JP2012014248A - System, method and program for software distribution corresponding to scale-out configuration - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a software distribution system and the like capable of appropriate software distribution (update) and scale-out even if there are standby machines for scale-out and the like.SOLUTION: Distribution object machines 110 include operating machines 111 and standby machines 112. The software distribution system 101 includes: first means for managing configuration information including the states of the machines 110 and software configuration; second means for distributing software 150 to the machines 110; and third means for operating the booting and stopping states of the machines 110. In the distribution, the software distribution system boots the standby machines 112 in a stopped state based on the configuration information, updates the software of the respective machines 110 and stops the standby machines 112 in a booting state. The software of the respective machines 110 is updated at a same time.

Description

  The present invention relates to a technology such as a software distribution system, and in particular, distribution (update) of software to a computer system or the like (including a scale-out machine or the like) having a scale-out configuration in changing (updating) the software configuration of a computer system. The present invention relates to a technology that is effective when applied to a system that enables the

  In a distributed / parallel processing computer system composed of a plurality of machines, the performance of the entire machine group can be improved by increasing the number of machines constituting the computer system. This is hereinafter referred to as “scale-out”. For the sake of explanation, the state where scale-out is performed (when scale-out is executed) is “ON”, and the state where scale-out is not performed (when scale-out is not executed) is “OFF”. Machines prepared for scale-out (“standby machine”) are scaled out based on conditions such as when the amount of processing in the computer system exceeds a certain standard (high load state) It is activated only during the interval (ON), and is stopped while the scale-out is not performed (OFF) based on conditions such as when the amount of processing performed by the computer system returns to a certain standard (low load state).

  On the other hand, the OS, applications, and other software (hereinafter collectively referred to as “software”) installed in the machine need to be installed with the latest version with bug fixes, security hole corrections, etc., for example. . If you keep using a version of software that still has security holes, not only the computer system that runs worst, but also the computer system that runs on the latest version may be disturbed. Therefore, it is necessary to quickly upgrade (update) the machine software without omission. A software distribution system is a system in which operations and processes such as software update are systemized. That is, it is possible to automatically update the distribution target system by distributing the latest software to the distribution target machine.

  As an example of the prior art, there is JP-A-4-268663 (Patent Document 1). Patent Document 1 describes data backup / synchronization between an active system and a standby system machine. Specifically, a cold standby computer system (standby machine) is periodically started up, and the cold standby computer system (standby) is backed up by backup information (backup data) of the online computer system (active machine). It is described that it is possible to update the data of the (system machine).

JP-A-4-2686633

  When the software of a machine group is updated by a conventional general software distribution system, it is not applied to a stopped machine, and an update omission occurs. For example, in a standby computer in a stopped state in a computer system having a conventional scale-out configuration, update omission occurs.

  In the above case, because the software configuration of the machine group is different due to omission of update, when performing scale out (ON), before that, by starting the standby machine in the stopped state and updating the software, It is necessary to perform a synchronization procedure to make the software configuration of the standby machine the same as that of the active machine (requires user operation or the like). Until this synchronization procedure is completed, the next procedure, scale-out (ON), cannot be performed.

  As a method for realizing the above-mentioned synchronization, the software configuration of the standby machine is obtained by periodically acquiring (copying) data from the active machine and updating (reflecting) it on the standby machine as in Patent Document 1. Can be synchronized with the active machine.

  However, when the software configuration is synchronized between the active and standby machines using the above synchronization method, the latest data (that is, the software of the active machine) and the previous (last) data at the time of scale-out (ON) There may be a difference between the data acquired at the time of synchronization (that is, the software of the standby machine). In other words, there is a difference between the time when the software of the active machine is updated and the time when data (software) is synchronized from the active machine to the standby machine. Is not applied, there occurs a time when there is a difference in software configuration between the active system and the standby system machine as described above (FIG. 7A).

  In the above case, when the scale-out is performed (ON), there is a possibility that the time when the above difference occurs is overlapped. In that case, because there is a problem due to the difference in the software configuration between the active system and the standby system (security and other problems described above), it is necessary to separately work to make the software configuration of the standby system the latest (same as the active system machine). (This requires user operation). There is a problem that the overload state of the computer system is prolonged due to such work and its time. Further, by simply copying the active system data (software) to the standby system by the synchronization method as described in Patent Document 1, it is necessary to change software settings for operating as a separate machine (standby system machine). There is a case.

  As described above, in the prior art, when the distribution (update) of software to a scale-out computer system (including active and standby machines) is considered, the software configuration of each machine is the latest (same as the synchronization etc.) ), There is a problem that a time (lead time) required for performing scale-out (ON) without any problem and a work burden on the user (operator, etc.) are generated.

  In addition, as a means for avoiding the above lead time from occurring during the scale-out (ON), the target machine is started before software distribution by the user's manual operation, or the target machine is stopped after software distribution. It is possible to do. However, in that case, there is a problem that the work burden on the user is increased because separate work is required for each software distribution and each target machine.

  In view of the above, the object of the present invention is to eliminate the need for a special work load by the user even when there is a standby machine for scale-out in a computer system to be distributed (updated), Another object of the present invention is to provide a system that can appropriately distribute (update) and scale out software with little or no lead time.

  In order to achieve the above object, the present invention is a technology such as a software distribution system, method, and program for distributing (updating) software to a computer system (machine group) such as a scale-out configuration. It is characterized by having.

  The system of the present invention enables automatic distribution (updating) to a standby machine for scale-out (stopped when turned off, activated when turned on), etc. when distributing (updating) the latest software, That is, the software configuration between the active system and the standby system machine can be maintained to be the same (latest), and appropriate scale-out (ON / OFF) is enabled.

  This software distribution system is, for example, a software distribution system that performs processing for distributing and updating software to a group of machines of a target computer system. The target machine group includes, for example, a first machine that is activated (for example, an active machine) and a second machine that is stopped (for example, a standby machine). The software distribution system includes a first means for managing configuration information including a state of a machine group and a software configuration, a second means for executing software distribution to the machine group, and a start and stop state of the machine group. And third means for operation (control). When executing software distribution by the second means, the software distribution system starts the stopped second machine by the third means based on the configuration information, and includes each of the first and second machines. Software data information is transmitted to the machine to update the software of each machine. After the update, the second machine in the activated state is stopped by the third means, and the result is converted to the configuration information by the first means. Process to reflect.

  With the above configuration, by automatically updating the software of each target machine at the same time (approximately the same timing), the software configuration of each machine can be maintained to be the same (latest), and the user's work load Appropriate software distribution (update) and scale-out can be realized with little or no lead time.

  According to the present invention, even when there is a standby machine for scale-out in a computer system to be distributed (updated), there is no need for a special work load by the user and there is no lead time. Alternatively, software distribution (update) and scale-out can be performed appropriately.

It is a figure which shows the structural example of the whole computer system containing the software distribution system of one embodiment of this invention, its distribution object machine, etc. (management object). It is a figure which shows the detailed content example of the machine information stored in the structure information storage part in this software distribution system. It is a figure which shows the example of the detailed content of the software installation information stored in the structure information storage part in this software distribution system. It is a figure which shows the example of the change location of the software installation information before (a) distribution before (b) distribution in this software distribution system. It is a flowchart of the software distribution process (distribution control process) in this software distribution system. It is a figure which shows the transition table of (a) active machine and (b) standby machine in this software distribution system. (A) A prior art example, (b) It is a figure which shows typically the example of the timing and state transition regarding software distribution (update) of various machines of a computer system, scale-out, etc. in the software distribution system of this Embodiment. .

  Hereinafter, a software distribution system according to an embodiment of the present invention will be described in detail with reference to the drawings. It should be noted that “distribution” of software includes the meaning of “update”.

[Overview]
The software distribution system 101 (FIG. 1, etc.) according to the present embodiment manipulates the state of each machine 110 including the standby machine 112 for scale-out, such as starting and stopping, for the distribution / update of the software 150 ( Control means (machine operation unit 104, distribution agent 120, etc.). In addition, the configuration information related to the configuration including the software version of each machine 110 is managed (distribution control unit 102, configuration information storage unit 105, etc.), and each machine of the target computer system including the standby machine 112 in the stopped state 110 has a function of constantly grasping the software configuration. As a result, at the time of distribution accompanying the upgrade of the software 150 or the like, the standby system machine 112 in the stopped state that needs to be activated is activated, and the active machine 111 that is already activated is activated. In each machine 110 including the standby machine 112, the software 150 (151 and 152) can be updated at the same time. As a result, the software configuration of the active machine 111 and the standby machine 112 can always be maintained without any difference. In this system, control related to the above distribution / update is automatically realized without the need for special work by a user (operator, etc.). In this system, since there is no difference in the software configuration at scale-out (ON / OFF), the lead time can be eliminated or reduced.

[system]
FIG. 1 shows an example of the overall configuration of a computer system including a software distribution system 101 according to an embodiment of the present invention and a software distribution target machine 110 that is the management target. The entire computer system has a machine (software distribution machine) 100 including a software distribution system 101 and a machine 110 (target machine (group)) of a target (software distribution target) computer system, such as a network. Connected through means. The machine 100 (software distribution system 101) connects (or incorporates) the configuration information storage unit 105. The target computer system includes one or more active machines 111 and one or more standby machines 112 as a plurality of machines 110. Each machine 110 (111, 112) includes a distribution agent 120 (121, 122) which is a part of the software distribution system 101. Each machine 110 (111, 112) is connected to a load balancer 130 corresponding to a scale-out configuration.

  The machine 100 includes software 150 (data information for update, etc.) to be distributed and managed. Each target machine 110 (111, 112) has software 151, 152 installed based on distribution management of the software 150. The configuration information storage unit 105 is configured by a database, storage, and the like, and includes machine information 200 (FIG. 2), software installation information 300 (FIGS. 3 and 4), and the like as configuration information to be stored.

  The software distribution system 101 includes (1) a distribution control unit 102, (2) a distribution execution unit 103, and (3) a machine operation unit 104 as components (program modules and the like). The software distribution system 101 according to the present embodiment is realized mainly by program processing using these components on the machine 100.

  The distribution control unit 102 has a function of comprehensively controlling software distribution to the machine 110 corresponding to the scale-out configuration, controls other parts (103, 104, etc.), and manages configuration information for the configuration information storage unit 105 ( Process). The distribution control unit 102 implements the above control by executing the flow of the distribution control process shown in FIG. The distribution control unit 102 always grasps the attributes and states of each machine 110, the software configuration, and the like, and manages the configuration information in the configuration information storage unit 105.

  The distribution execution unit 103 performs processing for executing distribution of the software 150 (151 and 152) in cooperation with the distribution agent 120 (121 and 122) installed in the target machine 110 (111 and 112). In accordance with the distribution control unit 102, the distribution execution unit 103 has a function of transmitting the target software 150 (update data information and the like) to the distribution agent 120 of the target machine 110, and the machine 110 performed by the distribution agent 120. A function of receiving (acquiring) execution result information such as installation and update of software and transmitting the information to the distribution control unit 102.

  The machine operation unit 104 has a function of performing operations (control) such as starting and stopping of each machine 110 related to software distribution remotely through a network or the like in cooperation with the distribution agent 120. This also grasps the state of each machine 110 such as starting and stopping.

  The distribution agent 120 (121, 122) is installed in advance on the target machine 110 (active machine 111, standby machine 112) and cooperates with the software distribution system 101 (distribution execution unit 103 or machine operation unit 104). The distribution agent 120 is realized by program processing on the machine 110 and has the following functions. (1) When the machine 110 is activated, information indicating completion (status) of activation is transmitted to the machine operation unit 104 (that is, a function for grasping the activation or deactivation status of the machine 110). (2) During distribution of the software 150, the software 150 (update data information, etc.) received from the distribution execution unit 103 is installed or updated in the machine 110. (3) Information on the result of execution of the distribution (update completion or the like) is transmitted to the distribution execution unit 103 (that is, a function for grasping the software configuration of the machine 110).

  The example of FIG. 1 is a system configuration that uses the load balancer 130. The machine 110 is, for example, a server device that provides service processing to a client terminal or the like. The load distribution device 130 distributes the load by distributing access to the machine 110 from a client terminal or the like according to a load or a scale-out situation.

  In the example of FIG. 1, for the sake of simplicity, one active machine 111 and one standby machine 112 are shown as the machines 110. That is, it is a case where scale-out (ON / OFF) in units of one unit is possible. However, the present invention is not limited to this, and a plurality of machines 111 and 112 may be used to enable scale-out in units of a plurality of machines.

  The present software distribution system 101 has the above-described functions (software distribution function corresponding to the scale-out configuration), so that each machine 110 can be an active machine 111 that is always started or normal (when scale-out is OFF). Identifies whether the standby machine 112 is in a stopped state, and enables automatic operation (control) of starting and stopping of the machine 110.

[Configuration management information]
Examples of configuration information stored in the configuration information storage unit 105 managed by the software distribution system 101 include machine information 200 illustrated in FIG. 2 and software installation information 300 illustrated in FIGS. 3 and 4.

  FIG. 2 shows a detailed content example of the machine information 200 for the machine 110 to be updated. The machine information 200 in FIG. 2 includes (1) machine identifier (ID), (2) machine name (mashine_name), (3) attribute (mashine_attribute), (4) status (status), and the like. The attributes of (3) include (1) “normal”, (2) “standby”, and the like. “Normal” indicates that the system is always activated in the active system, that is, in the normal business, regardless of the scale-out state (ON / OFF). “Standby” indicates that the system is activated only in the standby system, that is, in scale-out (ON). The state of (4) indicates a state relating to the current operation or stop of the machine 110. In this embodiment, (1) “Start A”, (2) “Start B”, (3) “Stop”, Etc. “Activation A” (normal business activation) represents a state of activation for normal business or the like. “Activation B” (maintenance activation) represents an activated state for maintenance including software distribution. “Stopped” represents a stopped state (for example, a power-off state or a sleep state).

  FIG. 3 shows an example of detailed contents of the software installation information 300. 3 includes (1) installed software information (identifier), (2) machine identifier (mashine_ID), (3) software name (software), (4) version information (version), and the like. . (1) is information (identifier) related to software installed in the machine. The machine identifier (2) is associated with the machine identifier (1) of the machine information 200.

  That is, by managing the configuration information as shown in FIGS. 2 to 3, the configuration including the version of the software (151, 152) provided in each machine 110 of the target computer system is grasped as needed.

  FIG. 4 shows an example of content change (update) by software distribution in the software installation information 300 of FIG. (A) is before distribution, (b) is after distribution.

  First, in (a), in this example, the active machine 111 in FIG. 1 has a machine identifier “1”, and the standby machine 112 has a machine identifier “2”. Each of the active machine 111 and the standby machine 112 has “Web server_1” installed as software, and its version is “2.3”.

  And as shown in (b), this example shows a case where the version of “Web server_1” of each machine 110 is updated from “2.3” to “2.4” by executing software upgrade and distribution. ing. With the distribution operation by the distribution execution unit 103 or the like, the version value of the software installation information 300 is updated from “2.3” to “2.4” as illustrated.

[Processing flow]
Next, with reference to FIG. 5, a software distribution process (distribution control process) in the software distribution system 101 will be described along with the operation of each component (S indicates a processing step or the like).

  First, in step S <b> 501, the distribution control unit 102 acquires the machine information 200 of the machine 110 to be distributed from the software installation information 300 in the configuration information storage unit 105. In the loop from S502 to S511, the processing for the number of the acquired target machines 110 is repeated in the same manner.

  In S <b> 503, the distribution control unit 102 checks and determines the attribute (attribute) and status (status) of the target machine 110 based on the machine information 200. In the determination in S503, if the attribute of the machine 110 is “standby” and the state is “stopped” (that is, the standby machine 112 in the stopped state), the process proceeds to S504, and the machine information 200 is transferred to the machine operation unit 104. To pass.

  In S504, the machine operating unit 104 activates the machine 110 (standby machine 112). After startup, in S505, the state of the machine 110 is changed (transitioned) from “stop” to “start B” (maintenance start). That is, the distribution control unit 102 updates the state of the machine 110 in the machine information 200 of the configuration information storage unit 105 to “activation B”. Then, the machine information 200 is transferred to the distribution execution unit 103, and the process proceeds to S506.

  An example of the contents of the machine activation process by the machine operation unit 104 in S504 is as follows. First, the machine operation unit 104 receives (acquires) the machine information 200 transmitted from the distribution control unit 102. Next, the machine operation unit 104 activates (starts up maintenance) the machine 110 to be activated. In the present embodiment, an instruction for activation (for example, power-on or return from sleep) is transmitted from the machine operation unit 104 of the machine 100 to the distribution agent 120 of the target machine 110. Accordingly, the distribution agent 120 activates the machine 110.

  Then, the distribution agent 120 of the target machine 110 transmits information on the result of the machine operation such as the start completion to the machine operation unit 104. The machine operation unit 104 waits for and receives the result information from the distribution agent 120 of the machine 110. Then, the result is notified from the machine operation unit 104 to the distribution control unit 102 (leading to S505). As a result, the “start-up B” state of the target machine 110 (that is, a state where software distribution is possible) or the like is confirmed.

  Further, when the target machine 110 exists in the virtual environment, the processing in S504 may be as follows. In this case, the distribution agent 120 is not required to start the machine 110, and can be realized (started) by an API (command) provided by software that creates a virtual environment.

  On the other hand, if it is determined in S503 that the attribute and state are other (other than “standby” and “stop”), the process proceeds to S506, and the machine information 200 of the machine 110 is passed to the distribution execution unit 103.

  In S <b> 506, the distribution execution unit 103 executes software distribution to the target machine 110. That is, the distribution execution unit 103 of the distribution machine 100 distributes (sends) data information for installation / update of the target software 150 (latest version) to the distribution agent 120 of the target machine 110 via the network or the like. To do. The distribution agent 120 of the target machine 110 installs / updates the target software (151, 152) of the target machine 110 using the received data information. Then, information on the result of distribution execution such as completion of the update is transmitted to the distribution execution unit 103 of the distribution machine 100. Note that the above is the case of the distribution execution process in units of one unit. However, when there are a plurality of units, the distribution execution process may be simultaneously performed on a plurality of units.

  After executing the distribution in S506, in S507, the distribution control unit 102 updates the contents of the software installation information 300 in the configuration information storage unit 105 according to the information on the result (update completion, etc.) received from the distribution execution unit 103. An example of the update contents is as shown in FIG. 4. For example, the software of both the active machine 111 (“Startup A” state) and the standby machine 112 (“Startup B” state) is updated to the same latest version. It will be.

  In step S <b> 508, the distribution control unit 102 determines the status (status) of the target machine 110 in the machine information 200. When the state of the machine 110 is “start-up B” (maintenance start-up) (that is, in the case of the standby machine 112 started up in S504), the process proceeds to S509, and the machine information 200 is passed to the machine operation unit 104.

  In S509, the machine operation unit 104 stops the machine 110. This stop process is performed by, for example, sending and receiving a stop instruction or the like using the distribution agent 120 as in S504. Then, based on the result of the stop in S509, in S510, the distribution control unit 102 updates the state of the machine 110 in the machine information 200 of the configuration information storage unit 105 from “start B” to “stop”.

[Case]
FIGS. 7A and 7B schematically show examples of timings and state transitions related to software distribution and scale-out of various machines of a computer system in this embodiment.

  (A) In the prior art example, first, the active machine ("started" state) and the standby machine ("stopped" state) have the same version (v1) software configuration and are not scaled out. (OFF) state. Next, it is assumed that distribution of the latest version (v2) software is executed for each machine at a certain time (t1) (t1 to t2). At this time, the distribution is applied to the active machine in the “started” state, but is not applied to the standby machine in the “stopped” state. The active machine has the latest configuration (v2) from the end of distribution (t2), but the standby machine remains in the old configuration (v1).

  Further, at a certain time (t4), it is assumed that synchronization at a regular timing is performed (t4 to t5) by the technique of Patent Document 1 or the like. As a result, by copying (reflecting) the data of the active machine (v2) to the standby machine (v1) in the “started” state, the software configuration (v1) of the standby machine is the same as that of the active machine Updated (synchronized) to (v2).

  In the above example, a time (t2 to t4) in which there is a difference (deviation) between the software configurations of the active system and the standby system is generated as in 701.

  In addition, when scale-out (ON) is performed based on conditions (such as a high load state of a computer system) at a certain time (t3) during the time (t2 to t4), there is a problem because there is a difference in the software configuration. There is. Therefore, it is necessary to update the software configuration (v1) of the standby machine by a user manual operation or the like.

  (B) In this embodiment, first, the active machine 111 and the standby machine 112 have the same version (v1) software configuration and are not scaled out (OFF), that is, the active machine Assume that 111 is in the “startup A” state, and the standby machine 112 is in the “stopped” state. Next, it is assumed that distribution of the latest version (v2) software is executed to each machine 110 at a certain time (t1). At this time, according to the distribution control (FIG. 5) described above, the standby machine 112 is activated for maintenance (transition from “stop” to “start B” state). The active machine 111 remains in the “startup A” state. Then, the software (v2) is distributed / updated for both (111, 112) (t1-t2). After the update is completed (t2), the active machine 111 returns to the “start A” state, and the standby machine 112 returns to the “stop” state.

  If the scale-out (ON) is performed based on conditions (such as a high load state of the computer system) at a certain time (t3) thereafter, the standby machine 112 is transitioned to the “startup A” state.

  In (b) above, since there is no time (701) in which a difference in software configuration occurs between the active system and the standby system as in (a), there is no problem associated with scale-out (ON / OFF) as in (a). .

  FIG. 6 shows (a) a transition table of active machines and (b) a transition table of standby machines in the present embodiment. Examples of the transition destination state to be transitioned according to the state (status) and condition of the transition source are collectively described. For example, in (a), the active machine is normally activated at all times and basically does not undergo state transition. Even in the stopped state, it is considered that there is some cause, and the transition to the activated state is not performed even during software distribution. In (b), the standby machine is not scaled out (ON / OFF) during software distribution.

[Effects]
As described above, according to the software distribution system 101 of the present embodiment, the user (operator or the like) distributes the software 150 to the computer system (machine 110) having the standby machine 112 having the scale-out configuration. Software distribution and scale-out can be performed appropriately without requiring a special work load (for example, the procedure such as synchronization described above) and lead time.

  The user does not need to be aware of the scale-out machine etc. (whether the system is active or standby), and can automatically and collectively select the target by simple operation such as selecting the software 150 to be updated. The software 150 can be distributed / updated to all the machines 110 that require the computer system (including the standby machine 112 in the stopped state). In this software distribution system 101, a scale-out configuration system can be handled in the same manner as other systems (non-scale-out configuration, etc.), and the software configuration of each machine 110 is the same at the time of scale-out execution (ON). Therefore, appropriate scale-out (ON / OFF) can be performed immediately without lead time.

  Further, in this system, the types of activation are managed separately, such as “activation A” and “activation B”. Thereby, even when software distribution is performed during the scale-out (ON) in which the standby machine 112 is operating, the standby machine 112 can return to the state before distribution.

  As mentioned above, the invention made by the present inventor has been specifically described based on the embodiment. However, the present invention is not limited to the embodiment, and various modifications can be made without departing from the scope of the invention. Needless to say.

  INDUSTRIAL APPLICABILITY The present invention can be effectively used for software update particularly in an environment in which a predetermined computer system is operated, such as a corporate information processing system and a server system in a cloud environment, particularly when having a scale-out configuration.

  DESCRIPTION OF SYMBOLS 100 ... Machine (software distribution machine) 101 ... Software distribution system 102 ... Distribution control part 103 ... Distribution execution part 104 ... Machine operation part 105 ... Configuration information storage part 110 ... Machine (target machine (group)) , 111 ... active machine, 112 ... standby machine, 120 (121, 122) ... distribution agent, 130 ... load balancer, 150, 151, 152 ... software, 200 ... machine information, 300 ... software installation information.

Claims (5)

A software distribution system that distributes and updates software to a group of machines in a target computer system,
The machine group includes first and second machines corresponding to a scale-out configuration, and the attribute of the first machine is an active system that is activated regardless of whether the scale-out is executed or not. The attribute of the second machine is a standby system that is stopped when scale-out is not executed and is started when scale-out is executed.
The software distribution system includes:
A first processing unit for managing configuration information related to the machine group;
A second processing unit for distributing software to the group of machines;
A third processing unit for operating a start or stop state of the machine group,
The software distribution system includes:
The first processing unit manages, as configuration information, attributes related to the scale-out configuration, status, and information regarding the software configuration for the machine group including machines in a stopped state.
When distributing the software by the second processing unit, determine the attribute and state of each machine based on the configuration information,
When the second machine is in a stopped state, the second processing unit is started by the third processing unit, and the first processing unit is caused to transition from the stopped state to the starting state by the configuration information.
The second processing unit transmits the software data information to each machine including the first and second machines at the same time to update the software of each machine, and the first processing unit Reflect the update result in the configuration information,
Based on the configuration information, attributes and states of each machine are determined. When the second machine is in an activated state, the third processing unit stops the machine, and the first processing unit performs the configuration information. A software distribution system characterized in that a transition is made from a start state to a stop state. The software distribution system according to claim 1, wherein
The software distribution system includes:
A distribution machine comprising the first, second and third processing units;
A storage unit for storing the configuration information;
An agent unit provided in each machine of the target computer system and cooperating with the first, second, and third processing units,
The agent unit of each machine has a function of transmitting information related to the start and stop states of the machine to the distribution machine, a function of performing installation or updating on the machine when distributing the software from the distribution machine, And a function of transmitting information on the result of the distribution to the distribution machine. The software distribution system according to claim 1, wherein
As a startup state of the machine, it has a first state and a second state,
Controlling the first machine to be in the first state;
Controlling the second machine to stop when scale-out is not executed and to enter the first state when scale-out is ON;
A software distribution system that controls the first machine to be in the first state and the second machine to be in the second state during the distribution. A software distribution method in a software distribution system that performs processing for distributing software to a group of machines of a target computer system,
The machine group includes first and second machines corresponding to a scale-out configuration, and the attribute of the first machine is an active system that is activated regardless of whether the scale-out is executed or not. The attribute of the second machine is a standby system that is stopped when scale-out is not executed and is started when scale-out is executed.
During the process of executing distribution of the software to the group of machines,
Based on the management of configuration information including the state of the machine group and the software configuration,
A processing step of starting the stopped second machine;
A processing step of transmitting data information of the software to each machine including the first machine and the activated second machine to update the software of each machine;
And a processing step of stopping the activated second machine after the update. A program that causes a computer to function as a software distribution system that distributes and updates software to a group of machines in a target computer system,
The machine group includes first and second machines corresponding to a scale-out configuration, and the attribute of the first machine is an active system that is activated regardless of whether the scale-out is executed or not. The attribute of the second machine is a standby system that is stopped when scale-out is not executed and is started when scale-out is executed.
During the process of executing software distribution to the group of machines,
Based on the management of configuration information including the state of the machine group and the software configuration,
A processing step of starting the stopped second machine;
A processing step of transmitting data information of the software to each machine including the first machine and the activated second machine to update the software of each machine;
And a processing step of stopping the activated second machine after the update.

Images