JP7537134B2 - CONTAINER MANAGEMENT DEVICE, CONTAINER MANAGEMENT METHOD, AND PROGRAM - Google Patents

Description

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

Patent Document 1 discloses a container that manages some of the associated resources and provides them as virtual resources. Patent Document 1 also discloses that a device management unit acquires utilization rates and actual performance from each managed device and stores them in a resource operation information database. Patent Document 1 also discloses that a virtual resource configuration provided to a tenant or application management server is managed for each container and stored in a container management table. Patent Document 1 also discloses that a correction coefficient is set in the table to commonly compare and evaluate the resource requirements of multiple heterogeneous applications, and the actual measured values of each performance index are corrected with the correction coefficient and the correction results are recorded.

International Publication No. 2015/049789

In Patent Document 1, in order to optimize resources, performance indicators are collected, and correction is performed using a specific magnification for each type of application, and the score calculation results are shown as the performance characteristics of the resource.
However, in Patent Document 1, it is not possible to eliminate the influence of resource allocation and obtain continuity of performance information.

Therefore, the object of the present invention is to provide a container management device, a container management method, and a program that solve the above-mentioned problems.

Some aspects of the present invention are made to solve the above-mentioned problems. The first aspect of the present invention is a container management device that includes a performance information correction means that corrects performance information related to one or more containers based on resource allocation information indicating resources allocated to a host computer or container platform on which the one or more containers operate.

A second aspect of the present invention is a container management method that corrects performance information related to one or more containers based on resource allocation information indicating resources allocated to a host computer or container platform on which the one or more containers operate.

A third aspect of the present invention is a program that causes a computer of a container management device to function as a performance information correction means that corrects performance information related to one or more containers based on resource allocation information that indicates resources allocated to a host computer or container platform on which the one or more containers operate.

According to some aspects of the present invention, it is possible to eliminate the influence of resource allocation and obtain continuity of performance information.

1 is a schematic configuration diagram of a container management system according to a first embodiment of the present invention. 5 is a flowchart showing a process of a monitoring manager according to the first embodiment of the present invention. 1 is a conceptual diagram showing the relationship between a host computer, a container base, and a container according to a first embodiment of the present invention. FIG. 13 is another conceptual diagram showing the relationship between the host computer, the container base, and the container according to the first embodiment of the present invention. 4 is a graph showing an example of resource allocation information collected by a resource allocation information collecting unit according to the first embodiment of the present invention. 4 is a graph showing an example of performance information collected by a performance information collecting unit according to the first embodiment of the present invention. 11 is a graph showing an example of corrected performance information generated by a performance information correcting unit according to the first embodiment of the present invention. FIG. 11 is a schematic configuration diagram of a container management system according to a second embodiment of the present invention. 10 is a flowchart showing a process of a monitoring manager according to a second embodiment of the present invention. 11 is a graph showing an example of resource allocation information collected by a resource allocation information collecting unit according to the second embodiment of the present invention. 13 is a graph showing an example of performance information collected by a performance information collecting unit according to the second embodiment of the present invention. 13 is a graph showing an example of corrected performance information generated by a performance information correcting unit according to the second embodiment of the present invention. FIG. 13 is a block diagram showing the configuration of a monitoring manager having a minimum configuration. 13 is a flowchart showing the processing of a monitoring manager having a minimum configuration.

[First embodiment]
First, a first embodiment of the present invention will be described.

1 is a schematic configuration diagram of a container management system 100a according to a first embodiment of the present invention. The container management system 100a includes a host computer 10a and a monitoring manager 20a (also referred to as a container management device).
The host computer 10a and the monitoring manager 20a are connected to each other via a wired or wireless network. Although Fig. 1 shows a case where there is one host computer 10a, there may be a plurality of host computers.

The host computer 10a includes a container board 11, a performance information collector 13, a performance information storage unit 14, a resource allocation information collector 15a, and a resource allocation information storage unit 16a.
The container base 11 controls the container 12 processed by the host computer 10a. The container 12 is a set of libraries and applications required for setting logical partitions on the host computer 10a and running applications, and is configured to be used like an individual server.
1 shows only one container 12, there may be a plurality of containers 12. The container board 11 can control a plurality of containers.

The performance information collector 13 collects performance information related to the container 12. Here, the container 12 collects information on the load of a CPU (Central Processing Unit) (also called a central processing unit) that processes the container 12, as the performance information.
The performance information accumulation unit 14 is a storage device such as a HDD (Hard Disk Drive), and stores various information related to the container 12, such as the performance information collected by the performance information collection unit 13.

The resource allocation information collection unit 15a collects resource allocation information, which is information on resources allocated to the container 12, from the container board 11. Here, the resource allocation information collection unit 15a collects, as the resource allocation information, information on the number of CPUs used for processing the container 12.
The resource allocation information storage unit 16a is a storage device such as an HDD, and stores the resource allocation information collected by the resource allocation information collection unit 15a.

On the other hand, the monitoring manager 20 a includes a performance information acquisition unit 21 , a performance information correction unit 22 a (also referred to as a performance information correction means), a performance information display unit 23 (also referred to as a performance information display means), and an administrator dialogue unit 24 .
The performance information acquiring unit 21 reads out the performance information stored in the performance information storage unit 14 of the host computer 10a, and outputs it to the performance information correcting unit 22a.

The performance information correction unit 22a reads out the resource allocation information stored in the resource allocation information accumulation unit 16a of the host computer 10a based on a correction instruction input from the administrator dialogue unit 24. The performance information correction unit 22a corrects the performance information read out from the performance information accumulation unit 14 using the resource allocation information read out from the resource allocation information accumulation unit 16a, and outputs the corrected performance information to the performance information display unit 23.

The performance information display unit 23 is a display device such as a display, and displays the corrected performance information output from the performance information correcting unit 22a as a graph.
The administrator dialogue unit 24 is an input device such as a keyboard, and based on the operation of the container management system 100a, acquires a correction instruction regarding whether or not to correct the performance information in the performance information correction unit 22a, and outputs the correction instruction to the performance information correction unit 22a.

FIG. 2 is a flowchart showing the processing of the monitoring manager 20a according to the first embodiment of the present invention.
First, the performance information acquiring unit 21 acquires performance information related to the container 12 by reading the performance information stored in the performance information storage unit 14 (step S11).

Next, the administrator dialogue unit 24 judges whether or not to correct the performance information based on whether or not a correction instruction instructing the performance information correction unit 22a to correct the performance information has been received (step S12). The correction instruction from the administrator includes information on whether or not to correct the performance information in the performance information correction unit 22a, and, if correction is to be performed, information on the correction period (for example, time t0 to tn (n is an integer)).
If it is determined that the performance information is not to be corrected (No in step S12), the process of step S15 described below is performed.

On the other hand, if it is determined that the performance information should be corrected (Yes in step S12), the performance information correction unit 22a reads the resource allocation information for the correction period from the resource allocation information accumulation unit 16a, and obtains information regarding the number of host computers 10a used to process the container 12 as resource allocation information of the container base 11 (step S13).

Next, the performance information correction unit 22a corrects the performance information acquired by the performance information acquisition unit 21 using the resource allocation information read from the resource allocation information accumulation unit 16a (step S14). For example, the performance information correction unit 22a increases or decreases the value of the performance information during the period in which the performance information is corrected so that the performance information at the end of the period before the performance information correction is started matches the performance information at the beginning of the period in which the performance information is corrected.

Next, the performance information display unit 23 displays the performance information corrected by the performance information correction unit 22a in a graph, thereby presenting the corrected performance information to the administrator of the container management system 100a (step S15).

3A and 3B are conceptual diagrams showing the relationship between a host computer 10a, a container board 11, and a container 12 according to the first embodiment of the present invention.
3A and 3B, a case will be described in which the host computer 10a (FIG. 1) is made up of a host computer 10-1 and a host computer 10-2.
When the container 12 is moved between the host computers 10-1, 10-2 with different performance, or when the CPU allocation available to the container 12 changes, the CPU resources available to the container 12 change.

FIG. 3A shows a case where one container 12 is processed by a host computer 10-1 that operates using four CPUs under the control of a container board 11.
FIG. 3B shows a case where one container 12 is processed by a host computer 10-2 that operates using two CPUs under the control of the container board 11.
The container board 11 controls whether one container 12 is processed by the host computer 10-1 or the host computer 10-2 by switching between the state of FIG. 3A and the state of FIG. 3B.

4A is a graph showing an example of resource allocation information collected by the resource allocation information collector 15a according to the first embodiment of the present invention. Note that the data of the graph shown in FIG. 4A is acquired by the performance information corrector 22a in step S13 of the flowchart in FIG.
In FIG. 4A, the horizontal axis indicates time, and the vertical axis indicates the number of CPUs used to process the container 12 indicating the resource allocation information.

Specifically, Fig. 4A shows that during the period from time t0 to t1, the container 12 is processed by the host computer 10-1, which operates using four CPUs. Also, Fig. 4A shows that during the period from time t1 to t2, the container 12 is processed by the host computer 10-2, which operates using two CPUs. Also, Fig. 4A shows that during the period from time t2 onwards, the container 12 is processed by the host computer 10-1, which operates using four CPUs.
In other words, in Figure 4A, the number of CPUs processing container 12 is reduced from four to two at time t1, and therefore, in container 12, the resources used to process container 12 are reduced by a certain amount during the period from time t1 to t2.

4B is a graph showing an example of performance information collected by the performance information collecting unit 13 according to the first embodiment of the present invention. Note that the data of the graph shown in FIG. 4B is acquired by the performance information acquiring unit 21 in step S11 of the flowchart in FIG.
In this case, the performance information is the CPU load when processing the container 12.
In FIG. 4B, the horizontal axis indicates time, and the vertical axis indicates the load per CPU when processing the container 12.

As explained in FIG. 4A, during the period from time t0 to t1, the container 12 is processed by the host computer 10-1, which uses four CPUs, but during the period from time t1 to t2, the container 12 is processed by the host computer 10-2, which uses two CPUs. FIG. 4B shows a case in which the number of available CPUs decreases during the period from time t1 to t2, causing the load per CPU to increase and exceed the monitoring threshold V1.

4C is a graph showing an example of corrected performance information generated by the performance information correcting unit 22a according to the first embodiment of the present invention. Note that the data of the graph shown in FIG. 4C is displayed by the performance information display unit 23 in step S15 of the flowchart in FIG. 2.
In FIG. 4C, the horizontal axis indicates time, and the vertical axis indicates the corrected load per CPU when processing the container 12.

Compared to FIG. 4B, in FIG. 4C, the performance information correction unit 22a reduces the value of the performance information during the period in which the performance information is corrected by a predetermined percentage along the vertical axis of FIG. 4C so that the performance information at the end of the period (time t0 to t1) before the start of the performance information correction matches the performance information at the beginning of the period in which the performance information is corrected (time t1 to t2). In FIG. 4C, the load per CPU does not exceed the monitoring threshold V1 even during the period from time t1 to t2.

As shown in Figures 3A and 3B, when the number of CPUs used to process the container 12 is changed, the detected CPU load increases (see times t1 to t2 in Figure 4B) although there is no change in the processing performed on the container 12. Since this change in CPU load is based on the CPU allocation of the host computer, there is no need to present this increase in CPU load during the period t1 to t2 as a problem to the administrator of the container management system 100a.

Conventional container management devices are unable to grasp performance information based on changes in the number of CPUs used to process such containers 12. Therefore, even if there is no change in the processing of containers 12 in the container management system 100a as a whole, as shown in Figure 4B, the CPU load may exceed the monitoring threshold V1, and the administrator may mistakenly believe that a fault has occurred in the container management system 100a.

In the first embodiment of the present invention, the performance information regarding the container 12 is presented to the administrator in the form of a graph as shown in FIG. 4C, rather than the graph as shown in FIG. 4B, which prevents the administrator from making the above-mentioned mistaken understanding, and allows the container management system 100a as a whole to accurately determine whether or not a failure has occurred.

[Second embodiment]
Next, a second embodiment of the present invention will be described.
Portions of the second embodiment that are similar to those of the first embodiment are given the same reference numerals and their description will be omitted.
In the second embodiment, a case will be described in which a graph of the CPU load of a container is displayed when the container is replicated due to a scale-out of the container environment.

Figure 5 is a schematic diagram of a container management system 100b according to a second embodiment of the present invention. The container management system 100b includes a host computer 10b and a monitoring manager 20b (also referred to as a container management device).

The host computer 10a according to the first embodiment includes a resource allocation information collection unit 15a and a resource allocation information accumulation unit 16a (see FIG. 1), whereas the host computer 10b according to the second embodiment includes a resource allocation information collection unit 15b and a resource allocation information accumulation unit 16b.

The resource allocation information collection unit 15b collects, from the container base 11, resource allocation information that is information on resources allocated to the containers 12. Here, the resource allocation information collection unit 15b collects, as the resource allocation information, information on the number of containers 12 being processed by the container base 11.
The resource allocation information storage unit 16b is a storage device such as an HDD, and stores the resource allocation information collected by the resource allocation information collection unit 15b.

The monitoring manager 20a according to the first embodiment includes a performance information correcting section 22a (see FIG. 1), but the monitoring manager 20b according to the second embodiment includes a performance information correcting section 22b (also referred to as performance information correcting means).
The performance information correction unit 22b reads out the resource allocation information stored in the resource allocation information accumulation unit 16b of the host computer 10b based on a correction instruction input from the administrator dialogue unit 24. The performance information correction unit 22b corrects the performance information read out from the performance information accumulation unit 14 using the resource allocation information read out from the resource allocation information accumulation unit 16b, and outputs the corrected performance information to the performance information display unit 23.

FIG. 6 is a flowchart showing the processing of the monitoring manager 20b according to the second embodiment of the present invention.
First, the performance information acquisition unit 21 acquires performance information related to the container 12 by reading the performance information stored in the performance information accumulation unit 14 (step S21).

Next, the administrator dialogue unit 24 judges whether or not to correct the performance information based on whether or not a correction instruction instructing the performance information correction unit 22b to correct the performance information has been received (step S22). The correction instruction from the administrator includes information on whether or not to correct the performance information in the performance information correction unit 22b, and, if correction is to be performed, information on the correction period (for example, time t0 to tn (n is an integer)).
If it is determined that the performance information is not to be corrected (No in step S22), the process of step S25 described below is performed.

On the other hand, if it is determined that the performance information should be corrected (Yes in step S22), the performance information correction unit 22b reads the resource allocation information for the correction period from the resource allocation information accumulation unit 16b, and obtains information regarding the number of containers 12 being processed by the container base 11 as the resource allocation information of the container base 11 (step S23).

Next, the performance information correcting unit 22b corrects the performance information acquired by the performance information acquiring unit 21 using the resource allocation information read from the resource allocation information accumulating unit 16b (step S24). For example, the performance information correcting unit 22b increases or decreases the value of the performance information during the period in which the performance information is corrected, so that the performance information at the end of the period before the performance information correction is started matches the performance information at the start of the period in which the performance information is corrected.
Next, the performance information display unit 23 displays the performance information corrected by the performance information correcting unit 22b in a graph, thereby presenting the corrected performance information to the administrator of the container management system 100b (step S25).

7A is a graph showing an example of resource allocation information collected by the resource allocation information collector 15b according to the second embodiment of the present invention. Note that the data of the graph shown in FIG. 7A is acquired by the performance information corrector 22b in step S23 of the flowchart in FIG. 6.
In FIG. 7A, the horizontal axis indicates time, and the vertical axis indicates the number of containers 12 being processed by the container board 11 indicating resource allocation information.

Specifically, FIG. 7A shows that during the period from time t0 to t3, the container board 11 is processing one container 12. Also, FIG. 7A shows that during the period from time t3 to t4, the container board 11 is processing two containers 12. Also, FIG. 7A shows that during the period from time t4 onwards, the container board 11 is processing one container 12.

7B is a graph showing an example of performance information collected by the performance information collecting unit 13 according to the second embodiment of the present invention. Note that the data of the graph shown in FIG. 7B is acquired by the performance information acquiring unit 21 in step S21 of the flowchart in FIG.
In this case, the performance information is the CPU load when processing the container 12.
In FIG. 7B, the horizontal axis indicates time, and the vertical axis indicates the CPU load when processing the container 12.

As explained in FIG. 7A, in the period from time t0 to t3, the container board 11 processes one container 12, but in the period from time t3 to t4, the container board 11 processes two containers 12. FIG. 7B shows a case in which the number of containers 12 being processed increases in the period from time t3 to t4, and therefore the CPU load per container decreases.

7C is a graph showing an example of corrected performance information generated by the performance information correcting unit 22b according to the second embodiment of the present invention. The data of the graph shown in FIG. 7C is displayed by the performance information display unit 23 in step S25 of the flowchart in FIG.
In FIG. 7C, the horizontal axis indicates time, and the vertical axis indicates the corrected load on the CPU when processing the container 12.

Compared to FIG. 7B, in FIG. 7C, the performance information correction unit 22b increases the value of the performance information during the period in which the performance information is corrected by a predetermined percentage along the vertical axis of FIG. 7C so that the performance information at the end of the period (time t0 to t3) before the start of the performance information correction matches the performance information at the beginning of the period in which the performance information is corrected (time t3 to t4).

As shown in Figures 7A and 7B, when the number of containers 12 processed by the container base 11 is changed, the number of CPUs processing the containers 12 does not change, but the detected CPU load decreases (see times t3 to t4 in Figure 7B). Since this change in CPU load is based on a change in the number of containers being processed, it may cause problems if the CPU load value for this period t3 to t4 is presented to the administrator of the container management system 100b as being no problem.

Conventional container management devices are unable to grasp performance information based on changes in the number of containers 12 being processed, so if there is no change in the load on the CPU processing one container, the administrator may mistakenly believe that no problem is occurring with the container management system 100b.

In the second embodiment of the present invention, the administrator is presented with a graph such as that shown in FIG. 7C, rather than a graph such as that shown in FIG. 7B, as performance information regarding the container 12. This makes it possible to prevent the administrator from making a mistaken understanding as described above, and allows the container management system 100b as a whole to accurately determine whether or not a fault has occurred.

In the above-described first and second embodiments, the case where the load information measured when the container 12 is processed by the CPU is used as the performance information is described, but this is not limited to this, and other performance information may be used.

In addition, in the first and second embodiments described above, the number of CPUs or the number of containers 12 are used as resource allocation information, but this is not limited to this, and CPU clock information may be used as resource allocation information.
In addition, in the above-described first and second embodiments, a case has been described in which one piece of resource allocation information (the number of CPUs or the number of containers) is used, but a plurality of pieces of resource allocation information may be used simultaneously.

Next, we will explain the monitoring manager 20c (also called the container management device) that has the minimum configuration.

Figure 8 is a block diagram showing the configuration of a monitoring manager 20c having a minimum configuration. The monitoring manager 20c includes a performance information correction unit 22c (also called a performance information correction means).

FIG. 9 is a flowchart showing the processing of the monitoring manager 20c having the minimum configuration.
The performance information correction unit 22c provided in the monitoring manager 20c corrects performance information regarding one or more containers based on resource allocation information indicating resources allocated to a host computer on which the one or more containers are running, or to a container base (step S31).

In addition, a program for realizing the functions of each part in Figures 1, 5, and 8 may be recorded on a computer-readable recording medium, and the program recorded on this recording medium may be read into a computer system and executed to process each part.

Note that the term "computer system" here includes hardware such as the OS (Operating System) and peripheral devices. Also, "computer-readable recording media" refers to portable media such as flexible disks, optical magnetic disks, ROMs (Read Only Memory), and CDs (Compact Disk)-ROMs, as well as storage devices such as HDDs built into computer systems. Furthermore, "computer-readable recording media" also includes devices that hold a program for a certain period of time, such as volatile memory (RAM (Random Access Memory)) inside a computer system that acts as a server or client when a program is transmitted over a network such as the Internet or over a communication line such as a telephone line.

Some aspects of the present invention can be applied to container management devices, container management methods, and programs that require the continuity of performance information to be obtained by eliminating the effects of resource allocation.

10a, 10b, 10-1, 10-2... host computers,
11...Container base,
12...Container 12,
13...Performance information collection unit,
14...performance information storage section,
15a, 15b...resource allocation information collection unit,
16a, 16b...resource allocation information storage unit,
20a, 20b, 20c... monitoring managers,
21... Performance information acquisition unit,
22a, 22b, 22c... performance information correction section,
23... Performance information display section,
24...Administrator dialogue section,
100a, 100b...Container management system

Claims (10)

a performance information correcting means for correcting performance information relating to the one or more containers based on resource allocation information indicating resources allocated to a host computer or a container platform on which the one or more containers operate ,
The performance information correction means
The performance information is corrected by increasing or decreasing a value of the performance information during a period in which the performance information is corrected so that first performance information at the end of the period before the start of the correction of the performance information and second performance information at the start of the period in which the performance information is corrected coincide with each other.
Container management device. 2. The container management apparatus according to claim 1, wherein the performance information correction means uses the number of central processing units of the host computer as the resource allocation information. 3. The container management device according to claim 2, wherein said performance information correction means corrects said performance information when the number of central processing units of said host computer changes, so that said performance information before and after the change becomes continuous. The container management device according to claim 2 , wherein the performance information correcting means uses, as the performance information, a load value of a central processing unit that processes the one or more containers. The container management device according to claim 1 , wherein the performance information correcting means uses the number of the one or more containers as the resource allocation information. The container management device according to claim 5 , wherein the performance information correcting means corrects the performance information when the number of the one or more containers changes so that the performance information before and after the change becomes continuous. The container management device according to claim 2 , wherein the performance information correcting means uses, as the performance information, a load value of a central processing unit that processes the one or more containers. The container management device according to claim 1 , further comprising a performance information display unit that displays the performance information corrected by the performance information correction unit as a graph. A container management method using a computer of a container management device, comprising:
correcting performance information related to the one or more containers based on resource allocation information indicating resources allocated to a host computer or a container platform on which the one or more containers operate;
In the correction of the performance information,
The performance information is corrected by increasing or decreasing a value of the performance information during a period in which the performance information is corrected so that first performance information at the end of the period before the start of the correction of the performance information and second performance information at the start of the period in which the performance information is corrected coincide with each other.
Container management methods. The computer of the container management device,
functioning as a performance information correcting means for correcting performance information relating to the one or more containers based on resource allocation information indicating resources allocated to a host computer or a container platform on which the one or more containers operate ;
The performance information correction means
The performance information is corrected by increasing or decreasing a value of the performance information during a period in which the performance information is corrected so that first performance information at the end of the period before the start of the correction of the performance information and second performance information at the start of the period in which the performance information is corrected coincide with each other.
program.

Images