MAC address, and the MAC address 23 of the connection destination switching device 3-1 is set as the destination MAC address for routing according to the existing invention. Set and send.

The connection destination switching device 3 receiving this signal
In step 1, for transmission to the server side, S100 (reference numeral 26) is converted to S200 (reference numeral 31) as the destination IP address, and when the destination IP address is determined, the destination MAC address is changed to the server 1 by the existing invention. -2 as the MAC address 28 and the source MAC address as the connection destination switching device 3-1
Is transmitted as the MAC address 29.

Next, the reply from the server to the client will be described with reference to the flowchart of FIG. The connection destination switching device 3-1 converts or does not convert the address of the signal by the address conversion mechanism 15 shown in FIG. In the address conversion mechanism 15, the address of the signal is analyzed by the address analysis unit 18 in FIG. 8, and when the signal is transmitted from the server 1-2 to the client 4-1, CL101 is set as the destination address.
S200 is obtained as the source address (step S5).

Next, the third table of the data held by the connection destination switching device 3-1 is searched, and the source address S20 is searched.
0 is obtained (step S6). Then the second
The table is searched to find that the destination address CL101 is on the second table (step S7).

As a result, the signal is transmitted by the address conversion unit with the source address rewritten to the content S100 of the first table. If the search on the third and second tables does not correspond, the address of the signal is transmitted by the client side interface unit without being converted as shown in the flowchart of FIG.

The above-mentioned conversion is traced again in terms of the signal content. Contents before and after conversion of a signal transmitted from the server 1-2 (address S100) to the client 4-1 (address CL101) and returned by the connection destination switching device 3-1 from the server 1-1 (address S100). Is shown in FIG. The server 1-2 uses CL101 (reference numeral 36) as the destination IP address and S2 as the source IP address.
00 (symbol 35) is set, the MAC address 34 of the server 1-2 is set as the source MAC address, and the MAC address 33 of the connection destination switching device 3-1 is set as the destination MAC address for routing according to the existing invention. Set and send. Upon receiving this signal, the connection destination switching device 3-1 sets S200 as the source IP address for transmission to the client side.
(Reference numeral 35) is converted to S100 (reference numeral 40), and the destination MAC address is used as the MAC address 38 of the client 4-1 and the source MAC address is used as the connection destination switching device 3 according to the existing invention.
It is transmitted as the MAC address 39 of -1.

The above description is a method for coping with a switching factor occurring in the server 1-1. However, when a factor has occurred in the server 1-2, the connection destination switching device 3-2 is operated. Is done.

Also, as shown in FIG.
N, the data is distributed from a switching device on each LAN to a server on another LAN via a router. When a server switching factor occurs in server 43-1 and server 43-2 and server 43-3 performs backup,
The signals of the clients 45-1 to 45-3 are transmitted to the connection destination switching device 44.
-1, the destination address is converted to the address of the server 43-3, and the router 47-1, the WAN 46, the router 47-
It is distributed to the server 43-3 via 2. Similarly, the signals of the clients 45-4 to 45-6 are converted from the destination address to the address of the server 43-3 by the connection destination switching device 44-2,
According to the existing invention, the data is distributed to the server 43-3 via the WAN 46 and the router 47-2.

Conversely, a reply to the clients 45-1 to 45-3 is transmitted from the server 43-3 to the clients 45-1 to 45-3, and the router 47-2 and the WAN 46 according to the existing invention. The source address is converted to the server 43-1 by the connection destination switching device 44-1 via the router 47-1, and is distributed to the clients 45-1 to 45-3.

Similarly, a reply to the clients 45-4 to 45-6 is transmitted from the server 43-3 to the clients 45-4 to 45-6, and is transmitted via the router 47-2 and the WAN 46 according to the existing invention. The connection destination switching device 44-2 converts the source address to the server 43-2 and distributes it to each of the clients 45-4 to 45-6.

In this case, if the switching device itself also functions as a router in the LAN of the other system or the distribution server to be controlled is under the control of the switching device, the connection destination can be dynamically changed on the LAN of the other system. It can be switched, and dynamic server switching on each LAN under a consistent address between WANs is possible.

[0031]

As described above, according to the present invention, in the existing system, the backup effect can be obtained only by attaching a part of the connection destination switching device, and even if the number of servers increases, the items in the table can be obtained. By increasing the number, backup can be performed in the same manner as described above, and switching destinations can be distributed. Further, the connection destination switching device may be disposed at a location where a signal between the server and the client always passes, and the number of connection destination switching devices is arbitrary.

[Brief description of the drawings]

FIG. 1 is a basic server backup system of the present invention.

FIG. 2 is an internal configuration diagram of a connection destination switching device of the present invention.

FIG. 3 is a diagram in which addresses are assigned to a server and a client in FIG. 1;

FIG. 4 is a diagram in which an address is set in a first table of the connection destination switching device.

FIG. 5 is a diagram in which addresses are set in a second table of the connection destination switching device.

FIG. 6 is a diagram in which addresses are set in a third table of the connection destination switching device.

FIG. 7 is a processing flowchart of the address conversion mechanism 1 of the connection destination switching device.

FIG. 8 is a processing flowchart of the address conversion mechanism 2 of the connection destination switching device.

FIG. 9 is a schematic diagram of a signal content.

FIG. 10 is a diagram showing a result of processing the flowchart of FIG. 7;

FIG. 11 is a diagram showing a result obtained by processing the flowchart of FIG. 8;

FIG. 12 is a diagram showing server backup via a WAN.

[Explanation of symbols]

1-1, 1-2: server, 2: management device, 3-1, 3-
2 switching device, 5 client-side interface unit, 7 address conversion mechanism.

 ────────────────────────────────────────────────── ─── Continuing on the front page (72) Inventor Takeyuki Mayumi 890 Kashimada, Saiwai-ku, Kawasaki-shi, Kanagawa Prefecture F-term in the Information Systems Division, Hitachi, Ltd. (Reference) 5B045 BB11 GG01 JJ44 5B089 GA11 GA21 HA06 HB02 JB17 KA12 KB04 KB06 KC24 KC28 KH03 ME04 5K030 GA12 HD09 KA05 LB08 MD01 5K033 AA06 EA04 EB06 EC04

Claims (1)

[Claims] 1. A management system for managing a server / client connection state in a network system including a plurality of servers, wherein one switching device and a plurality of clients connected under the switching device are connected to one server. At least one device is connected to an arbitrary position on the network, and a first table holding an address of a server to which a client under the switching device is currently connected in the switching device; There is a second table that holds client addresses by grouping clients in connection server units and a third table that holds addresses of connection destination servers that change the connection of clients under the switching device to other servers. When a factor for changing the server to which the client connects is generated, the management device The first table contains the address of the server to which the client whose connection is to be changed is currently connected, and the second table contains the address of the client associated with the server set in the first table. In the table, the address of the connection destination server to be changed in connection is set, and for the transmission from the client to the server, according to the contents of the table, the switching device transmits the signal from the client shown in the second table. By converting the destination address from the server address shown in the first table to the server address shown in the third table, the server connection can be changed,
For the transmission from the server to the client, the switching device transmits the signal to the client shown in the second table among the signals of the server shown in the third table according to the contents of the table. A server switching method for a client-server system, wherein a response can be made by converting a source address from a server address shown to a server address shown in a first table.