JP6772857B2 - Judgment program, judgment method and judgment device - Google Patents

Description

The present invention relates to a determination program, a determination method, and a determination device.

In response to the request received from the user's terminal device, the information processing system extracts the data, processes the data, and sends the result of the processing as a response to the terminal device that sends the request.

FIG. 1 is an example of the above-mentioned information processing system, and is called, for example, a three-layer system. In the example of FIG. 1, the terminal device 1 transmits a request to the information processing system 10. In the information processing system 10, the Web server 11 receives the request and transmits the contents of the argument of the request to the Application (AP) server 12. When access to the data stored in the database is required, the AP server 12 generates a query statement according to the contents of the received argument and sends the generated query statement to the Database (DB) server 13. Then, the AP server 12 executes the application using the data extracted from the database by the DB server 13 as input data, and transmits the processing result of the application to the Web server 11. The Web server 11 transmits the processing result of the received application to the terminal device 1 as a response to the request. The above is an example of processing in a three-layer system.

A plurality of information processing systems may be constructed by using one or a plurality of information processing devices. For example, as shown in FIG. 1, three information processing systems 10, 20, and 30 are constructed in an environment such as a data center provided with one or a plurality of information processing devices. Similar to the information processing system 10, the information processing system 20 includes a Web server 21, an AP server 22, and a DB server 23, and executes processing according to a request transmitted by the terminal device 2. Further, the information processing system 30 also includes a Web server 31, an AP server 32, and a DB server 33, like the information processing system 10, and executes processing according to a request transmitted by the terminal device 3. The information processing systems 10, 20, and 30 may have different users and system types (collection search system, in-house personnel information management system, etc.).

Each server such as the Web server, AP server, and DB server shown in FIG. 1 is a virtual computer (virtual machine, virtual machine, VM). In other words, each server such as the Web server, AP server, and DB server shown in FIG. 1 is a virtual computer into which server software is installed. For example, when the server virtualization technology is applied, the number of VMs as a server and the number of information processing devices as a physical machine do not necessarily have to match.

Special Table 2013-543171

In addition to the unique software corresponding to each server type, software such as anti-virus software, which is required regardless of the server type, is introduced into each VM as a server. Therefore, when the VM operates, the process of the unique software (server software) corresponding to each server type, the process of other software, the process executed by the Operating System (OS), etc. operate in the VM. It will be.

By the way, apart from the server virtualization technology described above, there is a technology for operating a plurality of servers with one VM, and is called, for example, a container technology, a virtual container, etc. (hereinafter, referred to as a container technology).

In order to improve cost performance, it is required to suppress excess and deficiency of VM resources (calculation processing capacity, memory amount, etc.) when a plurality of servers are operated by one VM. In the following, operating a plurality of servers with one VM may be described as "aggregating" the servers. Therefore, for example, the resource usage when each of the plurality of servers targeted for server aggregation operates in individual VMs is added up, and the maximum number of servers is selected as long as the total usage amount does not exceed the resource upper limit of one VM. It is conceivable to consolidate into one VM.

However, in the above-mentioned server aggregation, the total usage amount may be overestimated than the actual usage amount. The above-mentioned total usage is estimated as the consumption of resources by processes other than the unique software process corresponding to the server type is also generated by the number of servers to be aggregated. However, when servers are actually aggregated, resource consumption by processes other than the unique software process is considered to depend on the number of VMs, not the number of servers. That is, it is considered that the consumption of resources by processes other than the unique software process is equivalent to one server regardless of the number of servers. In other words, the above-mentioned total usage is considered to be overestimated by about (n-1) units of processes other than the unique software process, where n is the number of servers to be aggregated. Be done.

In order to suppress the above-mentioned overestimation, when one server operates in one VM, the process operating in the VM is either a unique software process or a process other than the unique software process. It is required to distinguish whether there is.

One aspect is to make it possible to distinguish whether or not each of a plurality of processes running in a virtual machine is a process generated by the server software installed in the virtual machine.

In one aspect, a specific process among a plurality of processes running in a virtual machine in which server software is installed in a computer is a process generated by the server software and a process generated by software other than the server software. Which of the above is determined based on the number of connections established for the virtual machine and the number of child processes or threads spawned by the particular process, and the particular process is spawned by the server software. When it is determined that the process is to be executed, the server software generates the plurality of processes other than the specific process based on the information stored in the storage unit indicating the relationship between the plurality of processes. A process for determining whether the process is a process or a process generated by software other than the server software is executed.

On one aspect, it is possible to distinguish whether or not each of the plurality of processes running in the virtual machine is a process generated by the server software installed in the virtual machine.

The outline of the information processing systems 10, 20, and 30 is shown. An overview diagram of this embodiment is shown. A functional block diagram illustrating a software configuration of the centralized control device 300 is shown. The functional block diagram of the storage unit 220 is shown. An example of the information stored in the configuration information storage unit 223 is shown. An example of the information stored in the network information storage unit 222 is shown. An example of the information stored in the operation information storage unit 221 is shown. An example of the information stored in the directory information storage unit 224 is shown. An example of the information stored in the disk access information storage unit 225 is shown. An example of the information stored in the transmission / reception amount information storage unit 226 is shown. A flowchart showing the procedure of the process determination process in this embodiment is shown. A flowchart illustrating the details of the process of step S1003 is shown. An example of a VM communication setting file is shown. A flowchart illustrating the details of the process of step S1004 is shown. A flowchart illustrating the details of the process of step S1004 is shown. A flowchart illustrating the details of the process of step S1005 is shown. A flowchart illustrating the details of the process of step S1006 is shown. An example of the bottleneck determination information storage unit 227 is shown. An image is shown in which the aggregation process when the determination result of the server-specific process according to this embodiment is used and the above-mentioned excessive estimation (comparative example) are compared. An example of the hardware configuration of the centralized control device 300 is shown.

Examples of carrying out the present invention will be described with reference to the drawings.

[Overview]
FIG. 2 is an overview view of this embodiment. The system 100 of this embodiment includes an information processing device 200, an integrated control device 300, and a terminal device 400.

A plurality of VMs operate on the information processing device 200, and each VM operates as a server of any type. Although FIG. 2 shows a mode in which one information processing device is used, a plurality of information processing devices 200 may be used. Further, for example, the information processing device 200 of FIG. 2 may be a data center provided with a plurality of information processing devices.

The centralized control device 300 is communicably connected to the information processing device 200 via a wired or wireless network 500. The centralized control device 300 specifies the resource usage for each process of the VM operating on the information processing device 200, and determines whether or not the process is a unique software process corresponding to the server type. .. In the following description, the unique software process corresponding to the server type may be referred to as "server-specific process" for convenience. The server-specific process can also be said to be a server software process in a VM, a process generated by the server software, or a process directly related to the operation of the server.

The terminal device 400 is communicably connected to the information processing device 200 via the network 500. The terminal device 400 transmits a request to an information processing system (such as the above-mentioned three-layer system) configured by a VM operating on the information processing device 200, and receives a response to the request. In the following description, unless otherwise specified, when simply described as VM, it means a VM that operates on the information processing device 200 and has one server installed on its own device.

[Functional configuration of centralized control device 300]
FIG. 3A is a functional block diagram illustrating a software configuration of the centralized control device 300. The aggregation control device 300 includes an input unit 201, a display unit 202, a communication unit 203, a processing unit 210, and a storage unit 220.

The input unit 201 receives the input of information to the aggregation control device 300.

The display unit 202 outputs information such as whether or not server aggregation, which will be described later, is possible.

The communication unit 203 serves as an interface when the centralized control device 300 communicates with another device.

The processing unit 210 includes an operation monitoring unit 211, a directory information acquisition unit 212, a determination unit 213, and an aggregation control unit 214.

The operation monitoring unit 211 monitors the operation of the VM operating in the information processing device 200 and collects the operation information. Details of the operation information collected by the operation monitoring unit 211 will be described later.

The directory information acquisition unit 212 acquires information about the execution file of the process operating in the VM. The directory information acquisition unit 212 stores the acquired information in, for example, the directory information storage unit 224 described later.

The determination unit 213 determines whether or not the process is a server-specific process for the process operating in the VM. In this embodiment, in the following description, the process of determining whether or not the process is a server-specific process is described as "process determination process".

More specifically, in this embodiment, the determination unit 213 includes a first determination unit 213-1, a second determination unit 213-2, and a third determination unit 213-3. The first determination unit 213-1 executes the process determination process for the VM process on which the Web server operates. The second determination unit 213-2 executes the process determination process for the VM process in which the AP server operates. The third determination unit 213-3 executes the process determination process for the VM process in which the DB server operates. However, in this embodiment, it is assumed that the server is one of a Web server, an AP server, and a DB server. Therefore, the type of server to be determined and the specific configuration of the embodiment according to the type of server are not limited to this embodiment.

The aggregation control unit 214 determines whether or not the servers can be aggregated and the combination of servers to be aggregated based on the determination result of the determination unit 213. Further, the aggregation control unit 214 can transmit an instruction to execute the aggregation of the servers to the information processing apparatus 200 based on the determined combination of servers.

FIG. 3B is a functional block diagram of the storage unit 220 shown in FIG. 3A. As shown in FIG. 3B, the storage unit 220 includes an operation information storage unit 221, a network information storage unit 222, a configuration information storage unit 223, a directory information storage unit 224, a disk access information storage unit 225, and a transmission / reception amount information storage unit. 226, a bottleneck determination information storage unit 227 is provided. Details of each will be described later along with specific examples of stored information.

Before explaining the process determination process in this embodiment, the collection of information used for the process determination process will be described.

[Structure information storage unit 223]
FIG. 4 is a diagram showing an example of information stored in the configuration information storage unit 223. The configuration information storage unit 223 stores information such as OS type 2231, Internet Protocol (IP) address 2232, host name 2233, and server type 2234 shown in FIG. 4, for example. The information stored in the configuration information storage unit 223 is information about each VM operating on the information processing device 200. The information stored in the configuration information storage unit 223 in this embodiment may be held by the centralized control device 300 in the storage unit of its own device. Alternatively, it may be stored in a configuration information database (Configuration Management Database, CMDB) outside the own device, which can be accessed by the centralized control device 300.

The OS type 2231 is information indicating what the OS of each VM is. In FIG. 4, it is described as "OS-A 64 bit", "OS-B 32 bit", and "OS-B 64 bit". However, the data format shown in FIG. 4 is only an example, and the specific data format can be appropriately changed as long as the information can specify what the OS of each VM is.

The IP address 2232 is information indicating the IP address assigned to the VM. In FIG. 4, "IP_X1", "IP_X2", and "IP_X3" are described for the sake of explanation, but in reality, the data format is compliant with, for example, IPv4 (version 4) or IPv6 (version 6).

The host name 2233 is information indicating the host name (computer name) given to each VM. The host name 2233 may not be limited to the mode shown in FIG. 4 as long as each VM can be uniquely identified.

The server type 2234 is information indicating the type of the server operating in each VM. In this embodiment, the type of server can also be said to be the type of business system installed in each VM. In this embodiment, Web server, AP server, and DB server are listed as the types of servers. Note that FIG. 4 shows information on VMs for a total of three VMs, one for each of the Web server, the AP server, and the DB server.

[Network information storage unit 222]
FIG. 5 is a diagram showing an example of information stored in the network information storage unit 222. The network information storage unit 222 stores, for example, the local side socket information 2221, the remote side socket information 2222, the state information 2223, and the Process Identifier (PID) 2224 shown in FIG. The information stored in the network information storage unit 222 is information regarding the state of inter-process communication of the VM operating on the information processing device 200. For example, the centralized control device 300 can process information by using the netstat command or the like. It can be collected from device 200. Specifically, for example, the centralized control device 300 instructs the VM operating on the information processing device 200 to execute the netstat command, so that information about the instructed VM is acquired. Then, by transmitting the information acquired by the instructed VM to the aggregation control device 300, the aggregation control device 300 can acquire the information. In the following description, the above interprocess communication may be described as a connection.

The local socket information 2221 is information including a local host name and a port number in the connection. Here, the local side refers to the VM side from which the information has been acquired. The local socket information 2221 may be, for example, information including the IP address of the host instead of the host name. The local socket information 2221 shown in FIG. 5 is a data format in which the host name and the port number are separated by ": (colon)", but if the host name and the port number in the data can be distinguished, the figure is shown. A data format different from 5 may be used. Further, in FIG. 5, the port number is described as, for example, "PORT_ # 113" for the sake of explanation, but the actual port number is, for example, a 16-bit (0 to 65535) numerical value.

The remote side socket information 2222 is information including a remote side host name and a port number in the connection. Here, the remote side refers to the VM side of the communication partner of the VM that has acquired the information. The specific data format of the remote socket information 2222 and whether or not a modification can be adopted are the same as those of the local socket information 2221.

The state information 2223 is information indicating the state of interprocess communication. In FIG. 5, for example, a state of "LISTEN" and a state of "ESTABLISHED" are shown. "LISTEN" is waiting for a connection request, that is, the connection has not been established. "ESTABLISHED" indicates that the connection has been established.

The PID 2224 is identification information that identifies the process to which the local socket belongs, which is shown in the local socket information 2221. In FIG. 5, for example, it is described in the format of "P_X110", but specifically, the PID is a number (numerical value) that uniquely identifies the process, for example.

From the information shown in FIG. 5, the state of the connection between the processes can be specified at the timing when the information is acquired. For example, taking line 222- # 1 in FIG. 5 as an example, the host name and port number of the local socket are "HOST_X1: PORT_ # 110", and the host name and port number of the remote socket are "HOST_U1: PORT_ # U1". It is specified that the connection is not established (LISTEN) and the PID of the process corresponding to the local socket is "P_X110". Further, for example, in line 222- # 2 of FIG. 5, the host name and port number of the local socket is "HOST_X1: PORT_ # 111", and the host name and port number of the remote socket are "HOST_U2: PORT_". # U2 ”, the connection is established (ESTABLISHED), and it is specified that the PID of the process corresponding to the local socket is“ P_X111 ”.

[Operation information storage unit 221]
FIG. 6 is a diagram showing an example of information stored in the operation information storage unit 221. The operation information storage unit 221 is, for example, shown in FIG. 6, PID2211, Parent Process Agent (PPID) 2212, Central Processing Unit (CPU) usage rate 2213, real memory usage 2214, virtual memory usage 2215, number of threads 2216, The execution file information 2217 is stored. The information shown in FIG. 6 includes information about each process running in the VM and information indicating the amount of resources used by each process. The information shown in FIG. 6 can be collected from the information processing device 200 by the centralized control device 300 by using, for example, a ps command or the like. Specifically, for example, when the centralized control device 300 instructs the VM operating on the information processing device 200 to execute the ps command, the information about the instructed VM is acquired. Then, by transmitting the information acquired by the instructed VM to the aggregation control device 300, the aggregation control device 300 can acquire the information.

The PID 2211 is identification information that identifies the process, and may have the same aspect as the PID 2224 of FIG. 5, for example.

PPID2212 is information indicating identification information of the parent process of the process shown in PID2211. That is, the process shown in PID 2211 is a child process of the process shown in PPID 2212.

The CPU usage rate 2213 indicates the CPU usage rate for each process for the CPU resources allocated to the VM.

The actual memory usage amount 2214 indicates the memory capacity (actual memory capacity) used by each process. The virtual memory usage amount 2215 indicates the virtual memory capacity used by each process.

The number of threads 2216 is information indicating the number of threads generated by the process.

Executable file information 2217 is information that identifies an executable file of each process. In this embodiment, the executable file information 2217 may be, for example, an executable file name or an execution command name of each process. In the aspect of FIG. 6, the execution file name of each process is stored in the value of the execution file information 2217.

[Directory information storage unit 224]
FIG. 7 is a diagram showing an example of information stored in the directory information storage unit 224. The directory information storage unit 224 stores, for example, the executable file information 2241 and the path information 2242 shown in FIG. 7.

The executable file information 2241 is information that identifies the executable file, and may have the same mode as the executable file information 2217 of FIG. 6, for example.

The path information 2242 is information indicating the full path of the directory in which the executable file shown in the executable file information 2241 is stored.

The information shown in FIG. 7 can be collected from the information processing device 200 by the directory information acquisition unit 212 of the centralized control device 300 by using, for example, the ls (LS) command. Specifically, for example, when the centralized control device 300 instructs the VM operating on the information processing device 200 to execute the ls command, information about the instructed VM is acquired. Then, by transmitting the information acquired by the instructed VM to the aggregation control device 300, the aggregation control device 300 can acquire the information.

[Disk access information storage unit 225]
FIG. 8 is a diagram showing an example of information stored in the disk access information storage unit 225. The disk access information storage unit 225 stores information indicating the access status of each process to the storage area (storage area, disk area) allocated to the VM, specifically, the amount of access data. For example, as shown in FIG. 8, the disk access information storage unit 225 stores the PID 2251, the read data amount 2252, and the write data amount 2253.

The PID 2251 is identification information that identifies the process, and may have the same aspect as the PID 2224 of FIG. 5, for example.

The read data amount 2252 indicates the amount of data read from the storage area allocated to the VM by the process per unit time.

The write data amount 2253 indicates the write data amount per unit time for the storage area allocated to the VM by the process.

The information shown in FIG. 8 can be collected from the information processing device 200 by the centralized control device 300 by using, for example, an iotop command or the like. Specifically, for example, the centralized control device 300 instructs the VM operating on the information processing device 200 to execute the iotop command, so that information about the instructed VM is acquired. Then, by transmitting the information acquired by the instructed VM to the aggregation control device 300, the aggregation control device 300 can acquire the information.

[Transmission / reception amount information storage unit 226]
FIG. 9 is a diagram showing an example of information stored in the transmission / reception amount information storage unit 226. The transmission / reception amount information storage unit 226 stores information that specifies the data transmission / reception amount of the Web server for the terminal device 400. The transmission / reception amount information storage unit 226 stores, for example, the network interface 2261, the transmission data amount 2262, and the reception data amount 2263.

The network interface 2261 is identification information of the Network Interface Card (NIC) of the Web server.

The transmission data amount 2262 is information indicating the data transmission amount per unit time from the Web server to the terminal device 400.

The received data amount 2263 is information indicating the amount of data received per unit time received by the Web server from the terminal device 400.

The information shown in FIG. 9 can be collected by the centralized control device 300 from the information processing device 200 by using, for example, an ifconfig command or the like. Specifically, for example, the centralized control device 300 instructs the VM of the Web server operating on the information processing device 200 to execute the ifconfig command, so that information about the instructed VM is acquired. To. Then, by transmitting the information acquired by the instructed VM to the aggregation control device 300, the aggregation control device 300 can acquire the information.

The information stored in each of the operation information storage unit 221, the network information storage unit 222, the disk access information storage unit 225, and the transmission / reception amount information storage unit 226 is acquired, for example, by the aggregation control device 300 a plurality of times in a predetermined cycle. The information stored in the configuration information storage unit 223 and the directory information 224 may be, for example, once acquired at a certain timing by the aggregation control device 300. This is because the information stored in the configuration information storage unit 223 and the directory information 224 is information that does not change unless the system configuration is changed. However, the information stored in the configuration information storage unit 223 and the directory information 224 may be, for example, the information acquired at the latest timing among the information acquired at a plurality of different timings.

[Details of judgment processing]
The process determination process in this embodiment will be described in detail.

FIG. 10 is a flowchart showing the procedure of the process determination process in this embodiment.

First, the determination unit 213 selects one VM from the VMs whose information is stored in the configuration information storage unit 223 and targets it for the process determination process (step S1001).

Next, the determination unit 213 determines the type of server operating on the VM selected in step S1001 (step S1002). For example, the determination unit 213 may determine the server type by referring to the server type 2234 associated with the host name 2233 of the VM selected in step S1001 in the configuration information storage unit 223. When the server type is a Web server, the first determination unit 213-1 executes the first determination process (step S1003). When the server type is an AP server, the second determination unit 213-2 executes the second determination process (step S1004) and the third determination process (step S1005). When the server type is a DB server, the third determination unit 213-3 executes the fourth determination process (step S1006). Each of steps S1003 to S1006 is a process of determining whether or not the VM process is a server-specific process. Details of each of steps S1003 to S1006 will be described later.

After any of the processes of steps S1003 to S1006 is executed, the determination unit 213 determines whether all the VMs have been selected as targets in step S1001 (step S1007). Here, all VMs may be all VMs in which information is stored in the configuration information storage unit 223, or are designated from all VMs in which information is stored in the configuration information storage unit 223. , May be some VM. If it is determined that not all VMs have been selected (step S1007, NO), the process returns to step S1001. On the other hand, when it is determined that all the VMs have been selected (step S1007, YES), the series of processes shown in FIG. 10 ends.

Hereinafter, the processing executed in each of steps S1003 to S1006 will be described in detail. In the following description, details may be described with reference to FIGS. 5 and 6, but in the description, it is assumed that the information shown in FIGS. 5 and 6 is acquired at the same timing.

[Details of step S1003]
FIG. 11 is a flowchart illustrating the details of the process of step S1003 of FIG. In this embodiment, the process of FIG. 11 is a process executed by the first determination unit 213-1 for the VM process whose server type is determined to be a Web server.

First, the first determination unit 213-1 acquires the port number of the port used when the target VM (Web server) receives the request from the terminal device 400 (step S1101). The port number of the port used when receiving the request is, for example, the port assigned to the Transmission Control Protocol (TCP) communication whose service name is "http" or "https" in the communication setting file of the target VM. It is a number.

FIG. 12 is a diagram showing an example of a VM communication setting file. In the communication setting file 1200 of FIG. 12, the service name is shown in the area 1201, and the port number and the communication protocol are shown in the area 1202.

Returning to the description of FIG. 11, after step S1101, the first determination unit 213-1 specifies the number of established connections (step S1102). Here, the number of established connections can be specified, for example, based on the information stored in the network information storage unit 222. For example, the first determination unit 213-1 specifies a connection corresponding to the condition that the port number acquired in step S1101 is included in the local side socket information 2221 and the value of the state information 2223 is “ESTABLISHED”. Then, the first determination unit 213-1 specifies the total number of the corresponding connections as the number of established connections. In this embodiment, the connections shown in the data rows 222- # 2, 222- # 3, and 222- # 4 in FIG. 5 correspond to each other, and the number of established connections is 3.

After the processing of step S1102 is completed, the first determination unit 213-1 specifies the number of child processes for the process selected in step S1102 (step S1103). The number of child processes may be, for example, the number of processes having the process selected in step S1102 as the parent process in the operation information acquired at a certain timing. Here, the first determination unit 213-1 stores the identification information (PPID) of the parent process and the number of the specified child processes in association with each other. For example, assuming that the processes in the range of 221 # 1 shown in FIG. 6 are acquired as the processes of the Web server, PPI D = P_X110 having 3 child processes, PPI D = P_X120 having 1 child process, and 1 child process. PPID = P_X130 and PPID = P_X140 having 1 child process are specified.

The first determination unit 213-1 selects one parent process from the parent processes for which the number of child processes has been specified in step S1103 (step S1104).

Then, the first determination unit 213-1 compares the number of established connections identified in step S1103 with the number of child processes of the parent process selected in step S1104. Then, the first determination unit 213-1 determines whether or not the number of established connections and the number of child processes match (step S1105).

When it is determined that the number of established connections and the number of child processes match (step S1105, YES), the first determination unit 213-1 determines the parent process selected in step S1104 and the child process of the parent process. It is determined that the process is server-specific (step S1106). On the other hand, if it is not determined that the number of established connections and the number of child processes match (step S1105, NO), the first determination unit 213-1 determines the parent process selected in step S1104 and the child process of the parent process. Is not a server-specific process (step S1107).

For example, in the example of FIG. 6, the process of PPID = P_X110 has 3 child processes, which matches 3 which is the number of established connections, and thus is determined to be a server-specific process. On the other hand, the process of PPI D = P_X120, the process of PPI D = P_X130, and the process of PPI D = P_X140 are determined not to be server-specific processes because the number of child processes is 1 and does not match the number of established connections of 3. Will be done.

After step S1106 or S1107, the first determination unit 213-1 determines whether or not all the parent processes have been selected in step S1104 (step S1108). If all parent processes are not selected (steps S1108, NO), the process returns to step S1104, selects an unselected parent process, and repeats the processes of steps S1105 to S1108. On the other hand, when all the parent processes are selected (step S1108, YES), the first determination unit 213-1 executes the process of step S1109.

After the server-specific process is specified by the processes of steps S1101 to S1108, the first determination unit 213-1 executes the additional identification process of the server-specific process (step S1109).

In step S1109, the first determination unit 213-1 specifies the directory in which the process execution file is stored for the specified server-specific process. When a plurality of server-specific processes are specified by the processing of steps S1101 to S1108, the plurality of server-specific processes and child processes are the targets of the processing of step S1109. In order to specify the directory in which the executable file is stored, for example, the first determination unit 213-1 is the execution information storage unit 221 shown in the executable file information 2217 associated with the PID of the server-specific process. Identify the file. Then, the first determination unit 213-1 specifies the directory corresponding to the specified execution file with reference to the directory information storage unit 224. Specifically, in the directory information storage unit 224, the executable file is stored based on the full path of the executable file shown in the path information 2242 associated with the specified executable file (value of the executable file information 2241). The directory is identified.

Then, the first determination unit 213-1 determines the server-specific process based on the positional relationship of the storage position of the executable file in the directory. Specifically, the first determination unit 213-1 determines that the process corresponding to the specified directory and the execution file stored in the directory under the specified directory is also a server-specific process.

For example, in FIGS. 6 and 7, the executable file name of the process of PPI D = P_X110 determined to be a server-specific process in the above example is CMD_ # 1, and the full path of the directory of the executable file is ". / Opt / local / bin / CMD_ # 1 ". The executable file with the executable file name CMD_ # 4 is stored in the same directory as the executable file with the executable file name CMD_ # 1. The executable file having the executable file name CMD_ # 5 is stored in a directory under the directory of the executable file having the executable file name CMD_ # 1. The executable file having the executable file name CMD_ # 6 is stored in a directory that is neither the same as the directory of the executable file whose executable file name is CMD_ # 1 nor under it. Therefore, the first determination unit 213-1 refers to the information in FIG. 6 to perform a process having PPI D = P_X120 having an executable file name of CMD_ # 4 and a process having PCID = P_X130 having an executable file name CMD_ # 5. Determined to be a server-specific process. On the other hand, the first determination unit 213-1 determines that the process with PPID = P_X140 whose execution file name is CMD_ # 6 is not a server-specific process. The process of PPID = P_X120 and the process of PPID = P_X130 are processes that are not determined to be server-specific processes in the examples described in the description of steps S1105 to S1107.

The server-specific process identified by the process of step S1109 is a process generated by an execution file stored in the same directory as the execution file of the server-specific process or a directory under the directory of the execution file of the server-specific process. .. By the process of step S1109, the server process can be extracted from the processes that could not be specified in steps S1101 to S1108.

After the process of step S1109 is completed, the series of processes shown in FIG. 11 is completed. The above description has described a series of processes based on the operating status of the process and the network status acquired at a certain timing. However, if the process operation status and network status are acquired at a plurality of timings (for example, periodically at a predetermined cycle), FIG. 11 shows the process operation status and network status at each timing. The series of processes shown may be executed a plurality of times. Then, the first determination unit 213-1 identifies the process determined to be a server-specific process as a server-specific process more than a predetermined number of times (for example, once) in a plurality of executions, and determines that the process is a server-specific process. It may be determined that the process that has been executed less than the predetermined number of times is not a server-specific process.

When the process (acceptance process) of the Web server receives the request from the terminal device 400, the connection with the terminal device 400 for the process is established (“ESTABLISHED”), and a child process is generated. After the child process is created, the processing corresponding to the request is taken over by the child process. Then, the connection with the terminal device 400 for the child process is established, and the connection with the terminal device 400 for the parent process is in the request waiting state (“LISTEN”). When the processing corresponding to the request is completed, the child process disappears, but the establishment of the connection is maintained until the completion, that is, while the child process exists. Therefore, the number of established connections to the Web server and the number of child processes of the Web server process (accepting process) are equal. This characteristic can be used to identify a server-specific process for a Web server. In the specific method of this embodiment, the server-specific process is specified by comparing the number of connections with the number of child processes. Therefore, for example, the determination accuracy is improved in that it can be determined that the process whose resource usage accidentally increases at the timing of receiving the request is not a server-specific process.

Further, in the directory of the executable file of the specified server-specific process, the executable file of the process that is the server-specific process is stored, although it cannot be specified by comparing the number of connections and the number of child processes. Therefore, the determination accuracy can be further improved by specifying the process whose execution file is stored in the same directory as the execution file of the specified server-specific process or the directory under it as the server-specific process. The process extracted here is, for example, a process for outputting a log, a process for collecting unnecessary resources, and the like, and is called, for example, a background process.

[Details of step S1004]
13 and 14 are flowcharts illustrating the details of the process of step S1004 of FIG. In this embodiment, the processes of FIGS. 13 and 14 are the processes executed by the second determination unit 213-2 for the VM process whose server type is determined to be the AP server.

First, the second determination unit 213-2 identifies the connection established with the Web server for the target VM (AP server) (step S1301). For example, the second determination unit 213-2 can identify the established connection for each of the target VM and the VM of the Web server based on the information stored in the network information storage unit 222.

More specifically, for example, the second determination unit 213-2 establishes a connection for the target VM (AP server) in which the information stored in the network information storage unit 222 satisfies all of the following conditions. Judge that there is.
Condition 1: The value (host name and port number) of the local socket information 2221 matches the value of the remote socket information 2222 of the VM of the Web server.
Condition 2: The value of the remote side socket information 2222 matches the value of the local side socket information 2221 of the VM of the Web server.
Condition 3: The state of the connection shown in the state information 2223 is "ESTABLISHED".

In addition to the above conditions 1 to 3, the following may be added as conditions.
Condition 4: The connection status shown in the VM status information 2223 of the Web server is also "ESTABLISHED".
Condition 5: The connection protocol is TCP.

A specific example will be described with reference to FIG. For example, suppose that the connections shown in the data lines 222- # 8 to 222- # 11 in FIG. 5 are specified as the AP server connections. In this case, the connection of the Web server shown in the data line 222- # 5-222- # 7, which satisfies the above conditions 1 to 3 for the connection of the data lines 222-# 9, 222- # 10, and 222- # 11. Exists. Therefore, the connections in data rows 222- # 9, 222- # 10, 222- # 11 are identified as established connections for the AP server. In the example of FIG. 5, in addition to the conditions 1 to 3, conditions 4 and 5 are also satisfied.

After executing step S1301, the second determination unit 213-2 identifies the corresponding process for each of the established connections identified in step S1301 (step S1302). Specifically, the second determination unit 213-2 identifies the process to which the local socket belongs for each of the established connections. In FIG. 5, the processes of PID = P_X211, P_X221, and P_X231 shown in data rows 222- # 9, 222- # 10, and 222- # 11 are applicable.

Further, the second determination unit 213-2 specifies the parent process of each process specified in step S1302 (step S1303). The parent process specified here targets only the processes that have established a connection with the Web server or are waiting for a connection from the Web server. The process waiting for a connection from the Web server includes, for example, the host name of the Web server in the value of the remote socket information 2222 shown in FIG. 5, and the connection state shown in the state information 2223 is "LISTEN". Can be specified as a condition. In this embodiment, from FIGS. 5 and 6, the process with PPI D = P_X210 is specified as the parent process.

The second determination unit 213-2 selects one process from the parent processes identified in step S1303 (step S1304).

The second determination unit 213-2 specifies the number of established connections for each process selected in step S1304 (step S1305).

For example, assuming that the process selected in step S1304 is the parent process of PPID = P_X210, the number of established connections is 1 from FIGS. 5 and 6.

The second determination unit 213-2 specifies the number of child processes for each of the processes selected in step S1304 (step S1306).

For example, if the process selected in step S1304 is the parent process of PPID = P_X210, the number of child processes is 3 from FIG.
The second determination unit 213-2 compares the number of established connections with the number of child processes for each of the processes selected in step S1304. Then, the second determination unit 213-2 determines whether or not the number of child processes is equal to or greater than the number of established connections (step S1307).

When it is determined that the number of child processes is equal to or greater than the number of established connections (step S1307, YES), the second determination unit 213-2 determines that the process selected in step S1304 is a server-specific process. (Step S1308). On the other hand, when it is determined that the number of child processes is not equal to or greater than the number of established connections (step S1307, NO), the second determination unit 213-2 does not consider the process selected in step S1302 to be a server-specific process. Determine (step S1309).

For example, if the process selected in step S1304 is the parent process of PPID = P_X210, the number of child processes is 3 and is 1 or more, which is the number of established connections, as in the above example. , The second determination unit 213-2 determines that the parent process (and its child process) of PPID = P_X210 is a server-specific process.

After completing the process of step S1308 or step S1309, the second determination unit 213-2 determines whether or not all the parent processes have been selected in step S1304 (step S1310). When it is determined that all the parent processes have not been selected (step S1310, NO), the process returns to step S1304, and the second determination unit 213-2 selects one unselected process and repeats the process.

When it is determined that all the parent processes have been selected (step S1310, YES), the second determination unit 213-2 executes the additional identification process of the server-specific process (step S1311). The process of step S1311 may be the same as the additional specific process described as step S1109 of FIG. 11, for example.

After the end of step S1311, the series of processes shown in FIGS. 13 and 14 ends. The above description has described a series of processes based on the operating status of the process and the network status acquired at a certain timing. However, if the process operation status and network status are acquired at a plurality of timings (for example, periodically at a predetermined cycle), FIG. 13 and FIG. 13 based on the process operation status and network status at each timing. The series of processes shown in FIG. 14 may be executed a plurality of times. Then, the second determination unit 213-2 identifies the process determined to be a server-specific process as a server-specific process more than a predetermined number of times (for example, once) in a plurality of executions, and determines that the process is a server-specific process. It may be determined that the process that has been executed less than the predetermined number of times is not a server-specific process.

[Details of step S1005]
FIG. 15 is a flowchart illustrating the details of the process of step S1005 of FIG.
In the present embodiment, the process of FIG. 15 is a process executed by the second determination unit 213-2 for the VM process whose server type is determined to be the AP server.

First, the second determination unit 213-2 identifies the connection established with the Web server for the target VM (AP server) (step S1501). In addition, the second determination unit 213-2 identifies the corresponding process for each of the established connections identified in step S1501 (step S1502). The processing of steps S1501 and S1502 may be, for example, the same processing as those of steps S1301 and S1302 described above, or the processing results of steps S1301 and S1302 may be diverted. In the example of FIG. 6, the processes of PID = P_X211 and PID = P_X221 and PID = P_X231 are specified.

Next, the second determination unit 213-2 specifies the number of established connections for each process specified in step S1502 (step S1503). In the example of FIG. 6, the number of established connections is 1 for each of the processes PID = P_X211 and PID = P_X221 and PID = P_X231.

Then, the second determination unit 213-2 selects one process from the processes specified in step S1502 (step S1504), and executes the processes of steps S1505 to S1509 described later for the selected process.

In step S1505, the second determination unit 213-2 specifies the number of threads of the process for the process selected in step S1504. For the number of threads of the process, for example, among the information of the number of threads 2216 stored in the operation information storage unit 221 shown in FIG. 6, the value (numerical value) of the data row corresponding to the PID of the process selected in step S1504 is referred to. It is possible to identify by.

After step S1505, the second determination unit 213-2 compares the number of established connections with the number of threads of the process for the process selected in step S1504. Then, the second determination unit 213-2 determines whether or not the number of threads is larger than the number of established connections (step S1506).

If it is determined that the number of threads is larger than the number of established connections (step S1506, YES), the second determination unit 213-2 determines that the process selected in step S1504 is a server-specific process. (Step S1507). On the other hand, when it is determined that the number of child processes is not equal to or greater than the number of established connections (step S1506, NO), the second determination unit 213-2 does not consider the process selected in step S1504 to be a server-specific process. Determine (step S1508).

For example, in the example of FIG. 6, the number of threads of the process of PID = P_X221 is 2, which is larger than the number of established connections of 1. Therefore, the second determination unit 213-2 determines that the process with PID = P_X221 is a server-specific process. On the other hand, the processes of PID = P_X211 and PID = P_X231 each have 1 thread, which is not larger than 1 which is the number of established connections. Therefore, the second determination unit 213-2 determines that the processes with PID = P_X211 and PID = P_X231 are not server-specific processes.

After completing the process of step S1507 or step S1508, the second determination unit 213-2 determines whether or not all the processes have been selected in step S1504 (step S1509). If it is determined that all the processes have not been selected (step S1509, NO), the process returns to step S1504, and the second determination unit 213-2 selects one unselected process and repeats the process.

When it is determined that all the processes have been selected (step S1509, YES), the second determination unit 213-2 executes the additional identification process of the server-specific process (step S1510). The process of step S1510 may be the same as the additional specific process described as step S1109 of FIG. 11, for example.

After the end of step S1510, the series of processes shown in FIG. 15 ends. The above description has described a series of processes based on the operating status of the process and the network status acquired at a certain timing. However, if the process operating status and network status are acquired at a plurality of timings (for example, periodically at a predetermined cycle), FIG. 15 shows the process operating status and network status at each timing. The series of processes shown may be executed a plurality of times. Then, the second determination unit 213-2 identifies the process determined to be a server-specific process as a server-specific process more than a predetermined number of times (for example, once) in a plurality of executions, and determines that the process is a server-specific process. It may be determined that the process that has been executed less than the predetermined number of times is not a server-specific process.

The Web server that receives the request from the terminal device establishes a connection with the application process on the AP server and sends the request according to the request content of the request. The process of the AP server that received the request spawns a child process or thread. The number of child processes or threads created at this time depends on the application, but one or more child processes or threads are created. Therefore, the number of child processes of the process of the AP server that generates the child process in response to the request has a characteristic that it is equal to or larger than the number of connections between the Web server and the AP server. In addition, the number of threads in the process of the AP server that creates threads in response to the request has a characteristic that it is larger than the connection between the Web server and the AP server (because the number of threads is acquired as 1 in the state before thread creation). ).

In the specific method of this embodiment, for the AP server, the server-specific process is specified by comparing the number of connections with the number of child processes or the number of connections with the number of created threads. Therefore, for example, the determination accuracy is improved in that it can be determined that the process whose resource usage accidentally increases at the timing of receiving the request is not a server-specific process.

Further, similarly to the above-mentioned Web server, a process in which the executable file is stored in the same directory as the executable file of the specified server-specific process or a directory under it is also specified as a server-specific process. As a result, the background process of the AP server can be specified, so that the determination accuracy can be further improved.

As in this embodiment, for the AP server, the server-specific process is specified by comparing the number of connections with the number of child processes or the number of connections with the number of generated threads. Therefore, a process determined to be a server-specific process by at least one of the comparison between the number of connections and the number of child processes or the comparison between the number of connections and the number of created threads may be regarded as a server-specific process. Also, it is possible to identify the server-specific process of the AP server by comparing the number of connections with the number of child processes, or the number of connections with the number of created threads, but both. The determination accuracy can be improved by performing the above processing.

[Details of step S1006]
FIG. 16 is a flowchart illustrating the details of the process of step S1006 of FIG. In the present embodiment, the process of FIG. 16 is a process executed by the third determination unit 213-3 with respect to the VM process whose server type is determined to be the DB server.

In the series of processes shown in FIG. 16, first, the third determination unit 213-3 identifies the period during which the response performance of the DB server is deteriorated (step S1601). In the following description, the period in which the response performance of the DB server is deteriorated may be described as the response deterioration period, and the period in which the response performance of the DB server is deteriorated may be described as the bottleneck.

The outline of the method for specifying the response deterioration period of the DB server in this embodiment will be described. In the description, it is assumed that the centralized control device 300 acquires VM operation information, network information, and the like at a plurality of different timings, for example, at a fixed cycle within a predetermined period. The response deterioration period is a period during which the response performance of the DB server is deteriorated in the predetermined period in which the VM operation information, network information, etc. are acquired.

If the response of the information processing system is not deteriorated, the ratio (transmission / reception ratio) of the data transmission amount and the data reception amount of the Web server in the information processing system has a characteristic of being constant (however, when it becomes constant). The ratio varies from system to system). In the following description, the transmission / reception ratio = data transmission amount / data reception amount will be described.

On the other hand, when the response performance of the information processing system is deteriorated, the transmission / reception ratio of the Web server is smaller than that in the normal state. That is, the value of the transmission / reception ratio is smaller than that in the normal case. By utilizing this characteristic, it is possible to specify the period in which the response of the information processing system is deteriorated based on the data transmission amount and the data reception amount of the Web server collected during the predetermined period.

Further, if the response performance of the information processing system is not deteriorated, the ratio of the number of connections between each server is also constant. That is, when the number of connections between the terminal device and the Web server is A, the number of connections between the Web server and the AP server is B, and the number of connections between the AP server and the DB server is C, A: B: There is a characteristic that C is constant. On the other hand, when the response performance of the information processing system deteriorates due to the DB server, as described above, the transmission / reception ratio of the Web server is smaller than that at the normal time, and the ratio of the data transmission amount is smaller and A: B. : C has the characteristic that it is the same as the normal time, or only C is larger than the normal time. Therefore, it is possible to specify the response deterioration period of the DB server based on the transmission / reception ratio of the Web server and the ratio of the number of connections between each server.

Here, the above-mentioned normal transmission / reception ratio and the number of connections may be, for example, an average value of the transmission / reception ratio and the number of connections during a predetermined period in which the operation information and the network information are collected. Alternatively, it may be the average value of the transmission / reception ratio and the number of connections in the period obtained by excluding the period to be determined whether or not the response is deteriorated from the predetermined period.

After step S1601, the third determination unit 213-3 selects one of the response deterioration periods specified in step S1601 (step S1602).

The third determination unit 213-3 specifies the type of bottleneck for the response deterioration period selected in step S1602 (step S1603). In this embodiment, the type of bottleneck can be said to be a factor in the deterioration of the response performance of the DB server. The third determination unit 213-3 identifies the type of bottleneck based on, for example, the resource usage status during the response deterioration period and the information stored in the bottleneck determination information storage unit 227 shown in FIG.

FIG. 17 is a diagram showing an example of the bottleneck determination information storage unit 227. For example, the bottleneck determination information storage unit 227 stores the bottleneck type 2271, the overall resource tendency 2272, and the process resource tendency 2273.

The bottleneck type 2271 is information indicating the type of bottleneck. Regarding the bottleneck type 2271 shown in FIG. 17, “CPU” indicates that the bottleneck is caused by the CPU. In addition, "storage (disk)" indicates that it is a bottleneck due to storage. Note that the storage refers to a storage means or a storage area allocated separately from the memory for the VM, and for example, a part of the storage area of the storage device such as the Hard Disk Drive (HDD) included in the information processing device 200 is allocated. It is a memory.

“Exclusive control” indicates that it is a bottleneck due to processing wait (lock) due to exclusive control in accessing the database. Further, "network" indicates that the VM of the DB server is a bottleneck due to the delay or congestion of the network used when communicating with other VMs. In this embodiment, the bottleneck type, that is, the cause of deterioration of the response performance of the DB server can be classified into the four shown in FIG.

The overall resource tendency 2272 is information indicating the tendency of the resource usage status of the entire VM during the response deterioration period for each type of bottleneck.

The process resource tendency 2273 is information indicating the tendency of the resource usage status of the server-specific process during the response deterioration period for each type of bottleneck.

Explaining the process of step S1603 with reference to FIG. 17, the third determination unit 213-3 specifies the resource usage status of the entire VM during the response deterioration period. Specifically, the third determination unit 213-3 determines, for example, whether the usage amount of each resource of the entire VM during the response deterioration period corresponds to "low", "no fluctuation", or "high". To do. The specific determination of "low", "no fluctuation", and "high" is, for example, that the usage of each resource of the entire VM during the response deterioration period is a predetermined threshold value with respect to the average usage of each resource of the entire VM. If it is larger than that, it may be judged as "high". Further, when the usage amount of each resource of the entire VM during the response deterioration period is smaller than the average usage amount of each resource of the entire VM by a predetermined threshold value or more, it may be determined as “low”. Then, when neither "high" nor "low" is determined, it may be determined that there is no change.

After identifying the usage status of each resource of the entire VM during the response deterioration period, the third determination unit 213-3 determines that the usage status of each resource of the identified VM as a whole matches any of the overall resource trends 2272 shown in FIG. Determine if you want to. Then, the third determination unit 213-3 specifies the bottleneck type 2271 associated with the overall resource tendency determined to match as the bottleneck type corresponding to the target response deterioration period.

After executing step S1603, the third determination unit 213-3 identifies the server-specific process from the processes in the VM of the DB server based on the type of bottleneck specified in step S1603 (step S1604). An example of the specific processing of step S1604 will be described again with reference to FIG.

First, the third determination unit 213-3 specifies the resource usage status for the target VM (DB server) process. Specifically, the third determination unit 213-3 determines, for example, whether the usage amount of each resource for each process during the response deterioration period corresponds to "low", "no fluctuation", or "high". To do. The specific determination of "low", "no fluctuation", and "high" may be, for example, the same as the resource usage status of the entire VM described above.

The third determination unit 213-3 determines whether or not the usage amount of each resource for each process matches the process resource tendency 2273 associated with the bottleneck type identified in step S1603. The third determination unit 213-3 determines that the process in which the usage amount of each resource matches the process resource tendency 2273 associated with the bottleneck type identified in step S1603 is a server-specific process. On the other hand, the third determination unit 213-3 determines that the process in which the usage amount of each resource does not match the process resource tendency 2273 associated with the bottleneck type identified in step S1603 is not a server-specific process.

In step S1603 and step S1604, the usage amount of each resource for each process is determined from, for example, the information stored in the operation information storage unit 221, the disk access information storage unit 225, and the transmission / reception amount information storage unit 226 for the process of the DB server. Can be calculated. Further, the usage amount of each resource in the entire VM can be the total usage amount of the DB server process for each resource. Further, the average usage amount may be, for example, the average value of the usage amount of each resource in a predetermined period, or the average value of the usage amount of each resource in a predetermined period excluding the data acquired at the timing of the determination target. Further, the threshold value used as the determination criterion can be, for example, a ratio (percentage) to the average usage amount or a difference in numerical values. The threshold value may be appropriately changed for each type of resource and for each determination of "high" and "low".

After executing the process of step S1604, the third determination unit 213-3 determines whether or not all the response deterioration periods have been selected in step S1602 (step S1605). When it is determined that all the response deterioration periods have not been selected (step S1605, NO), the process returns to step S1602, and the third determination unit 213-3 selects one unselected response deterioration period and steps S1603. The process of ~ S1605 is repeated.

When it is determined that all the response deterioration periods have been selected (step S1605, YES), the third determination unit 213-3 executes the additional identification process of the server-specific process (step S1606). The process of step S1606 may be the same as the additional specific process described as step S1109 of FIG. 11, for example. After the end of step S1606, the series of processes shown in FIG. 16 ends.

Most of the access to the DB server is a data read request from the AP server to the database. Therefore, the DB server is equipped with a large amount of memory (allocated to the VM) as compared with, for example, a Web server or an AP server, and uses the memory as a cache. Even for read requests for the same data, the load on resources (specifically, in particular, access to the disk) varies from time to time, depending on whether the requested data is stored in the cache.

In addition, the load of resources (specifically, access to a disk in particular) varies depending on the installed memory capacity. For example, if the memory capacity is such as an in-memory database that can store all the data in the database in the cache, disk access hardly occurs.

Further, since the size of the data required to be read fluctuates greatly for each request, the CPU usage rate and the memory usage fluctuate greatly for each request.

In addition, the DB server may have a function that detects the number of records in the database table or the number of deleted records and operates, and what kind of data read request was received in the past due to the influence of the function. Also, the CPU usage rate and the memory usage amount change.

From the above items, the behavior of the DB server is inconsistent when the process behavior is simply monitored, and it is not possible to specify the unique server process from the relationship between the request (number of connections) and the process behavior. It can be said that it is difficult.

However, the DB server has a characteristic that the life period of the process for processing the data read request becomes long when the response of the DB server deteriorates. Although the factors of response deterioration are not uniform, as described in this example, it is possible to classify them into patterns that are far fewer than the behavior of the above-mentioned process. Moreover, if the cause of the response deterioration can be identified, the behavior of the server-specific process according to the factor can be estimated.

Therefore, as described above, in the present embodiment, the third determination unit 213-3 first identifies the response deterioration period and the cause of the response deterioration for the DB server. Then, by specifying the process that shows the behavior that matches the behavior of the server-specific process estimated based on the factor of the response deterioration, the third determination unit 213-3 can specify the server-specific process, and the determination accuracy is improved. improves.

Further, similarly to the above-mentioned Web server, a process in which the executable file is stored in the same directory as the executable file of the specified server-specific process or a directory under it is also specified as a server-specific process. As a result, the background process of the DB server can be specified, so that the determination accuracy can be further improved.

[Aggregate processing]
The aggregation control unit 214 can execute the aggregation processing of the server by using the determination result of the server-specific process for each VM by the determination unit 213.

FIG. 18 is a diagram showing an image comparing the aggregation process when the determination result of the server-specific process according to the present embodiment is used and the above-mentioned excessive estimation (comparative example). FIG. 18 shows an example of determining whether or not the servers can be aggregated for two VMs in which one server operates.

Graph 1801 shows the transition of the resource usage amount in the first VM to be aggregated. Graph 1802 shows the transition of the resource usage in the VM to be aggregated in the second unit. In graphs 1801 and 1802, the resource usage by the server-specific process (denoted as "unique" in FIG. 18) and the resource usage by processes other than the server-specific process (denoted as "non-unique" in FIG. 18) are totaled. The resource usage of the VM as a whole (denoted as "total" in FIG. 18) is shown.

In this embodiment, a case where the servers operating on the second VM to be aggregated are aggregated in the VM to be aggregated for the first unit is used as an example. However, the server aggregation process may be performed, for example, by consolidating the servers on the VM of the second aggregation target, and is different from the VM of the first aggregation target and the VM of the second aggregation target. It may be aggregated in the VM of. It is assumed that the threshold value in the graph is a threshold value that serves as a criterion for determining whether or not the operating performance of the server is affected when aggregated. The threshold value that serves as a criterion for determining whether or not the operating performance of the server is affected is not uniquely limited, and is appropriately determined according to the mode of the information processing system, the amount of resources allocated to the VM when the servers are aggregated, and the like. It may be possible to decide and change.

Graph 1803 shows the usage amount estimated as the resource usage amount when the servers are aggregated in each of the comparative example and the present embodiment.

In the comparative example, the estimated usage amount is simply the sum of the resource usage amount of each of the resources of the two VMs as a whole without distinguishing between the server-specific process and the process other than the server-specific process. In the comparative example, since there is a period in which the usage amount estimated as the resource usage amount when the servers are aggregated exceeds the threshold value, it is determined that the server aggregation is impossible.

On the other hand, in this embodiment, the resource usage by the server-specific process is added up for the two VMs. However, for processes other than the server-specific process, the resource usage for one VM is added to the resource usage for the server-specific process for two VMs, and the estimated usage is calculated. According to this embodiment, since the usage amount estimated as the resource usage amount when the servers are aggregated does not exceed the threshold value, it is judged that the servers can be aggregated.

That is, even if it is determined in the comparative example that the server cannot be aggregated due to the excessive estimation of the resource usage, in this embodiment, the excessive estimation of the resource usage is suppressed, so that the server Can be judged to be possible.

According to the method described above, the aggregation control unit 214 may estimate the resource usage amount when server aggregation is performed for a plurality of VMs, and determine whether or not server aggregation is possible.

For example, the aggregation control unit 214 may display the identification information of a plurality of VMs that can be aggregated by the server on the display unit 202. Further, the aggregation control unit 214 may determine whether or not the servers can be aggregated for a plurality of VMs for which the identification information has been input by the input unit, and display the determination result on the display unit 202. good. Further, the centralized control unit 214 may display information indicating whether or not the process is a server-specific process for each VM process on the display unit 202, or in the manner shown in FIG. 18, each VM and each VM may be displayed. The resource usage of the VM after server aggregation may be displayed on the display unit 202. Alternatively, the aggregation control unit 214 may execute server aggregation by transmitting an instruction to the information processing device 200 to execute server aggregation for a plurality of VMs capable of server aggregation.

When consolidating servers, a plurality of VMs having the same server type may be targeted for aggregation. If the server types are the same, even if the servers are aggregated, the behavior of processes other than the server-specific processes will not change, or it will be less than when the server types are different. Is.

[Hardware configuration]
FIG. 19 is an example of the hardware configuration of the centralized control device 300 in this embodiment. Although FIG. 19 shows an example of the hardware configuration of the centralized control device 300, the same configuration can be adopted for the information processing device 200 and the terminal device 400 as described later.

The centralized control device 300 is, for example, an information processing device including a CPU 1902, a memory 1903, a storage device 1904, a NIC 1905, a medium reading device 1906, an input device 1907, and a display device 1908, each of which is connected to each other by a bus 1901. ..

The CPU 1902 controls various operations in the centralized control device 300. The memory 1903 and the storage device 1904 store a program that executes various processes described in this embodiment and various data used for various processes. The storage device 1904 is, for example, a storage medium such as an HDD or a Solid State Drive (SSD).

The CPU 1902 may read the program stored in the memory 1903 or the storage device 1904, perform processing, and execute control to realize the processing unit 210 shown in FIG. 3A and each functional unit included in the processing unit 210. The CPU 1902 can be replaced with a hardware circuit such as a Micro Processing Unit (MPU) or an Application Specific Integrated Circuit (ASIC). Further, each of the memory 1903 and the storage device 1904 can function as the storage unit 220 shown in FIGS. 3A and 3B.

NIC1905 is hardware used for transmitting and receiving data via a wired or wireless network. The NIC1905 can function as the communication unit 203 under the control of the CPU 1902.

The medium reading device 1906 is a device for reading data from a recording medium, for example, a disk drive for reading data stored in a disk medium such as a CD-ROM or a DVD-ROM, or a disk drive for reading data stored in a memory card. It is a card slot to read. A part or all of the data stored in the storage unit 220 described above may be stored in a recording medium that can be read by using the medium reading device 1906.

The input device 1907 is a device that receives input and designation from the user. Examples of the input device 1907 include, for example, a keyboard, a mouse, and a touch pad. The display device 1908 displays various information under the control of the CPU 1902. The display device 1908 is, for example, a liquid crystal display.

Since the information processing device 200 and the terminal device 400 can adopt the same hardware configuration as the centralized control device 300 shown in FIG. 19, the description thereof will be omitted. However, the specific hardware (model, performance, etc.) of the CPU, memory, storage device, NIC, medium reader, input device, and display device is the information processing device 200, the centralized control device 300, and the terminal device 400, respectively. It doesn't matter if it is different. For example, the information processing device 200 may be equipped with a CPU having a higher computing power than the centralized control device 300 and the terminal device 400, and a memory having a larger capacity.

The information processing device 200 may be, for example, one or a plurality of physical server devices, or the information processing device 200 in the present embodiment may be replaced with a data center including a plurality of computers. Further, when the information processing device 200 is replaced with a data center including a plurality of computers, the centralized control device 300 may be realized by using a part of the computer resources of the data center.

Further, the terminal device 400 may be, for example, a Personal Computer (PC), a smartphone, a Personal Digital Assistant (PDA), or the like. When the terminal device 400 is a smartphone, a PDA, or a tablet PC, a touch panel having the functions of the input device 1907 and the display device 1908 may be used.

[Summary]
As described above, according to the present embodiment, it is possible to determine whether the process operating in the VM is a unique software process (server-specific process) or a unique software process. is there. By determining the server-specific process, excessive estimation of resource usage when determining whether server aggregation is possible is suppressed. Therefore, it is possible to consolidate more servers within a range that does not adversely affect the operating performance of the servers after consolidating the servers. If more servers can be aggregated, the number of VMs operating in the information processing apparatus 200 can be reduced, and the cost for operating the VMs can be reduced. Further, in the operation management of the information processing system, it is possible to reduce the cost of patch application processing to the OS, which increases or decreases according to the number of VMs.

Further, in this embodiment, various commands used for collecting information used for determining whether or not the process is a server-specific process are, for example, in an OS such as UNIX (registered trademark) or Windows NT (registered trademark). Is a standard available command. Therefore, it is not necessary to introduce an additional tool for collecting information on the server. This is an advantage, for example, when the contract with the user who provides the system does not allow the introduction of additional tools.

Further, when the information processing apparatus 200 is a data center, in this embodiment, the information used for determining whether or not the process is a server-specific process may be collected for stable operation of the system. is there. Therefore, there is an advantage that new information collection processing is not required to realize this embodiment, and it is easy to continuously collect information in the operation of the system.

Further, in this embodiment, the background process of the server can be specified as the server-specific process based on the storage state of the process execution file in the directory, and the determination accuracy can be improved. This corresponds to the additional identification process of the server-specific process described in this embodiment. According to the inventor's examination, for example, in the case of a DB server, about 30% of the resource usage by the server-specific process is the usage by the background process.

In this embodiment, the types of servers are classified into Web server, AP server, and DB server. However, for example, for a server in which the Web server and the AP server are integrated, the process determination process for the Web server, the process determination process for the AP server, and the determination unit 213 may be executed. Then, the determination unit 213 may specify the process determined to be the server-specific process as at least one of the process determination process for the Web server and the process determination process for the AP server as the server-specific process.

200 Information processing device 300 Aggregate control device 400 Terminal device 500 Network 201 Input unit 202 Display unit 203 Communication unit 210 Processing unit 220 Storage unit 1901 Bus 1902 CPU
1903 Memory 1904 Storage Device 1905 NIC
1906 Media reader 1907 Input device 1908 Display device

Claims (17)

On the computer
Whether a specific process among a plurality of processes running in a virtual machine in which the server software is installed is a process generated by the server software or a process generated by software other than the server software. Determined based on the number of connections established for the virtual machine and the number of child processes or threads spawned by the particular process.
When it is determined that the specific process is a process generated by the server software, the plurality of processes other than the specific process are stored in the storage unit based on the information indicating the relationship between the plurality of processes. Whether the process of the above is a process generated by the server software or a process generated by software other than the server software.
A judgment program characterized by executing a process. On the computer
When the number of the connections and the number of the child processes are equal, it is determined that the specific process is a process generated by the server software.
The determination program according to claim 1, wherein the process is executed. On the computer
Determine the type of server software
When it is determined that the type of the server software is a web server, it is determined whether the specific process is a process generated by the server software or a process generated by software other than the server software. Execute the process to
The determination program according to claim 2, wherein the process is executed. On the computer
When the number of child processes is equal to or greater than the number of connections, it is determined that the specific process is a process generated by the server software.
The determination program according to claim 1, wherein the process is executed. On the computer
If the number of threads is greater than the number of connections, it is determined that the particular process is a process spawned by the server software.
The determination program according to claim 1, wherein the process is executed. On the computer
Determine the type of server software
When it is determined that the type of the server software is an application server, it is determined whether the specific process is a process generated by the server software or a process generated by software other than the server software. Execute the process to
The determination program according to claim 4 or 5, wherein the process is executed. The determination program according to any one of claims 1 to 6, wherein the connection is a TCP connection. The information indicating the relationship between the plurality of processes is information indicating the storage location of the executable file corresponding to the plurality of processes in the directory.
On the computer
Of the plurality of processes other than the specific process, the process executable file is stored in the same directory as the specific process executable file, or in a directory under the directory in which the specific process executable file is stored. Is determined to be a process generated by the server software.
The determination program according to any one of claims 1 to 7, wherein the process is executed. On the computer
The resource usage in the first virtual machine running the specific type of server software is combined with the resource usage by the process generated by the server software in one or more other virtual machines running the specific type of server software. When the total value of is equal to or less than a predetermined threshold value, it is determined that the server software of the first virtual machine and the server software of the other machine can be operated by one virtual machine.
The determination program according to any one of claims 1 to 8, wherein the determination program is characterized by the above. On the computer
Acquires information indicating the resource usage status by multiple processes running in the virtual machine on which the server software is installed.
It depends on each of the plurality of states that can be shown in the resource usage status by the plurality of processes, each of the operating states of the virtual machine corresponding to each of the plurality of states, and the process generated by the server software in each of the operating states. Based on the judgment information associated with the information indicating the resource usage status, the operating status of the virtual machine corresponding to the acquired information indicating the usage status is specified.
For each of the plurality of processes, the server software associated with the operating state specified in the determination information as to whether it is a process generated by the server software or a process generated by software other than the server software. Judgment based on whether the resource usage by the process generated by
A judgment program characterized by executing a process. On the computer
Determine the type of server software
When it is determined that the type of the server software is a database server, the specifying process and the determining process are performed.
The determination program according to claim 10, wherein the process is executed. The virtual machine is a virtual machine on which a database server included in an information processing system operates.
On the computer
The ratio of the amount of data transmitted to the amount of data received by the virtual machine in which the web server included in the information processing system operates during the specific period included in the predetermined period is larger than the average value of the ratio in the predetermined period. If it becomes smaller, it is determined that the specific period is a period in which the response performance of the information processing system has deteriorated.
When it is determined that the specific period is a period in which the response performance of the information processing system is deteriorated, the said is based on the number of connections between a plurality of virtual machines included in the information processing system in the specific period. It is determined whether or not the specific period is the period during which the response performance of the virtual machine on which the database server operates has deteriorated.
When it is determined that the specific period is the period during which the response performance of the virtual machine on which the database server operates has deteriorated, either a process generated by the server software or a process generated by software other than the server software. Performs the process of determining whether or not
The determination program according to claim 10, wherein the process is executed. On the computer
If there is a process determined to be a process generated by the server software, the process is among the plurality of processes that is stored in the storage unit and is specified based on the information indicating the relationship between the plurality of processes. A process having a predetermined relationship with a process determined to be a process generated by the server software is determined to be a process generated by the server software.
The determination program according to any one of claims 10 to 12, wherein the process is executed. The computer
Whether a specific process among a plurality of processes running in a virtual machine in which the server software is installed is a process generated by the server software or a process generated by software other than the server software. Determined based on the number of connections established for the virtual machine and the number of child processes or threads spawned by the particular process.
When it is determined that the specific process is a process generated by the server software, the plurality of processes other than the specific process are stored in the storage unit based on the information indicating the relationship between the plurality of processes. Whether the process of the above is a process generated by the server software or a process generated by software other than the server software.
Judgment method characterized by that. Whether a specific process among a plurality of processes running in a virtual machine in which the server software is installed is a process generated by the server software or a process generated by software other than the server software. A first determination unit that determines based on the number of connections established for the virtual machine and the number of child processes or threads spawned by the particular process.
When it is determined that the specific process is a process generated by the server software, the plurality of processes other than the specific process are stored in the storage unit based on the information indicating the relationship between the plurality of processes. A second determination unit that determines whether the process is a process generated by the server software or a process generated by software other than the server software.
A determination device comprising. The computer
Acquires information indicating the resource usage status by multiple processes running in the virtual machine on which the server software is installed.
It depends on each of the plurality of states that can be shown in the resource usage status by the plurality of processes, each of the operating states of the virtual machine corresponding to each of the plurality of states, and the process generated by the server software in each of the operating states. Based on the judgment information associated with the information indicating the resource usage status, the operating status of the virtual machine corresponding to the acquired information indicating the usage status is specified.
For each of the plurality of processes, the server software associated with the operating state specified in the determination information as to whether it is a process generated by the server software or a process generated by software other than the server software. Judgment based on whether the resource usage by the process generated by
Judgment method characterized by that. The computer
Acquires information indicating the resource usage status by multiple processes running in the virtual machine on which the server software is installed.
It depends on each of the plurality of states that can be shown in the resource usage status by the plurality of processes, each of the operating states of the virtual machine corresponding to each of the plurality of states, and the process generated by the server software in each of the operating states. Based on the judgment information associated with the information indicating the resource usage status, the specific unit that identifies the operating state of the virtual machine corresponding to the acquired information indicating the usage status, and the specific unit.
For each of the plurality of processes, the server software associated with the operating state specified in the determination information as to whether it is a process generated by the server software or a process generated by software other than the server software. Judgment unit that judges based on whether or not it matches the resource usage status by the process generated by
A determination device comprising.

Images