JP2021027398A - Container daemon, information processing device, container-type virtualization system, packet distribution method, and program - Google Patents

Abstract

To provide a container virtualization environment that enables QoS control of a network without modifying a container.SOLUTION: A container daemon includes a packet distribution portion that outputs a packet to OS on which a container operates or a packet high-speed processing layer on the basis of setting information of a distribution destination of a packet delivered from the container.SELECTED DRAWING: Figure 1

Description

The present invention relates to a container daemon, an information processing device, a container-type virtualization system, a packet distribution method, and a program.

A technique called container-type virtualization that provides virtualization of an OS (Operating System) on one physical host is known. In container-type virtualization, a plurality of containers can be operated on one physical host. For example, a video conferencing service can be provided by a single physical host that operates a container for an application that makes a voice call, a container for an application that distributes video, and a container for an application that distributes document materials. In order to provide a comfortable video conferencing service, it is necessary to apply Quality of Service (QoS) requirements (eg, delay time, jitter, bandwidth) according to the characteristics of the delivered content to the communication traffic of each container. preferable. That is, it is necessary to be able to apply different QoS requirements for each container.

Tools such as DPDK (Data Plane Development Kit) are provided as an example of technologies related to QoS improvement applicable to a container-type virtualization environment. For example, when DPDK is introduced into a container, the communication traffic of the container is accelerated. Furthermore, by introducing the DPDK Quality of Service (DPDK) framework into a container, it becomes possible not only to speed up communication traffic but also to adjust the priority of packet processing between containers. For example, in the case of the above-mentioned video conferencing service, it is possible to set the priority of the communication traffic of the voice call container higher than that of other containers (video, materials) and prevent delay of voice during the conference.

In addition, Patent Document 1 discloses a technique for automatically generating a container with respect to a container virtualization technique. For example, Patent Document 1 discloses an example of generating a container having characteristics according to an SLA (Service Level Agreement).
Further, Patent Document 2 discloses a technique in which a CPU (Central Processing Unit) or an accelerator performs processing of a packet related to communication from a container based on a lookup table setting. Patent Document 2 discloses that the capacity of resources allocated to the lookup table is changed according to the bandwidth of the packet flow. For example, if the capacity of resources allocated to the lookup table is reduced, resources can be allocated to communication processing accordingly.
Further, Patent Document 3 discloses a transmission management system that transmits content data by using a relay device according to a service level.

Special Table 2019-503535 Gazette JP-A-2019-333551 JP-A-2017-38251

However, it is not realistic to introduce a packet high-speed processing function such as DPDK to all containers. Moreover, due to hardware restrictions and the like, it is not always possible to introduce the packet high-speed processing function into all the devices in the container environment. In a container virtualization environment, there is a demand for a method for realizing high-speed network and QoS control without modifying the container. Patent Documents 1 and 3 do not disclose a method for solving such a problem. Further, the technique of Patent Document 2 is premised on the existence of an accelerator, and requires a load in the process of allocating the lookup table capacity.

Therefore, an object of the present invention is to provide a container daemon, an information processing device, a container-type virtualization system, a packet distribution method, and a program that solve the above-mentioned problems.

According to one aspect of the present invention, the container daemon outputs the packet to the OS on which the container operates or the packet high-speed processing layer based on the setting information of the distribution destination of the packet delivered from the container. It has a part.

Further, according to another aspect of the present invention, the information processing apparatus includes the above-mentioned container daemon, an OS, and a packet high-speed processing layer introduced into the OS.

Further, according to another aspect of the present invention, the information processing apparatus includes the above-mentioned container daemon and an OS having a function of performing priority control on the processing of the packet.

Further, according to another aspect of the present invention, the container-type virtualization system includes one or a plurality of the above-mentioned information processing devices and a container orchestrator that controls a container operating in the information processing devices. , And the container orchestrator sets the distribution destination of the packet to the container daemon of the information processing apparatus.

Further, according to another aspect of the present invention, in the packet distribution method, the container daemon transmits the packet to the OS or packet on which the container operates based on the setting information of the distribution destination of the packet delivered from the container. Output to the high-speed processing layer.

Further, according to another aspect of the present invention, the program sends the packet to the computer based on the setting information of the distribution destination of the packet delivered from the container to the OS or the packet high-speed processing layer in which the container operates. Execute the process to output to.

According to the present invention, in a container-type virtualization environment, it is possible to realize QoS control of a network according to SLA without modifying the container side.

It is the first figure which shows the configuration example of the container type virtualization system by one Embodiment of this invention. FIG. 2 is a second diagram showing a configuration example of a container-type virtualization system according to an embodiment of the present invention. It is a 1st flowchart which shows an example of the process by one Embodiment of this invention. It is a 2nd flowchart which shows an example of the process by one Embodiment of this invention. It is a figure which shows the setting example of packet distribution by one Embodiment of this invention. It is the first figure which shows the application example of the container type virtualization system by one Embodiment of this invention. It is a 2nd figure which shows the application example of the container type virtualization system by one Embodiment of this invention. It is a figure which shows the minimum structure of the container daemon by one Embodiment of this invention. It is a figure which shows an example of the hardware configuration of the container type virtualization system in one Embodiment of this invention.

Hereinafter, the container-type virtualization system according to the embodiment will be described with reference to FIGS. 1 to 9.
(Explanation of configuration)
FIG. 1 is a first diagram showing a configuration example of a container-type virtualization system according to an embodiment of the present invention.
As shown in FIG. 1, the container-type virtualization system 1 includes a container orchestrator 10, container operating devices 20, 20a, 20b, 20c, and a relay device 30.
The container orchestrator 10 is a computer on which software such as a master function of Kubernetes is installed. The container operating devices 20 to 20c are computers on which software such as Kubernetes node functions and Docker are installed. The container orchestrator 10 is a master, and the container operating devices 20 to 20c are nodes. In the container operating devices 20 to 20c, the container operates. The relay device 30 is, for example, a computer on which software such as HAproxy is installed. Each of the container orchestrator 10 and the container operating devices 20 to 20c is communicably connected by a message path NW1 (management plane). The container orchestrator 10 and the relay device 30 are communicably connected by a message path NW0 (management plane). Control information flows through the message paths NW0 and NW1. Further, the container operating devices 20 to 20c and the relay device 30 are communicably connected by a network NW2 (data plane). Further, the relay device 30 is connected to the external network NW3. For example, when a container operated by the container operating device 20 or the like is accessed from various devices such as a PC, tablet, or smartphone connected to the external network NW3, the relay device 30 mediates the access and operates by the container operating device 20 or the like. Realize communication with the container.

The container orchestrator 10 includes an instruction unit 11, a selection unit 12, and a storage unit 13. The instruction unit 11 gives an instruction regarding packet distribution to the container operating devices 20 to 20c. The selection unit 12 selects a container operating device that activates a container that provides the service requested by the user from the container operating devices 20 to 20c. The storage unit 13 stores information on whether or not the packet high-speed processing function is introduced for each of the container operating devices 20 to 20c. For example, the storage unit 13 stores information that the high-speed packet processing function has already been introduced in the container operating device 20, and the high-speed packet processing function has not been introduced in the container operating devices 20a to 20c.

The container operating device 20 includes a container group 21-1,21-2,21-3, a container daemon 22, a packet high-speed processing layer 24, an OS 25, a NIC (Network Interface Card) 28, and a designated NIC 29. Be prepared.
The container groups 21-1 to 21-3 are one or more containers, respectively. For example, application 211 operates in container group 21-1, application 212 operates in container group 21-2, and application 213 operates in container group 21-3.
The container daemon 22 manages the container groups 21-1 to 21-3 operating on the container operating device 20. For example, the container daemon 22 receives various commands for the container groups 21-1 to 21-3, and starts or stops the container groups 21-1 to 21-3. Further, the container daemon 22 includes a packet distribution unit 23. The packet distribution unit 23 distributes the packets delivered by the container groups 21-1 to 21-3 to either the OS 25 or the packet high-speed processing layer 24 according to the instruction from the instruction unit 11 of the container orchestrator 10. For example, when the services provided by the applications 211 and 212 have priority over the services provided by the application 213, the packet distribution unit 23 processes the packets delivered by the container groups 21-1, 21-2 at high packet speed. It is distributed to the layer 24, and the packet delivered by the container group 21-3 is distributed to the OS 25.

The packet high-speed processing layer 24 is software that speeds up packet processing such as DPDK. In the present embodiment, as an example, it is assumed that software capable of QoS control (for example, DPDK QoS framework) is installed in the container operating device 20. The packet high-speed processing layer 24 bypasses the high-overhead packet processing provided by the OS 25 and directly accesses the designated NIC 29 for communication. Therefore, the speed of the network can be increased. Further, the packet high-speed processing layer 24 includes a QoS control function 241. For example, if the communication traffic of the container group 21-1 has a higher priority than the communication traffic of the container group 21-2, the QoS control function 241 delivers the packet delivered by the container group 21-1 to the container group. It is processed with priority over the packet delivered by 21-2. The designated NIC 29 is mounted for the packet high-speed processing layer 24, and the packet high-speed processing layer 24 occupies the designated NIC 29.

The OS 25 includes an OVS 26 and a kernel 27. OVS26 is a virtual switch. The kernel 27 performs general packet processing. The packets distributed from the container daemon 22 to the OS 25 are processed by the kernel 27 and distributed via the NIC 28. The packet processing by the kernel 27 has a higher load and takes longer than the processing by the packet high-speed processing layer 24. The kernel 27 may have a function (coloring function) for changing the value of DCSP (Differentiated Services Code Point) or IP Precedence of the packet header. Since the priority of packet processing can be changed by changing the values of DCSP and IP Precedence, QoS control is possible even when packet processing is performed via the OS 25.

FIG. 2 is a second diagram showing a configuration example of a container-type virtualization system according to an embodiment of the present invention. In FIG. 1, the configuration of the container operating device 20 including the packet high-speed processing layer 24 has been described. Next, with reference to FIG. 2, the configuration of the container operating device 20a in which the packet high-speed processing layer 24 is not introduced will be described.

The container operating device 20a includes a container group 21-1a, 21-2a, 21-3a, a container daemon 22a, an OS 25a, and a NIC 28a. The container daemon 22a includes a packet distribution unit 23a, and the OS 25a includes an OVS 26a and a kernel 27a. The container operating device 20a does not have a configuration corresponding to the packet high-speed processing layer 24 and the designated NIC 29 in FIG.
In an actual container-type virtualization environment, the packet high-speed processing layer 24 may not be introduced due to various reasons such as hardware restrictions and costs, or even if it is planned to be introduced, it may not be introduced yet. The container operating device 20a is a configuration example of a container host in such a case.

The container groups 21-1a to 21-3a, the container daemon 22a, the packet distribution unit 23a, the OS25a, the OVS26a, the kernel 27a, and the NIC28a are the container groups 21-1 to 21-3 and the container daemon 22 described in FIG. 1, respectively. , The packet distribution unit 23, OS25, OVS26, kernel 27, and NIC28 have the same functions.
In the case of the configuration of FIG. 2, for example, when the service provided by the application 211a has the highest priority, the application 212a has the highest priority, and the application 213a has the highest priority, the indicator of the container orchestrator 10 has the highest priority. The container daemon 22a is responsible for distributing all the packets delivered by the container groups 21-1a to 21-3a to the OS25a and processing the packets delivered by the container groups 21-1a to 21-3a with the above priority. Instruct to.

The packet distribution unit 23a distributes the packets delivered by the container groups 21-1a to 21-3a to the OS 25a according to the instruction from the instruction unit 11. Further, in the OS 25a, the packet delivered by the container group 21-1a has the highest priority, the packet delivered by the container group 21-2a has the second priority, and the container group 21-3a delivers according to the instruction from the instruction unit 11 by the kernel 27a. The header of the packet is rewritten so that the packet to be processed is processed with the third priority. As a result, the QoS control of the communication traffic becomes possible even for the container operating device in which the packet high-speed processing layer 24 is not introduced.

(motion)
Next, the operation of the container-type virtualization system 1 will be described with reference to FIGS. 3 to 5.
FIG. 3 is a first flowchart showing an example of processing according to the embodiment of the present invention.
FIG. 5 is a diagram showing a setting example of packet distribution according to an embodiment of the present invention.
First, the relay device 30 receives the user traffic (step S1). User traffic includes the type of application requested by the user. For example, suppose application 211, application 212, and application 213 are requested. The relay device 30 notifies the container orchestrator 10 of the information received from the user through the message path NW0 (step S2).

(Example 1)
In the storage unit 13 of the container orchestrator 10, information in which the priority of communication traffic is set in advance for each application is registered. For example, the applications 211, 212, and 213 are set with priorities 1, 2, and 3, respectively (priority 1 has the highest priority).
Further, in the storage unit 13, whether or not the packet high-speed processing layer 24 is introduced is registered for each of the container operating devices 20 to 20c. For example, it is registered that the packet high-speed processing layer 24 has already been introduced in the container operating device 20, and the others have not been introduced.
Further, the storage unit 13 is registered with a setting that the packet high-speed processing layer 24 processes the communication traffic of the applications of priority 1 and 2.
These settings are registered in advance in the container orchestrator 10 by the administrator.

Based on these settings, the selection unit 12 of the container orchestrator 10 identifies the container operating device 20 that meets the priority requirements of the applications 211 to 213 from the container operating devices 20 to 20c (step S3).

Next, the container orchestrator 10 deploys the container group 21-1 in which the application 211 operates, the container group 21-2 in which the application 212 operates, and the container group 21-3 in which the application 213 operates, to the specified container operating device 20. (Step S4). At this time, the instruction unit 11 sets the container daemon 22 to instruct the container daemon 22 about the packet distribution method and the QoS control to be delivered by the container groups 21-1 to 21-3. The setting contents are illustrated in FIG. 5A. In the case of this example, priority is given to (1) outputting the packets delivered by the container group 21-1 and the container group 21-2 to the packet high-speed processing layer 24, and (2) processing the packets delivered by the container group 21-1. Set degree 1 and set priority 2 for processing packets delivered by container group 21-2, and (3) set to output packets delivered by container group 21-3 to OS25. Do. This setting is stored in the container operating device 20 as, for example, a setting file referenced by the container daemon 22.
The instruction unit 11 may also instruct the packet high-speed processing layer 24 to output the packet delivered by the container group 21-3.

(Example 2)
Further, as another operation example, applications 211a to 213a are requested by the user in step S1, and the storage unit 13 of the container orchestrator 10 has the application 211a, the application 212a, and the application 213a have priority 4, respectively. It is assumed that 5 and 6 are set (priority 4 has the highest priority). In this case, since the application for which the priorities 1 and 2 are set is not requested, it is possible to select the container operating devices 20a to 20c in which the packet high-speed processing layer 24 is not introduced. The selection unit 12 of the container orchestrator 10 specifies, for example, the container operating device 20a from the container operating devices 20 to 20c (step S3).

Next, the container orchestrator 10 deploys the container group 21-1a in which the application 211a operates, the container group 21-2a in which the application 212a operates, and the container group 21-3a in which the application 213a operates on the specified container operating device 20a. (Step S4). At this time, the instruction unit 11 sets the container daemon 22a to instruct the container group 21-1a to 21-3a about the packet distribution method and the QoS control. The setting contents are illustrated in FIG. 5 (b). In the case of this example, (1) the packets delivered by the container groups 21-1a to 21-3a are output to the OS25a, and (2) the priority 4 is set for the processing of the packets delivered by the container group 21-1a. A setting is made to instruct that the priority 5 is set for the processing of the packet delivered by the container group 21-2a and the priority 6 is set for the processing of the packet delivered by the container group 21-3a. This setting is stored in the container operating device 20 as, for example, a setting file referenced by the container daemon 22.

Next, a packet distribution process in a situation where packets are actually delivered from the container group 21-1 and the like will be described with reference to FIG.
FIG. 4 is a second flowchart showing an example of processing according to the embodiment of the present invention.
(In the case of Example 1)
First, packets are delivered independently from the container groups 21-1 to 21-3. The container daemon 22 acquires the delivered packet (step S11).
Next, the packet distribution unit 23 of the container daemon 22 operates depending on the presence or absence of the packet high-speed processing layer 24 (step S12). When there is a packet high-speed processing layer 24, the packet distribution unit 23 adjusts whether to distribute to the packet high-speed processing layer 24 or to the OS 25 based on the setting of the instruction unit 11 (step S13).

In the case of Example 1 above, the packet distribution unit 23 assigns priority 1 to the packets delivered from the container group 21-1 and gives priority 2 to the packets delivered from the container group 21-2. Then, it is output to the packet high-speed processing layer 24. In the packet high-speed processing layer 24, the QoS control function 241 performs priority control (step S14). The QoS control function 241 processes the packet with priority 1 with higher priority. The packet high-speed processing layer 24 delivers the packet delivered from the container to the user side through the designated NIC 29 according to the set priority (step S15).

Further, for example, the packet distribution unit 23 outputs the packet delivered from the container group 21-3 to the OS 25. In the OS 25, the kernel 27 performs predetermined packet processing and delivers the packet delivered from the container to the user side through the NIC 28 (step S16).

(In the case of Example 2)
The container daemon 22a acquires the packets delivered from the container groups 21a-1 to 21-3a (step S11).
Next, the packet distribution unit 23a of the container daemon 22a operates depending on the presence or absence of the packet high-speed processing layer (step S12). When there is no packet high-speed processing layer, the packet distribution unit 23a adjusts to distribute the packet to the OS 25 based on the setting from the instruction unit 11 (step S17). For example, in the case of Example 2 above, the packet distribution unit 23a assigns priority 4 to the packets delivered from the container group 21-1a, and gives priority 5 to the packets delivered from the container group 21-2a. Is attached, and the packet delivered from the container group 21-3a is given a priority of 6 and output to the OS 25a. In OS25a, kernel 27a performs predetermined packet processing. Further, the kernel 27a performs priority control (step S18). For example, the kernel 27a rewrites the IP Precedence or DCSP value of the packet header according to the priority instructed by the container daemon 22a to change the packet priority. The kernel 27a delivers the priority-controlled packet to the user side through the NIC 28 (step S19).

According to the present embodiment, it is delivered from the container only by making the necessary settings for the container orchestrator 10 and introducing the container daemon 22 without modifying the container group 21-1 or the like of the container operating device 20. It enables high-speed packets and QoS control.
For example, in the case of the container operating device 20 in which the packet high-speed processing layer 24 is introduced, communication is performed only by setting distribution to the packet high-speed processing layer 24 for the desired container group 21-1 or the like operating in the container operating device 20. Traffic can be accelerated. Furthermore, by setting the priority, the priority between the containers can be adjusted.

Further, in the case of the container operating device 20a in which the packet high-speed processing layer 24 is not introduced, the container group 21-1a and the like are set to be distributed to the OS 25a, and the priority of the container groups 21-1a to 21-3a is set. You can adjust the priority of communication traffic between containers.

As a result, regardless of whether or not the packet high-speed processing layer 24 is introduced, it is possible to perform QoS control according to the SLA for the communication traffic of the container operating in the container-type virtualization system 1. Further, the application developer can concentrate on the development of the application without considering the introduction of the packet high-speed processing layer 24 into the container. In addition, application development and provision can be performed quickly.

(Example 1)
FIG. 6 is a first diagram showing an application example of a container-type virtualization system according to an embodiment of the present invention.
FIG. 6 shows a container group 21-1 in which the application 211 for making a voice call operates, a container group 21-2 in which the application 212 for video distribution operates, and a container group 21- in which the application 213 for distributing materials operates. An embodiment of providing a video conferencing service by the container operating device 20 including 3 is shown.
The SLA levels of Gold are set in the application 211, Silver is set in the application 212, and Bronze is set in the application 213. For example, the container orchestrator 10 makes settings for the container daemon 22 as illustrated in FIG. 5A. The audio data delivered from the container group 21-1 and the video data delivered from the container group 21-2 are delivered to the user side via the packet high-speed processing layer 24. The data of the material distributed from the container group 21-3 is distributed to the user side via the OS 25.
In the conference, in order to reliably deliver the voice data, the voice data is processed as the highest priority (Gold) to realize low jitter and low delay delivery. Video streaming is set as an intermediate priority (Silver), and the priority of audio and video is differentiated by the QoS control function 241 of the packet high-speed processing layer 24, but audio data is given higher priority than video streaming. Deterioration of the quality of material sharing has a small impact on the meeting, so it is given a low priority (Bronze).

(Example 2)
FIG. 7 is a second diagram showing an application example of the container-type virtualization system according to the embodiment of the present invention.
FIG. 7 shows an example of application to processing and processing services of IOT (Internet of Things) data.
There is no problem even if there is a Gold or time lag in the application 211 that processes medical data for which processing that is life-threatening or that does not allow a decrease in processing speed, but for example, the position of a moving object provided by GPS. The application 212, which requires a certain degree of real-time performance such as data processing, is set to Silver, and the application 213, which processes data such as meteorological data after a certain period of time, is set to each SLA level of Bronze.
Also in this case, the container orchestrator 10 sets the container daemon 22 as illustrated in FIG. 5A. The medical data delivered from the container group 21-1 and the position data delivered from the container group 21-2 are delivered to the user side via the packet high-speed processing layer 24. The meteorological data delivered from the container group 21-3 is delivered to the user side via the OS 25.
Even when independent services are provided from the applications of each container as in this example, it is possible to perform priority control of communication traffic on a container-by-container basis according to the service level of each application.

With the shift to microservices of applications, the process from application development to delivery is becoming faster. However, the focus is on application development and delivery, and the provision of SLAs associated with networks has been postponed. According to this embodiment, it is possible to provide a service satisfying the SLA at the same time without impairing the advantages of container-type virtualization such as rapid application development and rapid service provision. Further, regardless of the presence or absence of the packet high-speed processing layer in the container environment, it is possible to provide a service that satisfies the SLA by preferentially controlling packet processing. In addition, it is possible to eliminate the troublesome and complicated operation management that occurs when applying the packet high-speed processing layer in the container environment.

FIG. 8 is a diagram showing a minimum configuration of a container daemon according to an embodiment of the present invention.
As shown in FIG. 8, the container daemon 22 includes at least a packet distribution unit 23.
The packet distribution unit 23 outputs the packet to the OS in which the container operates or the packet high-speed processing layer based on the setting of the distribution destination for the packet delivered from the container. The container daemon 22 may include setting information of a packet distribution destination based on the presence / absence of the packet high-speed processing layer and the priority of packet processing for each container.

FIG. 9 is a diagram showing an example of the hardware configuration of the container-type virtualization system according to the embodiment of the present invention.
The computer 900 includes a CPU 901, a main storage device 902, an auxiliary storage device 903, an input / output interface 904, and a communication interface 905. The container operating devices 20, 20a to 20c described above are mounted on the computer 900. The operation of each of the above-mentioned functional units is stored in the auxiliary storage device 903 in the form of a program. The CPU 901 reads the program from the auxiliary storage device 903, expands it to the main storage device 902, and executes the above processing according to the program. Further, the CPU 901 secures a storage area in the main storage device 902 according to the program. Further, the CPU 901 secures a storage area for storing the data being processed in the auxiliary storage device 903 according to the program.

In at least one embodiment, the auxiliary storage device 903 is an example of a non-temporary tangible medium. Other examples of non-temporary tangible media include magnetic disks, magneto-optical disks, CD-ROMs, DVD-ROMs, semiconductor memories, etc. connected via the input / output interface 904. When this program is distributed to the computer 900 via a communication line, the distributed computer 900 may expand the program to the main storage device 902 and execute the above processing. Further, the program may be for realizing a part of the above-mentioned functions. Further, the program may be a so-called difference file (difference program) that realizes the above-mentioned function in combination with another program already stored in the auxiliary storage device 903.

In addition, it is possible to replace the components in the above-described embodiment with well-known components without departing from the spirit of the present invention. Further, the technical scope of the present invention is not limited to the above-described embodiment, and various modifications can be made without departing from the spirit of the present invention.

1 ... Container-type virtualization system 10 ... Container orchestrator 11 ... Indicator 12 ... Selection 13 ... Storage units 20, 20a, 20b, 20c ... Container operating device 21-1 , 21-2, 21-3, 21-1a, 21-2a, 21-3a ... Container group 22, 22a ... Container daemon 23, 23a ... Packet distribution unit 24 ... Packet high-speed processing Layer 241 ... QoS control function 25, 25a ... OS
26, 26a ... OVS
27, 27a ... Kernel 28, 28a ... NIC
29 ... Designated NIC
30 ... Relay device NW0, NW1 ... Message path NW2 ... Network NW3 ... External network 900 ... Computer 901 ... CPU
902 ... Main storage device 903 ... Auxiliary storage device 904 ... Input / output interface 905 ... Communication interface

Claims (10)

A packet distribution unit that outputs the packet to the OS in which the container operates or the packet high-speed processing layer based on the setting information of the distribution destination of the packet delivered from the container.
A container daemon with. It includes setting information of the distribution destination based on the presence / absence of the packet high-speed processing layer and the priority of packet processing for each container.
The container daemon according to claim 1. The container daemon according to claim 1 or 2,
OS and
The packet high-speed processing layer introduced in the OS and
Information processing device equipped with. The container daemon outputs the packet having a priority higher than a predetermined priority to the packet high-speed processing layer.
The packet having a priority lower than the predetermined priority is output to the OS.
The information processing device according to claim 3. The packet high-speed processing layer has a function of performing priority control on the processing of the packet.
The information processing device according to claim 3 or 4. The container daemon according to claim 1 or 2,
An OS that has a function to perform priority control on packet processing, and
Information processing device equipped with. The information processing apparatus according to any one of claims 3 to 6, and the information processing apparatus according to any one of claims 3 to 6.
A container orchestrator that controls the container operating in the information processing device,
With
The container orchestrator sets the packet distribution destination for the container daemon of the information processing device.
Container-type virtualization system. The container orchestrator stores information on whether or not the packet high-speed processing layer has been introduced for each information processing device.
When starting a container that executes a service that requires high-speed packets, the information processing device in which the packet high-speed processing layer has been introduced is selected, and the container is started by the selected information processing device.
The container-type virtualization system according to claim 7. The container daemon
The packet is output to the OS in which the container operates or the packet high-speed processing layer based on the setting information of the distribution destination of the packet delivered from the container.
Packet distribution method. On the computer
A process of outputting the packet to the OS in which the container operates or the packet high-speed processing layer based on the setting information of the distribution destination of the packet delivered from the container.
A program that executes.

Images