JP2017173995A - Server device, cloud system, and web application switching method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable switchover to an update version of web application to be carried out in a short time in a server device having a public virtual IP and a private virtual IP allocated thereto.SOLUTION: A server device of the present invention includes an NAT table update unit 15 for updating the content of an NAT table stored in an NAT table storage unit 10 from first NAT information in which a first private virtual IP tied to a first node 301 and a public virtual IP are correlated to second NAT information in which a second private virtual IP tied to a second node 302 and the public virtual IP are correlated. An update version of web application is implemented in the second node 302 beforehand while an active version of web application is operating in the first node 301, so that switchover to the update version of web application is carried out by only updating the content of the NAT table.SELECTED DRAWING: Figure 2

Description

  The present invention relates to a server device, a cloud system, and a web application switching method, and in particular, a server device to which a public virtual IP address and a private virtual IP address are assigned, a cloud system using the server device, and a web application switching method in the server device. It is about.

  2. Description of the Related Art In recent years, a cloud system that provides a given service to a client by mounting a web application on a server device on the Internet and executing the web application in response to a request from the client has been widely used. In a cloud system, in order to realize a multi-tenant function, a virtual IP address system assigned to a server device is divided into public and private, and a public virtual IP address and a private virtual IP address are associated with each other to perform network address translation (NAT). By doing so, communication is possible.

  The public virtual IP address (hereinafter abbreviated as “public virtual IP”) is a virtual IP address for communicating with the public network on the client side (Internet side). The private virtual IP address (hereinafter abbreviated as private virtual IP) is a virtual IP address for communicating with a private network on the server side (LAN side). Private virtual IPs can be assigned freely, thereby making it easy to manage multi-tenancy.

  By the way, the web application needs to be updated appropriately for various reasons such as version upgrade and security measures. However, conventionally, switching for applying the updated version of the web application to the server device has been manually performed, so that the service has stopped for a relatively long time.

  In addition, when updating a web application, it is necessary to verify the legitimacy of the operation in advance. However, since the current version (production system) web application cannot be stopped for the verification, another verification system such as a verification system is required. In most cases, the environment is prepared to support the development of updated web applications. In this case, since the verification system is a different environment from the production system, when the updated version of the web application was migrated from the verification environment to the production environment, the verification system worked correctly. There was also a problem that it would not work properly.

  In addition, a technique capable of switching a server from an active server to a standby server is known (see, for example, Patent Document 1). In the network system described in Patent Document 1, an address of a specific server computer connected to the network is virtually set as an address of another server computer, and the server function of the specific server computer is set by the other server computer. Switch to an alternate server. In addition, a packet transmitted / received between a specific server computer and a client computer is converted into a packet transmitted / received between another server computer and a client computer whose servers are switched.

Japanese Patent Laid-Open No. 10-23074

  The present invention has been made to solve the above-described problems, and in a server device to which a public virtual IP and a private virtual IP are assigned, switching to an updated version of a web application is performed in a short time. The purpose is to be able to.

  In order to solve the above-described problem, in the present invention, a working version web application is implemented in a server device including a NAT table storage unit that stores a NAT table in which a public virtual IP address and a private virtual IP address are associated with each other. From the first NAT information in which the first private virtual IP address and the public virtual IP address associated with the first node are associated, the second node associated with the second node in which the updated version of the web application is implemented. The contents of the NAT table stored in the NAT table storage unit are updated to the second NAT information in which the private virtual IP address and the public virtual IP address are associated with each other.

  According to the present invention configured as described above, when the current version of the web application is running on the first node, the updated version of the web application is installed in advance on the second node, and then the NAT is executed. You can switch to an updated version of the web application simply by updating the contents of the table. In addition, since the updated version of the web application will continue to be used on the second node that has been verified for correctness before switching, the operating environment will not change after switching and will not cause any problems. Is never taken away. Thereby, the web application can be switched from the active version to the updated version in a short time.

It is a figure which shows the example of whole structure of the cloud system to which the server apparatus by 1st and 2nd embodiment is applied. It is a block diagram which shows the function structural example of the server apparatus by 1st Embodiment. It is a block diagram which shows the function structural example of the server apparatus by 2nd Embodiment. It is a sequence diagram which shows the operation example of the server apparatus by 1st Embodiment, and a 1st, 2nd node. It is a sequence diagram which shows the other operation example of the server apparatus by 1st Embodiment, and a 1st, 2nd node.

(First embodiment)
DESCRIPTION OF EXEMPLARY EMBODIMENTS Hereinafter, a first embodiment of the invention will be described with reference to the drawings. FIG. 1 is a diagram illustrating an example of the overall configuration of a cloud system to which the server device according to the first embodiment is applied. The cloud system illustrated in FIG. 1 includes a server device 100 according to the first embodiment, a client 200 connected to the public network on the Internet side with respect to the server device 100, and a private network on the LAN side with respect to the server device 100. The first and second nodes 301 and 302 are configured.

  The server apparatus 100 is a virtual server apparatus to which a public virtual IP address (hereinafter abbreviated as public virtual IP) and a private virtual IP address (hereinafter abbreviated as private virtual IP) are assigned. In the first embodiment, one public virtual IP is assigned to the server apparatus 100 and two private virtual IPs are assigned.

  On the other hand, the first and second nodes 301 and 302 are nodes to which private IP addresses (hereinafter abbreviated as private IPs) are assigned. For example, real server devices (Web servers) on which web applications are mounted It is. In the first embodiment, the first node 301 is a real server on which the current version of the web application is implemented, and the second node 302 is a real server on which the updated version of the web application is implemented. To do.

  Before the web application is switched, the client 200 accesses the first node 301 via the server device 100, and the first node 301 responds to a request from the client 200 according to a given service. I will provide a. The service provided here is executed by the current version of the web application.

  When the client 200 accesses the first node 301 via the server device 100, the server device 100 assigns the public virtual IP designated by the client 200 to the private IP (hereinafter, referred to as the first node 301). Communication between the two is realized by performing NAT for conversion to the first private IP.

  On the other hand, after the web application is switched, the client 200 accesses the second node 302 via the server device 100, and the second node 302 is located in response to a request from the client 200. Provide a given service. The service provided here is executed by an updated web application.

  When the client 200 accesses the second node 302 via the server device 100, the server device 100 assigns the public virtual IP designated by the client 200 to the private IP (hereinafter, referred to as the second node 302). Communication between the two is realized by performing NAT for conversion to the second private IP).

  FIG. 2 is a block diagram illustrating a functional configuration example of the server apparatus 100 according to the first embodiment. As shown in FIG. 2, the server apparatus 100 includes a NAT table storage unit 10 as a storage medium. In addition, the server device 100 includes a public side communication unit 11, a private side communication unit 12, an address conversion unit 13, a NAT table update unit 15, and an operation reception unit 16 as functional configurations.

  Each of the functional blocks 11 to 13, 15, and 16 can be configured by any of hardware, DSP (Digital Signal Processor), and software. For example, when configured by software, each of the functional blocks 11 to 13, 15, and 16 is actually configured by including a CPU, RAM, and ROM of a computer, and is stored in a recording medium such as a RAM, ROM, hard disk, and semiconductor memory. This is realized by operating the stored program.

  The NAT table storage unit 10 stores a NAT table in which a public virtual IP assigned to the server device 100 is associated with a private virtual IP. When using the first node 301 on which the current version of the web application is mounted, the NAT table storage unit 10 includes a public virtual IP and a first private virtual IP associated with the first node 301. The associated first NAT table (corresponding to the first NAT information in the claims) is stored. The first private IP is associated with the first private virtual IP.

  On the other hand, when the state is switched to the state in which the second node 302 on which the updated version of the web application is mounted is used, the NAT table storage unit 10 stores the public virtual IP and the second node associated with the second node 302. A second NAT table (corresponding to the second NAT information in the claims) associated with the private virtual IP is stored. The second private virtual IP is associated with the second private IP.

  The public-side communication unit 11 communicates with the client 200 using a public virtual IP. The private-side communication unit 12 communicates with the first node 301 or the second node 302 using a private virtual IP. Normally, the private-side communication unit 12 performs communication with the first node 301 on which the working version web application is mounted.

  The address conversion unit 13 performs address conversion between the public virtual IP and the private IP based on the NAT table stored in the NAT table storage unit 10. When using the first node 301 on which the current version of the web application is implemented, the address translation unit 13 uses the public virtual IP and the public virtual IP based on the first NAT table stored in the NAT table storage unit 10. Address conversion is performed with the first private IP.

  On the other hand, when switching to a state in which the second node 302 in which the updated version of the web application is installed is used, the address conversion unit 13 is based on the second NAT table stored in the NAT table storage unit 10. The address is converted between the public virtual IP and the second private IP.

  The NAT table update unit 15 obtains an updated version of the web from the first NAT table that associates the first private virtual IP and the public virtual IP associated with the first node 301 on which the current version of the web application is implemented. The contents of the NAT table stored in the NAT table storage unit 10 to the second NAT table in which the second private virtual IP and the public virtual IP address associated with the second node 302 on which the application is mounted are associated with each other. Update.

  The updating of the NAT table by the NAT table updating unit 15 is executed when the operation receiving unit 16 receives a user operation for making a web application switching request. As a premise for the switching, when the first node 301 on which the current version of the web application is mounted is used, the user has mounted the updated version of the web application on the second node 302 in advance. A second NAT table is also created in advance.

  After the updated web application is mounted on the second node 302 and the verification is completed, the user inputs the second NAT table to the server device 100 and instructs the switching of the web application. Perform the operation. When the operation accepting unit 16 accepts this user operation, the NAT table updating unit 15 receives a web application switching request from the operation accepting unit 16 and stores the first NAT table stored in the NAT table storage unit 10. Rewrite the input second NAT table.

  After the NAT table stored in the NAT table storage unit 10 is rewritten, the access from the client 200 is performed to the second node 302 instead of the first node 301. Thus, the service is provided to the client 200 by the updated version of the web application.

  As described above in detail, in the first embodiment, in the server device 100 to which the public virtual IP and the private virtual IP are allocated, the first node associated with the first node 301 in which the active version web application is mounted. Corresponding the second private virtual IP and the public virtual IP associated with the second node 302 on which the updated version of the web application is mounted from the first NAT table in which the private virtual IP and the public virtual IP are associated with each other. The contents of the NAT table stored in the NAT table storage unit 10 are updated to the second NAT table.

  According to the first embodiment configured as described above, the first version of the web application installed in the first node 301 can be updated only by updating the contents of the NAT table in the NAT table storage unit 10. During the operation of the node 301, it is possible to switch to the updated version of the web application that is previously installed in the second node 302. In addition, since the updated version of the web application is continuously used in the second node 302 whose operation has been validated before switching, the operation environment does not change after switching and causes a problem. No time is taken. Thereby, the web application can be switched from the active version to the updated version in a short time.

(Second Embodiment)
Next, a second embodiment of the present invention will be described with reference to the drawings. The second embodiment is for solving the problem that the NAT table cannot be switched due to the influence of the HTTP keep-alive function. The keep alive function is a function that maintains a TCP session connected to the first node 301 in response to an access from the client 200 even after a response to the client 200 is made.

  When the keep alive function is off, a pair of HTTP request and response is performed in one TCP session. In other words, after establishing a session with the first node 301, when a response from the first node 301 is returned to the client 200 via the server device 100 in response to a request from the client 200, the session is disconnected. Therefore, when a plurality of requests are made from the client 200 to the first node 301 via the server device 100, session establishment and disconnection are repeated each time.

  On the other hand, when the keep alive function is on, a plurality of combinations of requests and responses can be executed in one session. That is, after establishing a session with the first node 301, when a predetermined timeout has passed without a request from the client 200, or when the total number of requests reaches a predetermined number of times, the first node 301 By transmitting a FIN packet to the client 200 via the server device 100, the session is disconnected.

  As described above, when the keep alive function is valid, a plurality of requests are processed by one TCP connection, and therefore the same TCP connection is continuously used until a time-out occurs or the maximum number of requests is exceeded. For this reason, even if the NAT table stored in the NAT table storage unit 10 is rewritten while the TCP connection is continued, the change is not reflected.

  In other words, it is desired to switch communication from the first node 301 to the second node 302 by dynamically rewriting the NAT table stored in the NAT table storage unit 10, but only when the TCP session is established. Since the specification is to evaluate the above address translation rule, the update of the NAT rule is not applied to an established session in which the keep-alive function is valid. When the keep-alive function is valid, the session establishment is performed again when the timeout period elapses or when the maximum number of requests is reached. Therefore, the NAT rule is switched at that time, but there is a problem that the intended timing is not reached. In the second embodiment, this problem is solved by the means described below.

  FIG. 3 is a block diagram illustrating a functional configuration example of the server apparatus 100 ′ according to the second embodiment. In FIG. 3, those given the same reference numerals as those shown in FIG. 2 have the same functions, and thus redundant description is omitted here. As shown in FIG. 3, the server device 100 ′ according to the second embodiment further includes a session control unit 14. Further, a NAT table update unit 15 ′ is provided instead of the NAT table update unit 15.

  The session control unit 14 disconnects the session maintained by the keep-alive function when the operation receiving unit 16 receives a user operation for making a web application switching request.

  Specifically, when the first node 301 in which the current version of the web application is mounted is used (the state before the web application is updated), the first node 301 is sent from the client 200 after the session is established. The keep alive function is turned on by accepting the keep alive setting request included in the incoming request.

  Thereafter, when the NAT table update unit 15 ′ updates the contents of the NAT table in response to a user operation for making a web application switching request, the session control unit 14 first disconnects the session and rejects keep-alive acceptance. 1 node 301 is instructed. In response to this instruction, the session control unit 14 transmits the FIN packet output from the first node 301 to the client 200 to disconnect the session in which the connection is maintained.

  Note that the session controller 14 instructs the first node 301 to reject keepalives, and a new request from another client arrives before the NAT table update performed by the NAT table updater 15 ′. In this case, the session is disconnected each time by setting the keep alive invalid operation.

  The NAT table update unit 15 ′ rewrites the first NAT table stored in the NAT table storage unit 10 to the second NAT table after the session is disconnected by the session control unit 14.

  After rewriting the NAT table, the session is established again by accessing the server apparatus 100 'from the client 200. At this time, since the contents of the rewritten second NAT table in the NAT table storage unit 10 are reflected as a new NAT rule, a session is established with the second node 302. Thereafter, in response to the keep alive setting request included in the request sent from the client 200, the second node 302 permits the request, so that the keep alive function is turned on again. Thereafter, communication with the second node 302 is performed.

  FIG. 4 is a sequence diagram showing an operation example of the server device 100 ′ and the first and second nodes 301 and 302 according to the second embodiment configured as described above. If the client 200 accesses the server device 100 ′ when the first node 301 on which the current version of the web application is installed is used, a session is established between the client 200 and the first node 301. Is established (step S1). At this time, the contents of the first NAT table stored in the NAT table storage unit 10 are evaluated and applied as a NAT rule.

  Thereafter, a request including a keep alive setting request is transmitted from the client 200 to the first node 301 via the server device 100 ′ (step S2), and the first node 301 accepts the keep alive for the request. By returning the response including the response, the keep alive function is turned on (step S3). Thereafter, a plurality of pairs of requests and responses are executed between the client 200 and the first node 301 within the session established in step S1 (steps S4 to S7).

  Thereafter, when the operation receiving unit 16 receives a user operation for making a web application switching request, the session control unit 14 starts preparation for updating the NAT table (step S8). Then, the session control unit 14 instructs the first node 301 to disconnect the session and reject the keep-alive acceptance (step S9).

  Next, the session control unit 14 disconnects the session in which the connection is maintained by transmitting the FIN packet output from the first node 301 to the client 200 in response to the above instruction (step S10). After the session disconnection, the NAT table update unit 15 'rewrites the first NAT table stored in the NAT table storage unit 10 with the second NAT table (step S11).

  When the client 200 accesses the server apparatus 100 ′ after rewriting the NAT table, the content of the second NAT table rewritten in the NAT table storage unit 10 is evaluated and applied as a NAT rule. And a second node 302 are established (step S12).

  Thereafter, a request including a keep alive setting request is transmitted from the client 200 to the second node 302 via the server device 100 ′ (step S13), and the second node 302 accepts the keep alive for the request. Is returned to the response, the keep-alive function is turned on (step S14). Thereafter, a plurality of sets of requests and responses are executed between the client 200 and the second node 302 within the session established in step S12 (steps S15 to S16).

  As described above in detail, according to the second embodiment, even when the HTTP keep-alive function is set to ON, the NAT is performed without delay after the user performs the web application switching request operation. Communication can be switched from the first node 301 to the second node 302 by reflecting the rewritten second NAT table in the table storage unit 10.

  In the second embodiment, the example has been described in which the keep-alive function is always turned on and the keep-alive function is temporarily turned off when the web application is switched. However, the present invention is not limited to this. For example, the keep alive function may be always turned off by rejecting a keep alive setting request in a request sent from the client 200 after the session is established.

  FIG. 5 is a sequence diagram illustrating an operation example of the server apparatus 100 ′ and the first and second nodes 301 and 302 in this case. When the keep alive function is always turned off, the session is established and disconnected every time the client 200 accesses the server apparatus 100 'as in steps S21 to S24 and steps S25 to S28 in FIG. In this case, the FIN packet for session disconnection is transmitted from the client 200 side each time the client 200 receives a response including a keep-alive acceptance refusal from the server device 100 ′.

  When the user performs a web application switching request operation in such a state, the session is disconnected at that time, or even if connected, the NAT table update unit 15 ′ rewrites the NAT table. Can be performed immediately (step S29).

  When the client 200 accesses the server apparatus 100 ′ after rewriting the NAT table, the content of the second NAT table rewritten in the NAT table storage unit 10 is evaluated and applied as a NAT rule. And a second node 302 are established (step S30).

  Thereafter, a request including a keep alive setting request is transmitted from the client 200 to the second node 302 via the server device 100 ′ (step S31), and the second node 302 accepts the keep alive for the request. By returning a response including rejection, the keep alive function is turned off (step S32). Therefore, when a pair of a request and a response is executed between the client 200 and the second node 302 within the session established in step S30 (steps S31 and S32), the FIN packet is transmitted from the client 200. Is sent and the session is disconnected (step S33).

  In the first and second embodiments described above, an example in which the first and second nodes 301 and 302 are Web servers has been described, but the present invention is not limited to this. For example, the first and second nodes 301 and 302 may be load balancers or cluster servers.

  In addition, each of the first and second embodiments described above is merely an example of a specific example for carrying out the present invention, and the technical scope of the present invention should not be interpreted in a limited manner. It will not be. That is, the present invention can be implemented in various forms without departing from the gist or the main features thereof.

DESCRIPTION OF SYMBOLS 10 NAT table memory | storage part 14 Session control part 15,15 'NAT table update part 100,100' Server apparatus 200 Client 301 1st node 302 2nd node

Claims (7)

A server device to which a public virtual IP address and a private virtual IP address are assigned,
A NAT table storage unit for storing a NAT table in which the public virtual IP address and the private virtual IP address are associated with each other;
The updated version of the web application is implemented from the first NAT information that associates the first private virtual IP address associated with the first node on which the current version of the web application is implemented with the public virtual IP address. Update the contents of the NAT table stored in the NAT table storage unit to the second NAT information in which the second private virtual IP address associated with the second node is associated with the public virtual IP address. A server apparatus comprising a NAT table update unit. Session control unit that disconnects the session that is maintained when a web application update request is received when the keep-alive function that maintains the session connected with the client after the response is turned on. Further comprising
The server apparatus according to claim 2, wherein the NAT table update unit updates the content of the NAT table after the session is disconnected by the session control unit.   The server apparatus according to claim 2, wherein the session control unit instructs the first node to reject keepalive acceptance when the session is disconnected. A cloud system comprising a server device to which a public virtual IP address and a private virtual IP address are assigned, and a plurality of nodes connected to the server device on the private network side,
The server device is
A NAT table storage unit for storing a NAT table in which the public virtual IP address and the private virtual IP address are associated with each other;
The updated version of the web application is implemented from the first NAT information that associates the first private virtual IP address associated with the first node on which the current version of the web application is implemented with the public virtual IP address. Update the contents of the NAT table stored in the NAT table storage unit to the second NAT information in which the second private virtual IP address associated with the second node is associated with the public virtual IP address. A cloud system comprising a NAT table update unit. Session control unit that disconnects the session that is maintained when a web application update request is received when the keep-alive function that maintains the session connected with the client after the response is turned on. Further comprising
The cloud system according to claim 4, wherein the NAT table update unit updates the content of the NAT table after the session is disconnected by the session control unit.   The first node and the second node reject the keep alive setting request included in the request sent from the client after the session is established, so that the keep alive function is always turned off. The cloud system according to claim 4, wherein: In a server device to which a public virtual IP address and a private virtual IP address are assigned, a method for switching a web application to be applied,
A first step of determining whether or not the NAT table updating unit of the server device has accepted the web application switching request;
When the NAT table update unit of the server device accepts the switching request, the first private virtual IP address associated with the first node on which the current version of the web application is mounted is associated with the public virtual IP address. From the attached first NAT information to the second NAT information in which the second private virtual IP address associated with the second node on which the updated version of the web application is mounted is associated with the public virtual IP address. And a second step of updating the contents of the NAT table stored in the NAT table storage unit.