KR20190048924A - System and method for parallel processing flow-based data - Google Patents

Abstract

Provided are a data parallel processing system based on flow and an operation method thereof. Since the data parallel processing system based on flow includes a data reception unit, a flow table, a data management unit including a queue management unit, a kernel unit including at least one core, and an application unit, the data parallel processing system with high efficiency and high accuracy may maintain order of data and perform parallel processing, provide low latency and/or bandwidth guarantee for data with high priority by applying a priority of data, and may perform automatic core scaling depending on traffic conditions for data with low priority.

Description

[0001] SYSTEM AND METHOD FOR PARALLEL PROCESSING FLOW-BASED DATA [0002]

The present invention relates to a flow-based data parallel processing system and an operation method thereof, and more particularly, to a flow-based data parallel processing system that is processed in a multicore environment and a method of operating the same.

Cloud Computing is a method of using computing services such as server, storage, software, and data analysis over the Internet. In other words, unlike existing computing services, in which users own and manage resources, cloud computing has the characteristic that users are provided with the necessary resources in virtualized form over the Internet.

As a result, cloud computing is able to provide services on various platforms because the availability of computers is high and it is not necessary to construct individual contents for each terminal, and real-time data management is possible, Are used by users of the < / RTI >

However, as the number of users has increased recently, the amount of data processing required for cloud computing has increased, and data processing delays have arisen in data analysis and transmission and reception processes.

In addition, security problems such as personal information leakage cases are emerging due to external attacks of cloud computing.

As a result, edge cloud computing has recently emerged. Edge cloud computing is studying a system to prevent data latency and security problems by positioning a computing server close to a network close to a terminal, in contrast to cloud computing where existing computing servers are located close to the cloud central server.

However, this edge cloud computing technology also requires processing of sequential data in a multicore environment to handle services requiring low latency / bandwidth guarantees such as real-time IOT service, telemedicine service, high quality video service, AR / System is required.

Here, the data processing system in the multicore environment is a processing technology for speeding up the network traffic performance. Such a data processing system should maintain the order of the data in order even if the sequenced data is processed by one or more multicore cores in parallel at the same time.

In addition, in data parallel processing technology, a method for scaling a core of a processor according to the occurrence of a processing delay problem due to re-ordering of data and a situation of network traffic is required.

SUMMARY OF THE INVENTION It is an object of the present invention to provide a high-performance flow-based data parallel processing system with improved data processing speed and an operation method thereof.

Another object of the present invention is to provide a highly efficient flow-based data parallel processing system that prevents data re-ordering and an operation method thereof.

Another object of the present invention is to provide a highly efficient flow-based data parallel processing system capable of performing core scaling of a processor according to a network traffic situation and a method of operating the same.

According to another aspect of the present invention, there is provided a method of operating a flow-based data parallel processing system in a multicore environment, the method comprising: generating at least one flow by classifying received data according to attributes; Queuing the flow generated in at least one of the queues mapped on a priority basis with reference to a flow table including information on at least one specific queue corresponding to the flow, And allocating cores differently according to the priority of the flow stored in the core.

According to an aspect of the present invention, there is provided a method of operating a flow-based data parallel processing system, the method comprising: queuing the generated flows in at least one queue mapped according to a priority order, To a first queue having a higher priority on the flow table.

According to another aspect of the present invention, there is provided a method of operating a flow-based data parallel processing system, the method comprising the steps of: classifying the flows according to a priority order and queuing the flows in at least one queue, And queuing to a second queue having a lower priority on the flow table.

According to another aspect of the present invention, there is provided a method of operating a flow-based data parallel processing system, the method comprising: allocating a dedicated core to the first queue mapped to the high- Step < / RTI >

According to another aspect of the present invention, there is provided a method of operating a flow-based data parallel processing system, the method comprising: allocating cores differently to at least one second queue mapped to the low- The core can be dynamically allocated according to the situation.

According to another aspect of the present invention, there is provided a method of operating a flow-based data parallel processing system, the method comprising: when the generated flow is a new flow not mapped on the flow table, And queuing the new flow with reference to a queue table providing real-time information.

According to another aspect of the present invention, there is provided a method of operating a flow-based data parallel processing system, the method comprising: queuing the new flow of at least one second queue having a utilization rate below a threshold on the queue table; And if so, queuing the new flow to at least one of the second queues according to the set mode.

According to another aspect of the present invention, there is provided a method of operating a flow-based data parallel processing system, the method comprising: queuing the new flow, when a queue having a usage ratio lower than a threshold value does not exist on the queue table, And queuing the new flow to a newly created new queue.

According to another aspect of the present invention, there is provided a flow-based data parallel processing system including a power saving mode, A large amount of data can be allocated to the queued first sub-queue.

In order to achieve the above object, in the flow-based data parallel processing system according to an embodiment of the present invention, when the set mode is a process mode, the new flow is divided into a smallest of the at least one second queue Data can be allocated to the queued second sub-queue.

According to an aspect of the present invention, there is provided a flow-based data parallel processing system for performing a multi-core environment in a multi-core environment, the system including: a processor for receiving ordered data and classifying the received data according to an attribute of the data; And a data management unit including a data receiving unit for storing at least one flow and a flow table for providing information on at least one specific queue corresponding to a specific flow in order of priority, And queues the stored at least one flow in at least one queue that is mapped to the flow table.

A flow-based data parallel processing system and an operation method thereof according to an embodiment of the present invention classify and store at least one data having an order received from the outside into a flow of a specific criterion by a data receiving unit, A flow-based data parallel processing system capable of performing data processing by using a data processing method and an operation method thereof.

In addition, the flow-based data parallel processing system and the operation method thereof according to the embodiment of the present invention dynamically allocate a queue mapped according to the priority of the flow with reference to a flow table, thereby achieving a high- A flow-based data parallel processing system and an operation method thereof can be provided.

In the flow-based data parallel processing system and the operation method thereof according to an embodiment of the present invention, in the case of a new flow, a core is allocated according to a traffic situation in a second aggregation unit by a queue table, A flow-based data parallel processing system and an operation method thereof can be provided.

In addition, a flow-based data parallel processing system and an operation method thereof according to an embodiment of the present invention are provided with a flow-based data parallel processing system capable of high-speed data parallel processing by parallel operation of cores in a kernel unit .

1 is a conceptual diagram for explaining a flow-based data parallel processing system according to an embodiment of the present invention.
2 is a conceptual diagram for explaining a flow-based data parallel processing system according to an embodiment of the present invention.
3 is a conceptual diagram for explaining a flow-based data parallel processing system according to an embodiment of the present invention.
4 is a flowchart illustrating an operation method of a flow-based data parallel processing system according to an embodiment of the present invention.
5 is a flowchart illustrating an operation method of a flow-based data parallel processing system according to an embodiment of the present invention.

While the invention is susceptible to various modifications and alternative forms, specific embodiments thereof are shown by way of example in the drawings and will herein be described in detail. It should be understood, however, that the invention is not intended to be limited to the particular embodiments, but includes all modifications, equivalents, and alternatives falling within the spirit and scope of the invention. Like reference numerals are used for like elements in describing each drawing.

The terms first, second, A, B, etc. may be used to describe various components, but the components should not be limited by the terms. The terms are used only for the purpose of distinguishing one component from another. For example, without departing from the scope of the present invention, the first component may be referred to as a second component, and similarly, the second component may also be referred to as a first component. The term " and / or " includes any combination of a plurality of related listed items or any of a plurality of related listed items.

It is to be understood that when an element is referred to as being "connected" or "connected" to another element, it may be directly connected or connected to the other element, . On the other hand, when an element is referred to as being "directly connected" or "directly connected" to another element, it should be understood that there are no other elements in between.

The terminology used in this application is used only to describe a specific embodiment and is not intended to limit the invention. The singular expressions include plural expressions unless the context clearly dictates otherwise. In the present application, the terms "comprises" or "having" and the like are used to specify that there is a feature, a number, a step, an operation, an element, a component or a combination thereof described in the specification, But do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, or combinations thereof.

Unless defined otherwise, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Terms such as those defined in commonly used dictionaries are to be interpreted as having a meaning consistent with the contextual meaning of the related art and are to be interpreted as either ideal or overly formal in the sense of the present application Do not.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. In order to facilitate the understanding of the present invention, the same reference numerals are used for the same constituent elements in the drawings and redundant explanations for the same constituent elements are omitted. Hereinafter, preferred embodiments according to the present invention will be described in detail with reference to the accompanying drawings.

The present invention relates to a system and method for providing network virtualization, and more particularly, to a flow-based data parallel processing system and apparatus for providing network virtualization in a multi-core based edge cloud system environment.

1 and 2 are block diagrams of a flow-based data parallel processing system according to an embodiment of the present invention. More specifically, FIG. 1 is an overall configuration diagram of a flow-based data parallel processing system, and FIG. 2 is a partial configuration diagram for explaining a data management unit of a flow-based data parallel processing system.

Referring to FIGS. 1 and 2, the flow-based data parallel processing system may include a data management unit 1000, a kernel unit 2000, and an application unit 3000.

The data management unit 1000 may be configured to process a series of processes from the externally received data to the core to be described later.

More specifically, the data management unit 1000 may include a data receiving unit 1100, a flow table 1300, and a queue management unit 1500. According to an embodiment, the data management unit 1000 may be a network interface card (NIC).

The data receiving unit 1100 can receive at least one data input from the outside. At this time, the received data may be provided in the form of a packet with data having an order.

The data receiving unit 1100 may classify the received data according to an attribute and store the classified data in a flow.

According to the embodiment, the data receiving unit 1100 may extract header information from the received data. For example, the header information may be any one of SIP (Source Internet Protocol), DIP (Destination Internet Protocol), SPORT (Source Port), DPORT (Destination Port) or PROTOCOL.

Thereafter, the data receiving unit 1100 may classify the at least one data into specific criteria using the extracted header information.

The data receiving unit 1100 may separately store at least one data classified into a specific criterion into at least one flow according to a specific criterion. In other words, the flow may be an ordered set of data having the same attributes by a certain criterion.

According to an exemplary embodiment, the data receiver 1100 may generate a flow based on a bit masking value of the extracted at least one header information.

According to another embodiment, the data receiving unit 1100 may map each extracted header information to a predetermined size using a hash function. Thereafter, a flow may be generated by classifying data based on the mapping value.

Thereafter, the data receiving unit 1100 can request mapping information for the flow from the flow table 1300 to be described later.

According to the first embodiment, the data receiving unit 1100 can call the flow table 1300, which will be described later, according to a mapping value of a specific flow.

According to the second embodiment, the data receiving unit 1100 can use a value obtained by combining a mapping value of a specific flow and tenant information of data stored in a specific flow as an address of a flow table 1300 to be described later .

According to the third embodiment, the data receiving unit 1100 may use a value obtained by combining the mapping value of a specific flow and the characteristic information of each type of the virtual network function as an address of a flow table 1300 to be described later.

According to the fourth embodiment, the data receiving unit 1100 receives a mapping value of a specific flow and a value obtained by combining QOS class information of a virtual service function in which data stored in a specific flow is processed, with a flow table 1300 ) Can be used as the address.

The data receiving unit 1100 may quench data to at least one or more queues matching the flow according to the information of the flow table 1300 to be described later. In other words, the data receiving unit 1100 can receive the queue information mapped to a specific flow by referring to the flow table 1300 to be described later, and queue it to the queue.

On the other hand, when there is no information corresponding to a specific flow in the flow table 1300 to be described later, the data receiving unit 1100 transmits the data having the order stored in the specific flow to the queue management unit 1500 To the queue queuing unit 1530. [ The queue management unit 1500 and the queue queuing unit 1530 will be described in more detail below.

The data receiving unit 1100 may include a mode setting unit 1110. The mode setting unit 1110 may include a power saving mode and a process mode.

The mode setting unit 1110 can control a queue allocation for a specific flow in a queue processing unit 1570 to be described later. This will be described in more detail in the description of the queue processor 1570 to be described later.

The flow table 1300 may include at least one queue information mapped to each particular flow, as described above.

More specifically, the flow table 1300 classifies flow information according to a priority order and can provide queue information allocated for each flow. According to an embodiment, the flow table 1300 may arrange at least one flow on a priority basis. For example, the priorities can be sequentially prioritized according to whether low latency and bandwidth need simultaneous assurance, whether low latency is needed, and whether bandwidth assurance is needed.

According to one embodiment, when a specific flow is classified as a flow having a high priority on the flow table 1300, the flow is a dedicated queue having a high priority And may be queued to the first queue. At this time, the flow may be guaranteed to have low delay and bandwidth depending on the data attribute.

According to another embodiment, when a specific flow is classified as a flow having a low priority on the flow table 1300, the flow may be classified into at least one queue among the second queues Can be queued. Accordingly, an automatic core scale corresponding to the traffic situation can be provided. The queue allocation according to the priority of the flow will be described in more detail below.

The flow table 1300 can be set dynamically by statically setting queue allocation information for a specific flow through setting initialization or by adding or deleting a new priority flow. According to an embodiment, the flow table 1300 may initialize low-order flow information deactivated at regular intervals.

3 is a block diagram illustrating a queue management unit of a flow-based data parallel processing system according to an embodiment of the present invention.

1 to 3, the queue management unit 1500 may manage at least one queue in which at least one data having a sequence transmitted from the data reception unit 1100 is stored. The queue management unit 1500 may include a queue set unit 1510, a queue waiting unit 1530, a queue generating unit 1550, a queue processing unit 1570, and a queue setting unit 1590.

 The queue aggregation unit 1510 may include a first aggregation unit 1511 and a second aggregation unit 1515.

The first gathering unit 1511 may include at least one first queue. Here, the first queue is a queue having a high priority, and a high priority flow can be queued. For example, the first queue may be a dedicated queue dedicated to traffic of the flow.

The second gathering unit 1515 may include at least one second queue. Here, the second queue is a queue having a low priority, and a low priority flow can be queued. At this time, the at least one second queue may include a threshold value for the usage amount. The threshold of the second queue will be described in more detail below.

The queue aggregation unit 1510 may include queues having different attributes according to the flow environment of the data receiving unit 1100 according to the first and sixth embodiments described above.

According to an embodiment, when the data receiving unit 1100 uses a value obtained by combining tenant information as an address of the flow table 1300, the queue set unit 1510 may include at least one queue independent of each tenant.

According to another embodiment, when the data receiving unit 1100 uses a value obtained by combining characteristic information of each type of virtual network function as an address of the flow table 1300, the queue aggregating unit 1510 may allocate at least one independent Lt; / RTI > At this time, the number of queues can be provided differently according to the types of virtual network functions.

According to another embodiment, when the data receiver 1100 uses a value obtained by combining the QOS class information of the virtual service function as the address of the flow table 1300, the queue aggregator 1510 allocates the QOS class information of the virtual service function, And may include at least one queue. At this time, the number of queues may be differently provided for each QOS class of the virtual service function.

The queue queuing unit 1530 may be a space for temporarily waiting for a new flow to which the queue is not allocated. Here, the new flow may be a flow not mapped to the flow table 1300 among the specific flows received by the data receiving unit 1100. In other words, the new flow may be a flow having a lower priority.

As described above, when there is a new flow temporarily stored in the queue waiting unit 1530, the queue generating unit 1550 can generate a new queue by the signal transmitted from the queue waiting unit 1530. [

The queue processing unit 1570 can refer to the queue table and allocate a queue to at least one piece of data temporarily stored in the queue waiting unit 1530. [ At this time, the queue table may include real-time activation information of a queue included in the queue set unit 1510.

More specifically, the queue processor 1570 may check the utilization threshold of the second queue having a low priority before allocating the queue to the temporarily stored flow. The threshold may be set around the length of the queue.

According to one embodiment, when the usage rate of at least one second queue is less than the threshold value, the queue processing unit 1570 sets a new flow according to the mode setting unit 1110 to the first sub-queue and / You can optionally map to a queue.

Here, the first sub-queue may be a queue in which the largest amount of data among the second queues having a usage rate lower than the threshold value is queued, and the second sub-queue may be a queue in which the smallest data among the second queues having a usage rate below the threshold is queued have.

For example, when the mode setting unit 1110 is in the power save mode, the queue processing unit 1570 may map a new flow having a low priority to the first sub-queue.

On the other hand, when the mode setting unit 1110 is in the process mode, the queue processing unit 1570 can map a new flow having a low priority to the second sub-queue.

According to another embodiment, when all the queues of the queue set section 1510 indicate a usage ratio higher than the threshold value, the queue processing section 1570 can be mapped to the new queue generated by the queue generating section 1550, which will be described later.

In addition, the queue processor 1570 can allocate a core to be described later corresponding to a queue in the queue set 1510, respectively. More specifically, the queue processing unit 1570 can allocate cores corresponding to the respective queues in order to process at least one data queued in each queue in the queue set unit 1510. In other words, the queue processing unit 1570 can provide core scaling. Here, the core is a processor dedicated to at least one queue, and can transmit the processing result to the application unit 3000 to be described later.

The queue processing unit 1570 can allocate data having a sequence belonging to the same flow to a single core. Thus, the order of data can be maintained when processing data by the core.

In addition, the queue processing unit 1570 can parallel-process data having at least one flow information by operating at least one of the cores in parallel.

At this time, in the case of at least one data queued in the first queue as the first gathering unit 1511, the queue processing unit 1570 can allocate cores according to the data processing purpose.

According to one embodiment, when at least one piece of data mapped to the first queue of the first gathering unit 1511 needs to guarantee both low latency and bandwidth, the queue processing unit 1570 exclusively processes the dedicated queue At least one dedicated core may be allocated.

According to another embodiment, if at least one piece of data mapped to the first queue of the first aggregation unit 1511 is to provide low latency, then the queue processing unit 1570 may perform at least one Lt; / RTI > dedicated core or a single dedicated processing core that handles a plurality of dedicated queues.

According to another embodiment, when at least one piece of data mapped to a dedicated queue of the first aggregation unit 1511 needs to guarantee the bandwidth, the queue processing unit 1570 allocates at least one Lt; / RTI > dedicated core or a single dedicated processing core that handles a plurality of dedicated queues. The first queue may be a dedicated queue, as described above.

The queue setting unit 1590 can control whether a queue in the queue set unit 1510 is activated or not. More specifically, the queue setting unit 1590 can dynamically control whether the queue is activated by checking the flow table 1300 and the queue table at regular intervals.

According to the embodiment, the queue setting unit 1590 may refer to the flow table 1300 and the queue table to select a queue that can be deactivated, and then dynamically deactivate the queue. At this time, not only the queue but also the core processing the corresponding queue can be dynamically deactivated.

Referring again to FIG. 1, the kernel unit 2000 may include at least one core. As described above, the core can process at least one piece of data queued in a particular queue by the queue setting unit 1590. [

Application 3000 may include one or more user applications. The application unit 3000 can receive the processed data processing result from at least one core in the kernel unit 2000, as described above. For example, the application unit 3000 may include at least one application.

The flow-based data parallel processing system according to the embodiment of the present invention has been described above.

Hereinafter, an operation method of the flow-based data parallel processing system will be described.

4 is a flowchart illustrating a flow-based data parallel processing system according to an embodiment of the present invention.

1 to 4, a flow-based data parallel processing system may be operated. The flow-based data parallel processing system can receive at least one data from the outside (S1000). At this time, the data may be provided in the form of a packet, with data having an order.

Thereafter, the data may be queued (S2000). The step of queuing the data will be described in more detail with reference to FIG. 5 to be described later.

FIG. 5 is a flowchart illustrating a method of queuing data among operation methods of a flow-based data parallel system according to an embodiment of the present invention.

1 to 5, the data receiving unit 1100 may generate a flow from data (S2100).

As described above, the data receiving unit 1100 may be a flow in which the new flow does not have any information matched on the flow table 1300 from the received data.

According to an embodiment, the data receiving unit 1100 may generate a flow based on a bit masking value of the extracted at least one header information.

According to another embodiment, the data receiving unit 1100 may generate a flow by mapping each header information extracted by the hash function to a predetermined size.

Thereafter, referring to the flow table 1300, it is possible to check whether there is queue information corresponding to the generated flow (S2200).

If the generated flow belongs to a high priority flow on the flow table 1300 (S2300), the flow is transferred to the high priority flow 1310 on the flow table 1300 May be queued in the first queue which is the mapped high priority (S2310).

On the other hand, if the generated flow does not belong to the high priority flow 1310 on the flow table, it can be confirmed whether the flow is a new flow (S2350).

If the generated flow is not a new flow, as with the high priority flow, the flow may be a low priority flow mapped to the low priority flow 1350 on the flow table 1300 May be queued in the second queue (S2355).

If the generated flow is a new flow that does not exist on the flow table 1300, the generated new flow can be temporarily stored in the queue waiting unit 1530.

Then, in order to allocate a queue to a new flow stored in the queue queuing unit 1530, the queue table may be referred to (S2500).

If at least one queue on the queue table exceeds the threshold value (S2600), a new flow may be queued to the new queue generated by the queue generating unit 1550 (S2610). Thereafter, the new flow information may be updated on the flow table 1300 (S2615).

On the other hand, if at least one queue on the queue table does not exceed the threshold value, the setting mode of the mode setting unit 1110 can be confirmed (S2650). The queue table can provide real-time queuing information of the queues included in the queue set unit 1510.

According to one embodiment, when the mode setting unit 1110 is in the power save mode, the new flow may be queued to the first sub-queue (S2651). At this time, the first sub-queue may be a queue in which the most data among the at least one second queue is queued.

According to another embodiment, when the mode setting unit 1110 is in the process mode, the new flow may be queued to the second sub-queue (S2655). At this time, the second sub-queue may be a queue in which the least data among the at least one second queue is queued.

Thereafter, the new flow information may be updated on the flow table 1300 (S2700).

Referring again to FIG. 4, in order to process at least one data in the queued flow, each core may be assigned to at least one queue (S3000).

At this time, the core may be allocated differently according to the priority of the flow and / or the processing purpose of the data. In other words, core scaling can be provided.

Also, at least one core may operate in parallel. Thus, data in at least one flow can be processed in parallel.

Thereafter, the processed data may be transmitted to the application unit 3000 to be described later (S4000).

The flow-based data parallel processing system and its operation method according to the present invention have been described above.

The flow-based data parallel processing system and its operation method include a data receiving unit, a flow table, a data managing unit including a queue managing unit, a kernel unit including at least one core, and an application unit, It is possible to apply low priority data to high priority data by applying priority to data and to provide low latency and / or bandwidth guarantee for high priority data. A flow-based data parallel processing system can be provided.

The flow-based data parallel processing system and its operation method according to an embodiment of the present invention can be implemented as a computer-readable program or code in a computer-readable recording medium. A computer-readable recording medium includes all kinds of recording apparatuses in which data that can be read by a computer system is stored. The computer-readable recording medium may also be distributed and distributed in a networked computer system so that a computer-readable program or code can be stored and executed in a distributed manner.

Also, the computer-readable recording medium may include a hardware device specially configured to store and execute program instructions, such as a ROM, a RAM, a flash memory, and the like. Program instructions may include machine language code such as those produced by a compiler, as well as high-level language code that may be executed by a computer using an interpreter or the like.

While some aspects of the invention have been described in the context of an apparatus, it may also represent a description according to a corresponding method, wherein the block or apparatus corresponds to a feature of the method step or method step. Similarly, aspects described in the context of a method may also be represented by features of the corresponding block or item or corresponding device. Some or all of the method steps may be performed (e.g., by a microprocessor, a programmable computer or a hardware device such as an electronic circuit). In some embodiments, one or more of the most important method steps may be performed by such an apparatus.

In embodiments, a programmable logic device (e.g., a field programmable gate array) may be used to perform some or all of the functions of the methods described herein. In embodiments, the field programmable gate array may operate in conjunction with a microprocessor to perform one of the methods described herein. Generally, the methods are preferably performed by some hardware device.

It will be apparent to those skilled in the art that various modifications and variations can be made in the present invention without departing from the spirit or scope of the present invention as defined by the following claims It can be understood that

1000: Data management unit 1100: Data receiving unit
1150: Mode setting section 1300: Flow table
1310: High Priority Flow 1350: Low Priority Flow
1500: queue management unit 1510:
1511: First set part 1515: Second set part
1530: queue queuing 1550:
1570: a queue processing unit 1590: a queue setting unit
2000: Kernel part 3000: Application part dddd

Claims (1)

A method of operating a flow-based data parallel processing system in a multi-core environment,
Classifying the received data according to an attribute to generate at least one flow;
Queuing the generated flow in at least one queue mapped on a priority basis with reference to a flow table containing information about at least one specific queue corresponding to a particular flow;
And allocating cores differently according to the priority of the flow stored in the queue for the at least one queue.

Images