Distributed computing system for fault detection based on stream filtering and supervision

Abstract

A distributed computing system includes one or more computing devices and one or more communication links in electronic communication with each other. Each communication link includes one or more nodes and a port switch. One or more nodes are in electronic communication with each other through a communication link, and the communication link is connected to the distributed computing system through a port switch. The port switch executes instructions to identify an online unique flow configuration document for a particular flow and determine an expected unique flow configuration document bias during a specified time interval. The port switch determines a difference between the offline unique flow profile and the online unique flow profile, compares an absolute value of the difference to an expected unique flow profile bias, and in response to determining that the difference is greater than the expected unique flow profile bias, determines that a fault has occurred.

Description

Technical Field

[0001] The present disclosure relates to a distributed computing system for detecting failures based on flow filtering and policing. Background Art

[0002] A distributed computing system includes multiple computing devices (also called nodes) that send and receive information via communication links. The communication links can be implemented in hardware (e.g., a bus-based communication network) or wirelessly. For example, a distributed computing system can employ a 10Base-T network communication protocol, in which nodes communicate electronically with each other via twisted pair cables.

[0003] A specific example of a distributed computing system using the 10Base-T protocol includes one or more multi-drop buses that communicate with one or more computing devices. In this example, the node can represent an intelligent electrical device, such as an intelligent sensor or actuator with limited processing power. The computing device has greater processing power than the node, so functions and logical calculations can be performed on the computing device. Each bus includes a port switch that connects the node to the remaining buses that are part of the distributed computing system and computing devices. Sometimes, one or more nodes or one or more buses belonging to the distributed computing network will experience a failure causing one or more nodes to go offline. A node may go offline due to a loss of connection between the relevant bus and the relevant node, or when the relevant node becomes inoperable. When one or more nodes or buses go offline, the computing devices that are part of the distributed computing network need to receive notification.

[0004] Therefore, while distributed computing systems achieve their intended purposes, there remains a need in the art for a method of determining when one or more nodes or buses are no longer connected to the distributed computing system. Summary of the invention

[0005] According to several aspects, a distributed computing system includes: one or more computing devices and one or more communication links for electronic communication with the one or more computing devices, wherein each communication link includes one or more nodes and a port switch, the one or more nodes electronically communicate with each other through the communication link, and the communication link is connected to the distributed computing system through the port switch. The port switch executes instructions to determine whether the distributed computing system is online. In response to determining that the distributed computing system is online, the port switch identifies an online unique stream profile for a specific stream, wherein the specific stream corresponds to an application executed by one of the nodes in electronic communication with the port switch during a specified interval. The port switch determines the expected unique stream profile deviation during the specified time interval. At the end of the specified time interval, the port switch determines the difference between an offline unique stream profile and an online unique stream profile, wherein the offline unique stream profile represents the behavior of the specific stream when the distributed computing system is offline. The port switch compares the absolute value of the difference with the expected unique stream profile deviation. In response to determining that the difference is greater than the expected unique stream profile deviation, the port switch determines that a failure has occurred within the communication link.

[0006] In another aspect, in response to determining that a failure has occurred within the communication link, the port switch sends a notification to each computing device through the distributed computing system.

[0007] In yet another aspect, the failure applies to one of: one of the nodes and a communication link.

[0008] In one aspect, a particular flow generated by an application during a specified time interval includes a single traffic parameter.

[0009] On the other hand, the port switch determines that a fault has occurred by: subtracting the value of a single flow parameter of a specific flow represented by an offline unique flow configuration document from the value of a single flow parameter of a specific flow represented by an online unique flow configuration document to determine a single flow parameter difference; comparing the single flow parameter difference with an expected unique flow configuration document deviation; and in response to determining that the single flow parameter difference is greater than the expected unique flow configuration document deviation, determining that a fault has occurred.

[0010] In yet another aspect, a particular flow generated by an application during a specified time interval includes more than one flow parameter.

[0011] In one aspect, the port switch determines the failure by setting the expected unique flow profile deviation equal to a correlation factor of a first vector representing an offline unique flow profile and a second vector representing an online unique flow profile.

[0012] In another aspect, an expected unique flow profile deviation represents a predetermined amount of deviation that one or more traffic parameters of a particular flow are allowed to have without determining a failure within the communication link.

[0013] In yet another aspect, the traffic parameters include one or more of: packet transmission frequency, packet length, burst size, and packet priority.

[0014] In one aspect, the port switch performs per-flow filtering and policing based on the Institute of Electrical and Electronics Engineers (IEEE) 802.1Qci standard.

[0015] In another aspect, the system also includes a switching network connecting the communication link with the one or more computing devices.

[0016] In one aspect, the communication link is a bus implemented based on the 10Base-T1S network communication protocol.

[0017] On the other hand, a method for detecting a fault through a distributed computing system is provided. The method includes determining whether the distributed computing system is online through a port switch, wherein the distributed computing system includes one or more communication links, and each communication link includes one or more nodes and a port switch, and the one or more nodes communicate electronically with each other through the communication link, and the communication link is connected to the distributed computing system through the port switch. In response to determining that the distributed computing system is online, the method includes identifying, by the port switch, an online unique flow configuration document of a specific flow, wherein the specific flow corresponds to an application executed by one of the nodes that communicate electronically with the port switch during a specified time interval. The method includes determining, by the port switch, a unique flow configuration document deviation expected during the specified time interval. The method also includes determining, by the port switch at the end of the specified time interval, a difference between an offline unique flow configuration document and an online unique flow configuration document, wherein the offline unique flow configuration document represents the behavior of a specific flow when the distributed computing system is offline. The method includes comparing, by the port switch, the absolute value of the difference with the expected unique flow configuration document deviation. In response to determining that the difference is greater than the expected unique flow configuration document deviation, the method includes determining that a fault has occurred in the communication link.

[0018] In another aspect, in response to determining that a failure has occurred within the communication link, the method includes transmitting, by the port switch through the distributed computing system, a notification to one or more computing devices.

[0019] In yet another aspect, a distributed computing system includes one or more computing devices and one or more communication links for electronic communication with the one or more computing devices, wherein each communication link includes one or more nodes and a port switch, the one or more nodes electronically communicating with each other through the communication link, the communication link being connected to the distributed computing system through the port switch, and the port switch performing flow-by-flow filtering and supervision based on the IEEE 802.1Qci standard. The port switch executes instructions to determine whether the distributed computing system is online. In response to determining that the distributed computing system is online, the port switch identifies an online unique flow configuration document for a specific flow, wherein the specific flow corresponds to an application executed by one of the nodes in electronic communication with the port switch during a specified interval. The port switch determines the expected unique flow configuration document deviation during the specified time interval. At the end of the specified time interval, the port switch determines the difference between the offline unique flow configuration document and the online unique flow configuration document, wherein the offline unique flow configuration document represents the behavior of the specific flow when the distributed computing system is offline. The port switch compares the absolute value of the difference with the expected unique flow configuration document deviation. In response to determining that the difference is greater than the expected unique flow configuration document deviation, the port switch determines that a failure has occurred within the communication link. In response to determining that a failure has occurred within the communication link, the port switch sends a notification to each computing device through the distributed computing system.

[0020] In another aspect, the fault applies to one of: one of the nodes and a communication link.

[0021] In yet another aspect, a particular flow generated by an application during a specified time interval includes a single traffic parameter.

[0022] In one aspect, the port switch determines that a fault has occurred by: subtracting the value of a single flow parameter of a specific flow represented by an offline unique flow configuration document from the value of a single flow parameter of a specific flow represented by an online unique flow configuration document to determine a single flow parameter difference; comparing the single flow parameter difference with an expected unique flow configuration document deviation; and in response to determining that the single flow parameter difference is greater than the expected unique flow configuration document deviation, determining that a fault has occurred.

[0023] In another aspect, a particular flow generated by an application during a specified time interval includes more than one traffic parameter.

[0024] In yet another aspect, the port switch determines the failure by setting the expected unique flow profile deviation equal to a correlation factor of the first vector representing the offline unique flow profile and the second vector representing the online unique flow profile.

[0025] Further areas of applicability will become apparent from the description provided herein.It should be understood that the description and specific examples are intended for purposes of illustration only and are not intended to limit the scope of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] The drawings described herein are for illustration purposes only and are not intended to limit the scope of the present disclosure in any way.

[0027] Figure 1 is a schematic diagram of a vehicle including a disclosed distributed computing system including one or more computing devices and one or more communication links including one or more nodes according to an exemplary embodiment;

[0028] Figure 2A According to an exemplary embodiment, Figure 1 A software diagram of one of the nodes of a portion of a distributed computing system shown in ;

[0029] Figure 2B According to an exemplary embodiment, Figure 1 a software diagram of an Ethernet switch that is part of the distributed computing system shown; and

[0030] Figure 3 is a process flow diagram illustrating a method for detecting a failure within one of the communication links as part of a distributed computing system according to an exemplary embodiment. DETAILED DESCRIPTION

[0031] The following description is merely exemplary in nature and is not intended to limit the present disclosure, application, or uses.

[0032] Reference Figure 1 , a schematic diagram showing an exemplary distributed computing system 10 as part of a vehicle 12 is shown. Figure 1 In the non-limiting embodiment shown, the distributed computing system 10 is implemented as part of a vehicle, such as, but not limited to, a car, truck, sport utility vehicle, van, or RV. However, it should be understood that Figure 1 The disclosed distributed computing system 10 is merely exemplary in nature and is not limited to vehicles. In fact, the distributed computing system 10 can be implemented in various other applications, such as industrial automation control systems and the Internet of Things (IoT).

[0033] The distributed computing system 10 includes one or more communication links 22 that are in electronic communication with each other through a switching network 24. The switching network 24 includes multiple interconnected switches or a single switch. The one or more communication links 22 are also in electronic communication with one or more computing devices 32. Each communication link 22 is connected to the distributed computing system 10 through a port of the switch, where each port is referred to as a Figure 1 3 is a port switch 34. In the case of adopting a wireless protocol, the port switch 34 is a wireless access point.

[0034] In such Figure 1 In the illustrated embodiment, the distributed computing system 10 includes three communication links 22 and three computing devices 32, however, it should be understood that Figure 1 The communication link 22 is implemented in hardware such as a bus communication system or wirelessly. Figure 1 In the non-limiting embodiment shown, the communication link 22 is a bus implemented based on the 10Base-T1S network communication protocol. In another embodiment, the network communication protocol is based on switched Ethernet.

[0035] Each communication link 22 includes up to i nodes 30, where i is any integer greater than zero. In other words, each communication link 22 includes one or more nodes 30. Figure 1 In the illustrated embodiment, the communication link 22 is implemented based on the 10Base-T1S network communication protocol, and each communication link 22 can include up to eight nodes 30 or i=8, however, it should be understood that the maximum number of nodes 30 supported by each communication link 22 varies based on the network communication protocol. In a non-limiting embodiment, the node 20 can represent a device with limited computing capabilities, such as a smart sensor and a smart actuator. It should be understood that in an embodiment, one or more computing devices 32 have greater processing capabilities than the node 30, so the functions and logical calculations of the distributed computing system 10 can occur on the computing device 32.

[0036] Node 30 is represented as Where m represents the total number of communication links 22 included as part of the distributed computing system 10, and m>0. The communication links 22 are each denoted as B1, B2, ..., B m , the port switches 32 are each denoted as S1, S2, ... S l , and one or more computing devices 32 are represented as C1, C2, . . . C k , where l represents the total number of port switches 34 and k represents the total number of computing devices 32 included as part of the distributed computing system 10.

[0037] Figure 2A is a software diagram of one of the nodes 30 of the distributed computing system 10, where the node 30 is represented as Each node 30 executes at most j number of applications Among the applications is the form of service, and j is any integer greater than zero, or j>0. Each node 30 includes a physical address It is a unique value for the distributed computing system 10. For example, in the embodiment shown in the figure, the communication link 22 is implemented based on the 10Base-T1S network communication protocol, and the physical address is the Media Access Control (MAC) address. Each application of a particular node 30 includes a port value Unique to each application for a particular node 30 . Representation application The service identifier, and for the application The specific service being performed is unique. Representation application The service instance identifier of is unique to a specific service instance. Representation application The generated stream.

[0038] Figure 2B is a software diagram of an exemplary Ethernet switch representing one of the port switches 32, which is denoted as S l . It should be understood that each port switch 34 as part of the distributed computing system 10 performs per-flow filtering and policing based on the Institute of Electrical and Electronics Engineers (IEEE) 802.1Qci standard. It should also be understood that each port switch 34 as part of the distributed computing system 10 is configured to generate notifications based on unicast and multicast methods and forward the notifications to the computing devices 32. Each port switch 34 includes one or more processors for performing per-flow filtering and policing. For example, in one embodiment, the port switch 34 includes one or more central processing units (CPUs) based on a reduced instruction set computer (RISC) architecture.

[0039] Reference Figure 1 , Figure 2A and Figure 2B , offline only stream configuration document Indicates that when the distributed computing system 10 is offline, the specified time interval An application program executed by a particular node 30 during The specific stream generated It should be understood that when the distributed computing system 10 is online, the offline unique flow configuration document can be updated through the software update of the port switch 34. Offline unique stream configuration document is defined during the design time of the distributed computing system 10 and each port switch 34 is aware of the offline unique flow configuration document Offline unique stream configuration document is the corresponding physical address of a specific node 30 Corresponding port value Corresponding service identifier and the corresponding service instance identifier function, or Offline unique stream configuration document Indicates a specific flow One or more traffic parameters of the offline unique flow configuration document. Some examples of traffic parameters include, but are not limited to, packet transmission frequency, packet length, burst size, and packet priority. It should be understood that the offline unique flow configuration document Can be a scalar value or a multidimensional vector, depending on the number of flow parameters included.

[0040] It should be understood that the communication link 22 includes a multi-point topology. Therefore, each flow generated by a node 30 that is part of a particular communication link 22 passes through a corresponding port switch 34. Once the distributed computing system 10 is online, the particular flow The online unique flow configuration document is detected and identified by the port switch 34. Specifically, each flow generated by one of the nodes 30 of the distributed computing system 10 All of them pass through the corresponding port switch 34 and are detected and identified by the corresponding port switch 34 based on the IEEE802.1Qci standard. The corresponding port switch 34 detects the corresponding unique flow Online only stream configuration document Among them, the only online stream configuration document Included with offline only stream configuration document The same size. It should be understood that during design time, and by the time the distributed computing system 10 goes live (i.e., is broadcast), each port switch 34 of the distributed computing system 10 is configured with a unique flow and flow profile, where a flow refers to a data flow and a flow profile refers to a specific traffic behavior followed by a unique flow. m Each node 30 executes all applications as part of Connected to communication link B m The corresponding port switch S l Configure a specific stream in Offline only stream configuration document

[0041] The port switch 34 corresponding to the relevant communication link 33 determines the specified time interval Expected unique stream configuration document deviation during The only stream configuration document expected to deviate from Indicates a specific flow One or more traffic parameters are not determined to be faulty within the corresponding communication link 22 The predetermined deviation amount is determined when the distributed computing system 10 is offline. It should be understood that the predetermined deviation amount varies depending on the accuracy required for the specific application implementing the distributed computing system 10. It should be understood that in embodiments, one or more nodes 30 may not be able to generate a conventional stream configuration document. In this case, a low resource consumption application may be used instead. Alternatively, a composite stream configuration document may be determined in place of a regular stream configuration document.

[0042] At a specified time interval At the end, the corresponding port switch 34 determines the offline unique flow configuration document and online only stream configuration document and compares the absolute value of the difference with the expected unique stream configuration document deviation In response to determining that the offline unique stream configuration document With online only stream configuration document The difference between is greater than the expected unique stream configuration document deviation The corresponding port switch 34 determines that a failure has occurred in the communication link 22 It is expressed as |∏-π|>Δ∏. Specifically, the fault Suitable for a particular application executed by one of the nodes 30 as part of the communication link 22 In response to determining that the fault The corresponding port switch 34 sends a notification through the distributed computing system 10 to each computing device 32 that is part of the distributed computing system 10. The notification indicates that a failure has occurred.

[0043] It should be understood that this fault Applicable to one of the nodes 30 that is part of a particular communication link 22 or to the particular communication link 22 itself. If all applications executed by a particular node 30 of a particular communication link 22 have this fault Then the fault applies to that particular node 30. However, if the fault exists for all applications executed by all nodes 30 that are part of a particular communication link 22 Then the fault Applicable to the specific communication link 22.

[0044] The determination of the fault by the specific communication link 22 will now be described. It should be understood that if at a specified time interval During the application The specific stream generated If a single flow parameter is included, the port switch 34 determines that a failure has occurred by: The value of a single traffic parameter for a specific flow represented by minus the online unique flow configuration document The value of a single flow parameter for a specific flow indicated to determine a single flow parameter difference; the single flow parameter difference is deviated from the expected unique flow configuration document In response to determining that the single flow parameter difference is greater than the expected unique flow configuration document deviation Port switch 34 determines that the fault has occurred. In this example, the expected unique stream configuration document deviates Equal to a specific stream The single flow parameter does not determine the fault The predetermined deviation allowed under the circumstances. Determine the fault As shown in Formula 1:

[0045] |Ww|> ΔW Formula 1

[0046] Where W represents the offline unique stream configuration document The value of a single traffic parameter for a specific flow; w represents an online unique flow configuration document The value of a single flow parameter of a specific flow represented by ΔW represents the value of a single flow parameter of a specific flow. The single flow parameter does not determine the fault The predetermined deviation allowed in the case of The service instance identifier elements of, or in Represents a set of real numbers.

[0047] If at the specified time interval During the application The specific stream generated If more than one flow parameter is included, the port switch 34 deviates from the expected unique flow configuration document by Set to equal to the offline only stream configuration document The first vector and represents the only online stream configuration document The second vector of correlation factors determines the fault Specifically, in one example, the first vector is represented as a multidimensional vector [W, L], where W represents the first flow parameter, L represents the second flow parameter, and in in represents a set of real numbers. The second vector is represented as a multidimensional vector [w,l], where w represents the first flow parameter, l represents the second flow parameter, and in In one embodiment, the correlation factor is equal to cos(π / 4), however, it should be understood that the value of the correlation factor can range from 0 to 1. In one embodiment, when Equation 2 is true, the port switch 34 determines that the fault has occurred, where Formula 2 is expressed as:

[0048]

[0049] In one embodiment, when Equation 3 is true, the port switch 34 determines that the fault Wherein formula 3 is expressed as:

[0050]

[0051] Figure 3 is a process flow diagram illustrating a method 300 for detecting a failure within one of the communication links 22 or a node 30 as part of a distributed computing system 10. Referring generally to Figure 1 , Figure 2A , Figure 2B and Figure 3 , method 300 may begin at decision block 302. In decision block 302, port switch 34 of one of communication links 22 continues to monitor communication link 22 until distributed computing system 10 is online. Once port switch 34 determines that distributed computing system is online, method 300 may proceed to block 304.

[0052] In block 304, in response to determining that the distributed computing system 10 is online, the port switch 34 identifies a particular flow Online only stream configuration document Among them, a specific flow corresponding to a call made by one of the nodes 30 in electronic communication with the port switch 34 at a specified time interval Applications executed during The method 300 may proceed to block 306 .

[0053] In block 306, the port switch determines that at a specified time interval Unique stream configuration document deviation expected during The method may then proceed to block 308 .

[0054] In block 308, at specified time intervals At the end, the port switch 34 determines that the offline unique flow configuration document With online only stream configuration document The difference between the offline unique stream configuration document Indicates a specific flow when the distributed computing system 10 is offline Then, the method 300 may proceed to decision block 310 .

[0055] In decision block 310, the port switch 34 deviates the absolute value of the difference from the expected unique flow profile. In response to determining that the difference is less than or equal to the expected unique stream configuration document deviation The port switch 34 determines that no failure has occurred and the method 300 may terminate.

[0056] In response to determining that the difference is greater than the expected unique stream configuration document deviation The method 300 may proceed to block 312. In block 312, the port switch determines that the failure has occurred within the communication link 22. The method 300 may then proceed to block 314 .

[0057] In block 314, the port switch 34 sends a notification to each computing device 32 via the distributed computing system, wherein the notification indicates the failure has occurred. Method 300 may then terminate.

[0058] Referring generally to the accompanying drawings, the disclosed distributed computing system provides various technical effects and benefits. Specifically, the distributed computing system provides a method for detecting faults within a communication link using existing per-flow filtering and policing methods. It should be understood that because the disclosed method for detecting faults utilizes standard per-flow filtering and policing already in place within the system, no additional hardware is required. Compared to existing methods, the disclosed method can improve detection time and improve bandwidth because no additional packet transmission is required to detect faults.

[0059] Port switches and computing devices may refer to electronic circuits such as in a system on chip, combinatorial logic circuits, field programmable gate arrays (FPGAs), processors (shared, dedicated, or groups) that execute code, or a combination of some or all of them, or a portion thereof. In addition, the controller may be microprocessor-based, such as a computer having at least one processor, memory (RAM and / or ROM), and associated input and output buses. The processor may operate under the control of an operating system residing in the memory. The operating system may manage computer resources so that computer program codes embodied as one or more computer software applications (e.g., applications residing in the memory) may have instructions executed by the processor. In an alternative embodiment, the processor may execute the application directly, in which case the operating system may be omitted.

[0060] The description of the present disclosure is merely exemplary in nature, and variations that do not depart from the gist of the present disclosure are intended to fall within the scope of the present disclosure. These variations should not be regarded as departing from the spirit and scope of the present disclosure.

Claims

1. A distributed computing system, comprising: one or more computing devices; and one or more communication links in electronic communication with the one or more computing devices, wherein each communication link comprises one or more nodes and a port switch, and wherein the one or more nodes are in electronic communication with each other via the communication links, and the communication links are connected to the distributed computing system via the port switches, wherein the port switches execute instructions to: determining whether the distributed computing system is online; responsive to determining that the distributed computing system is online, identifying an online unique flow configuration document for a particular flow, wherein the particular flow corresponds to an application executed by one of the nodes in electronic communication with the port switch during a specified time interval; determining the expected unique stream configuration document deviation during a specified time interval; At the end of the specified time interval, determining a difference between an offline unique flow configuration document and an online unique flow configuration document, wherein the offline unique flow configuration document represents the behavior of the particular flow when the distributed computing system is offline; comparing the absolute value of the difference to the expected unique stream configuration document deviation; and In response to determining that the difference is greater than the expected unique flow profile deviation, it is determined that a failure has occurred within the communication link.

2. The distributed computing system according to claim 1, wherein: The port switch executes instructions to: In response to determining that the failure has occurred within the communication link, a notification is sent to each computing device through the distributed computing system.

3. The distributed computing system according to claim 1, wherein: The failure applies to one of: one of the nodes and the communication link.

4. The distributed computing system according to claim 1, wherein: The particular flow generated by the application during the specified time interval includes a single flow parameter.

5. The distributed computing system according to claim 4, wherein: The port switch determines that a fault has occurred by: Subtract the value of the single traffic parameter of the specific flow represented by the offline unique flow configuration document from the value of the single traffic parameter of the specific flow represented by the online unique flow configuration document to determine a single traffic parameter difference; comparing the single flow parameter difference value to the expected unique flow configuration document deviation; as well as In response to determining that the single flow parameter difference is greater than the expected unique flow profile deviation, it is determined that the failure has occurred.

6. The distributed computing system according to claim 1, wherein: The particular flow generated by the application within the specified time interval includes more than one traffic parameter.

7. The distributed computing system according to claim 6, wherein: The port switch determines the failure by setting the expected unique flow profile deviation equal to a correlation factor of a first vector representing the offline unique flow profile and a second vector representing the online unique flow profile.

8. The distributed computing system according to claim 1, wherein: The expected unique flow profile deviation represents a predetermined amount of deviation that is permitted for one or more traffic parameters of the particular flow without determining the fault within the communication link.

9. The distributed computing system according to claim 8, wherein: The traffic parameters include one or more of the following: packet transmission frequency, packet length, burst size and packet priority.

10. The distributed computing system according to claim 1, wherein: The port switches perform per-flow filtering and policing based on the Institute of Electrical and Electronics Engineers (IEEE) 802.1Qci standard.

Images