Distributed network data analysis system and method supporting business instantaneous anomaly detection

Through the distributed network data joint analysis system, the problem that existing network analyzers cannot meet the instantaneous abnormality detection of high-bandwidth services is solved, and millisecond capture and analysis of high-bandwidth data is realized, and synchronous detection of multi-level basic network equipment is supported. It has stronger flexibility and detection capabilities to help quickly locate abnormal problems.

CN115766491BActive Publication Date: 2024-12-03SHANGHAI CHENYU INFORMATION TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202211324112.3
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-10-27
Publication Date
2024-12-03
Estimated Expiration
2042-10-27

AI Technical Summary

Technical Problem

Existing network analyzers cannot meet the detection requirements of instantaneous abnormalities of high bandwidth services, especially in the real-time video Ethernet application field, which cannot capture high bandwidth service data for a long time, and cannot locate deep-level problems across multiple basic network devices. .

Method used

A distributed network data joint analysis system was designed, including a distributed network data capture device, a distributed network data joint analysis platform and a distributed network data joint analysis client, which connects the basic network switching equipment through standard network link interfaces to realize Distributed data capture and joint analysis support instantaneous anomaly detection of high-bandwidth service data.

Benefits of technology

It realizes millisecond detection cycle of high-bandwidth service data, supports dynamic adjustment, can quickly locate synchronization problems of multi-level basic network equipment, has stronger detection capabilities and flexibility, and can capture and analyze abnormal data in high-bandwidth environments. , help users quickly locate the cause of the problem.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115766491B_ABST
    Figure CN115766491B_ABST
Patent Text Reader

Abstract

The present invention belongs to the field of network communication technology, and is specifically a distributed network data analysis system and method that supports instantaneous service anomaly detection. The system includes a distributed network data capture device, a distributed network data joint analysis platform, and a distributed network data joint analysis client. The system supports multiple distributed network data capture devices distributed at any different locations of the network, and unified management through the distributed network data joint analysis platform; the system can set specific joint capture conditions to realize the detection and analysis of instantaneous anomalies of network services, thereby assisting users to quickly locate unstable factors that affect services in the entire network and restore the normal operation of the network.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the field of network communication technology, and specifically relates to a distributed network data joint analysis system and method supporting business instantaneous anomaly detection. Background Art

[0002] In the field of network communication technology, the transmission of business data is based on the data packet exchange function provided by the basic network facilities. During the transmission of business data, it will go through multiple levels of data relay and exchange, and some abnormalities often occur in the middle process. These abnormalities may be caused by the physical interface of the basic network facilities, such as loose network interface and poor contact; they may also be caused by the transmission bottleneck of the link relay facilities, such as insufficient aggregation bandwidth leading to data loss; or they may be caused by some problems in the business system itself, rather than the basic network facilities.

[0003] Of course, basic network facilities provide some basic diagnostic methods. For example, general switches will provide basic parameters such as port status, traffic, and packet loss to help users locate problems. However, some deep-seated problems cannot be located through these basic diagnostic methods of basic network facilities. Therefore, third-party network diagnostic equipment is needed to determine and analyze the root cause of the abnormality.

[0004] Network analyzers are the most commonly used third-party network diagnostic equipment. With the help of network analyzers, users can capture network data that meets specific conditions and then perform offline analysis. Some common conditions include MAC address, IP address, network protocol, etc.

[0005] However, the network analyzers currently on the market still cannot meet the requirements for detecting instantaneous anomalies in high-bandwidth services. For example, in the field of real-time video Ethernet applications, the service bandwidth can reach G bit / s. The storage capacity limitation of the network analyzer makes it impossible to capture such a large bandwidth of service data for a long time, and the entire service data path needs to span multiple basic network devices (such as switches). Summary of the invention

[0006] The purpose of the present invention is to further improve the shortcomings of existing network analyzers and propose a distributed network data joint analysis system and method that supports business transient anomaly detection, which can quickly locate the cause of business transient anomalies.

[0007] To achieve the above purpose, the present invention adopts the following technical solution.

[0008] A distributed network data joint analysis system supporting business instantaneous anomaly detection includes a distributed network data capture device, a distributed network data joint analysis platform and a distributed network data joint analysis client; the distributed network data capture device, the distributed network data joint analysis platform, the distributed network data joint analysis client and the basic network switching device are connected through a standard network link interface; wherein:

[0009] The distributed network data joint analysis client provides a user human-computer interaction interface, receives analysis parameters input by the user in the interface, and displays the data capture process status and joint analysis results to the user;

[0010] The distributed network data joint analysis platform receives analysis parameters sent by the distributed network data joint analysis client, converts the analysis parameters into interface instructions including capture detection trigger conditions and publishes them to each distributed network data capture device, monitors the status of each distributed network data capture device, uploads the captured data of each distributed network data capture device, performs joint analysis, and notifies the distributed network data joint analysis client of the joint analysis results;

[0011] There are several distributed network data capture devices, which are distributed in different locations of the network, and the number of configurations is not less than the number of levels of basic network switching equipment. They capture the business data of a certain level of basic network switching equipment according to the configuration parameters and instructions of the distributed network data joint analysis platform.

[0012] In the present invention, the distributed network data capture device includes a network interface module, a parameter configuration module, a time synchronization module, a data packet filtering module, a capture detection module, a capture data writing module, a data storage module, a capture data reading module, and a capture data uploading module; wherein:

[0013] The network interface module is directly connected to the basic network switching device. The basic network device copies the port carrying the service data to this module by mirroring to obtain all data packets of the service port.

[0014] The parameter configuration module obtains the parameters issued by the distributed network data joint analysis platform, including data packet filtering conditions, capture detection trigger conditions and time synchronization information;

[0015] A time synchronization module receives time synchronization information and sets the local time of the distributed network data capture device to keep it consistent with the time of the distributed network data joint analysis platform;

[0016] The data packet filtering module filters the data sent by the network interface module according to the received data packet filtering conditions. Only the data that meets the filtering conditions can pass through this module;

[0017] The capture detection module generates a detection trigger condition according to the received capture detection trigger condition, and once the detection trigger condition is met, after a certain period of time, notifies the capture data writing module to stop writing to the data storage module;

[0018] The capture data writing module receives the write command sent by the capture detection module, and then continuously writes data into the data storage module; when receiving the stop write command sent by the capture detection module, the write operation is stopped;

[0019] The data storage module receives the data writing instruction from the captured data writing module and saves the data; at the same time, it receives the reading instruction from the captured data reading module and retrieves the corresponding data;

[0020] The captured data reading module receives the read-back command from the captured data uploading module, then takes out the business data and sends it to the captured data uploading module;

[0021] The capture data upload module receives the capture readback instruction issued by the distributed network data joint analysis platform, notifies the capture data reading module to read the business data, and then frames the received business data into data packets and sends them back to the distributed network data joint analysis platform.

[0022] In the present invention, the analysis parameters set in the distributed network data joint analysis client include: business data source IP address, MAC address, destination IP address, MAC address, business anomaly detection cycle, business anomaly bandwidth.

[0023] In the present invention, the basic network switching device is a switch.

[0024] In the present invention, the network interface module in the distributed network data capture device includes a Gigabit Ethernet interface, a 10 Gigabit Ethernet interface and a 100G optical fiber Ethernet interface.

[0025] The present invention further provides a method for detecting service transient anomalies using the above-mentioned distributed network data joint analysis system, and the specific process is as follows:

[0026] (1) configuring a corresponding number of distributed network data capture devices according to the number of levels of network switching devices, assuming that the number of levels of basic network devices is N, and N is a positive integer, then the number of configured distributed network data capture devices is at least N;

[0027] (2) connecting the network interface of the distributed network data capture device to the corresponding basic network device, and configuring the port mirroring associated with the current level of service data to the network interface of the current level of distributed network data capture device;

[0028] (3) Set the capture trigger conditions according to the service characteristics, set the minimum detection period T of the service data, and set the abnormal data bandwidth range [Sa, Sb] of the service;

[0029] (4) The distributed network data joint analysis platform sends the capture trigger conditions to the distributed network data capture devices at all levels and configures the trigger delay Td;

[0030] (5) The distributed network data joint analysis platform sends a capture start instruction to distributed network data capture devices at all levels to start the trigger capture function;

[0031] (6) When any distributed network data capture device meets the trigger condition, it stops capturing after a delay of Td and notifies the distributed network data joint analysis platform that the capture is successful;

[0032] (7) After receiving the information that any distributed network data capture device has successfully captured data, the distributed network data joint analysis platform immediately stops the capture work of all other distributed network data capture devices and pushes the capture success information to the distributed network data joint analysis client;

[0033] (8) After the distributed network data joint analysis client obtains the capture success information, it informs the user; if the user wants to analyze this capture, the data analysis command is started, and the distributed network data joint analysis platform uploads the data captured by the distributed network data capture devices at all levels and analyzes it; after the analysis is completed, the results are pushed to the user, and the user can quickly locate the cause of the abnormal problem based on the analysis results.

[0034] Compared with the prior art, the distributed network data joint analysis system and method supporting business instant anomaly detection disclosed in the present invention has the following beneficial effects:

[0035] 1. Supporting the detection of instantaneous anomalies of high-bandwidth business data: The detection cycle of the system of the present invention supports dynamic adjustment. The general equipment flow detection takes several minutes, while the detection cycle of the present invention can be supported to the millisecond level. Once the flow within the detection cycle is abnormal, it can be captured immediately; it has stronger detection capability and flexibility;

[0036] 2. It supports synchronous joint detection of multi-level basic network equipment, which can obtain data of different network nodes at the same time, facilitate global analysis and quickly locate problems. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 The structure diagram of a distributed network data joint analysis system that supports business instant anomaly detection.

[0038] Figure 2It is a structural block diagram of a distributed network data capture device. DETAILED DESCRIPTION

[0039] The present invention is further described in detail below with reference to the accompanying drawings and embodiments.

[0040] Example 1

[0041] like Figure 1 As shown, a distributed network data joint analysis system supporting business instant anomaly detection includes a distributed network data capture device, a distributed network data joint analysis platform, and a distributed network data joint analysis client. The distributed network data capture device, the distributed network data joint analysis platform, and the distributed network data joint analysis client are connected to the basic network switching equipment through a standard network link interface.

[0042] The distributed network data capture device captures the business data of a certain level of basic network switching equipment according to the configuration parameters and instructions of the distributed network data joint analysis platform; the distributed network data joint analysis platform receives the parameters configured by the user through the distributed network data joint analysis client, converts these parameters into interface instructions and sends them to each distributed network data capture device; the distributed network data joint analysis client mainly satisfies the user's interface interaction and receives the parameters configured by the user through the interface.

[0043] The structure and workflow of the distributed network data capture device, the distributed network data joint analysis platform, and the distributed network data joint analysis client are specifically introduced as follows.

[0044] For the convenience of description, in this embodiment, the basic network switching equipment is set to 3 levels, where the first level is the access switch, the network interface is a Gigabit Ethernet port, the second level is the aggregation switch, the network interface is a 10G fiber Ethernet interface, and the third level is the core switch, the network interface is a 100G fiber Ethernet interface; the detection period is 1 millisecond, the service abnormal bandwidth is greater than or equal to 0, and less than 200Mb per second;

[0045] 1) Distributed network data joint analysis client

[0046] The distributed network data joint analysis client provides a human-computer interaction interface, receives the analysis parameters entered by the user in the interface, and displays the data capture process status and joint analysis results to the user. Its workflow is as follows:

[0047] 1. The user connects to the distributed network data joint analysis platform through the network and logs in successfully;

[0048] 2. Searching for distributed network data capture devices in the network to obtain online distributed network data capture devices;

[0049] 3. Create a joint capture group and add the required distributed network data capture devices to the joint capture group;

[0050] 4. Create a joint capture trigger condition. In this embodiment, the service data information to be captured is added, such as the service data source IP address, MAC address, destination IP address, MAC address, service anomaly detection cycle, service anomaly bandwidth, etc.

[0051] 5. Enable the capture trigger condition, send it to the distributed network data joint analysis platform, and start the distributed network data capture device in the joint capture group;

[0052] 6. Monitor the messages of the distributed network data joint analysis platform. Once the capture is successful, the user will be notified immediately by means of pop-up windows or sounds;

[0053] 7. After receiving the capture success, the user notifies the distributed network data joint analysis platform to start data transmission and joint analysis;

[0054] 8. After the distributed network data joint analysis platform completes the data analysis, it notifies the client and displays the analysis results to the user;

[0055] 2) Distributed network data joint analysis platform

[0056] The distributed network data joint analysis platform receives the analysis parameters sent by the distributed network data joint analysis client, generates capture parameter commands based on the analysis parameters sent by the client and publishes them to each distributed network data capture device, monitors the status of each distributed network data capture device, uploads the captured data of each distributed network data capture device, and performs joint analysis, and notifies the distributed network data joint analysis client of the analysis results. Its workflow is as follows:

[0057] 1. Establish a connection with the distributed network data joint analysis client through the network to verify the user and password;

[0058] 2. Monitor the status of each distributed network data capture device in the network and report it to the network data joint analysis client;

[0059] 3. Receive the joint capture group parameters from the distributed network data joint analysis client, parse the parameters to generate capture parameter commands and publish them to each distributed network data capture device in the joint capture group;

[0060] 4. Upon receiving the capture start instruction from the distributed network data joint analysis client, each distributed network data capture device in the joint capture group is immediately started to trigger capture;

[0061] 5. Monitor the capture status of each distributed network data capture device in the joint capture group. Once any device captures successfully, stop all other distributed network data capture devices in the joint capture group and send a capture success message to the client;

[0062] 6. Upon receiving the data feedback and analysis instructions from the distributed network data joint analysis client, poll and collect the service data captured by each distributed network data capture device in the capture group, and analyze them simultaneously;

[0063] 7. After the analysis is completed, the analysis results are sent to the distributed network data joint analysis client, which is displayed to the user;

[0064] 3) Distributed Network Data Capture Device

[0065] A distributed network data capture device includes a network interface module, a parameter configuration module, a time synchronization module, a data packet filtering module, a capture detection module, a capture data writing module, a data storage module, a capture data reading module, and a capture data uploading module; wherein:

[0066] The network interface module is directly connected to the basic network equipment (such as a switch). The basic network equipment copies the port carrying the service data to this module by mirroring, so that all data packets of the service port can be obtained;

[0067] The parameter configuration module obtains the parameters issued by the distributed network data joint analysis platform, including data packet filtering conditions, capture detection trigger conditions, time synchronization information, etc.

[0068] The time synchronization module receives time synchronization information and sets the local time to keep it consistent with the time of the distributed network data joint analysis platform;

[0069] The data packet filtering module filters the data sent by the network interface module according to the parameters issued by the distributed network data joint analysis platform. Only data that meets the conditions can pass through this module;

[0070] The capture detection module generates a detection trigger condition according to the parameters issued by the distributed network data joint analysis platform. Once the detection trigger condition is met, after a certain period of time, the capture data writing module is notified to stop writing to the data storage module;

[0071] The capture data writing module receives the write command sent by the capture detection module, and then continuously writes the data into the data storage module; when receiving the stop write command sent by the capture detection module, the write operation is stopped;

[0072] The data storage module receives the data writing instruction from the captured data writing module and saves the data; at the same time, it receives the reading instruction from the captured data reading module and retrieves the corresponding data;

[0073] The captured data reading module receives the read-back command from the captured data uploading module, then takes out the business data and sends it to the captured data uploading module;

[0074] The capture data upload module receives the capture readback instruction issued by the distributed network data joint analysis platform, notifies the capture data reading module to read the business data, and then frames the received business data into data packets and sends them back to the distributed network data joint analysis platform.

[0075] In this embodiment, the system also supports the detection of transient anomalies of high-bandwidth services, and the specific detection process is as follows:

[0076] 1. Configure a corresponding number of distributed network data capture devices according to the number of levels of the basic network switching device. If the number of levels of the basic network device is N (N is a positive integer), then the number of distributed network data capture devices configured is at least N. In this embodiment, N = 3;

[0077] 2. Connect the network interface of the distributed network data capture device to the corresponding basic network device, and configure the port mirroring associated with the current-level business data to the network interface of the current-level distributed network data capture device; in this embodiment, the first-level distributed network data capture device is connected to the first-level access switch through a Gigabit Ethernet interface; the second-level distributed network data capture device is connected to the second-level aggregation switch through a 10G optical fiber Ethernet interface; the third-level distributed network data capture device is connected to the third-level core switch through a 100G optical fiber Ethernet interface;

[0078] 3. Set the capture trigger conditions according to the service characteristics, set the minimum detection period T of the service data, and set the abnormal data bandwidth range [Sa, Sb] of the service; in this embodiment, the detection period is T = 1 millisecond, Sa = 0, Sb = 200Mb;

[0079] 4. The distributed network data joint analysis platform sends the capture trigger conditions to the distributed network data capture devices at all levels and configures the trigger delay Td; in this embodiment, Td = 200 milliseconds;

[0080] 5. The distributed network data joint analysis platform sends a capture start instruction to distributed network data capture devices at all levels to start the trigger capture function;

[0081] 6. When any distributed network data capture device meets the triggering condition, it will stop capturing after a delay of 200 milliseconds and notify the distributed network data joint analysis platform that the capture is successful;

[0082] 7. After receiving the information that any distributed network data capture device has successfully captured, the distributed network data joint analysis platform immediately stops the capture work of all other distributed network data capture devices and pushes the capture success information to the distributed network data joint analysis client;

[0083] 8. After the distributed network data joint analysis client obtains the capture success information, it informs the user; if the user wants to analyze this capture, the data analysis command is started, and the distributed network data joint analysis platform uploads the data captured by the distributed network data capture devices at all levels and analyzes it; after the analysis is completed, the results are pushed to the user, and the user can quickly locate the cause of the abnormal problem based on the analysis results.

[0084] Detection example:

[0085] Thousands of streaming media services are running in a certain network center. When the decoder switches the media stream, signal freezes and the picture becomes pixelated occasionally. The phenomenon can be automatically restored after a few seconds. The probability of occurrence is about one in ten thousand. Since the signal source and decoder of the media stream are distributed on different switches or routers in the network, and the abnormality occurs for a very short time, existing network analysis methods cannot locate it.

[0086] By adopting the solution of the present invention, the problem can be quickly located: the distributed capture devices are arranged at different signal source ends and decoder ends, the abnormality detection cycle is adjusted to 100 milliseconds, and then all distributed capture devices are started at the same time, and the decoder media stream is continuously switched to reproduce the phenomenon. Once the phenomenon occurs, the distributed capture device must have captured the abnormality, and all distributed capture devices are triggered to stop at the same time, thereby providing data analysis for the distributed network data joint analysis platform. Finally, the instance is located as one of the first-level switches, when the data is switched, the routing table entries will occasionally have a delay, thus causing this phenomenon; by upgrading the firmware of the switch at this level, the problem is solved.

[0087] As described above, the distributed network data joint analysis system supporting instantaneous anomaly detection of services provided in the embodiments can support joint analysis and diagnosis of link data of complex distributed systems with multi-level basic network switching equipment. The detection intermediate process can cover all relay processes of service data. In particular, the system supports the detection of instantaneous anomalies of high-bandwidth service data, and the detection parameters can be adjusted and configured according to the service characteristics, with stronger detection capabilities and flexibility.

[0088] Obviously, those skilled in the art will intentionally make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A method for detecting transient service anomalies using a distributed network data joint analysis system, characterized in that: The distributed network data joint analysis system includes a distributed network data capture device, a distributed network data joint analysis platform and a distributed network data joint analysis client; the distributed network data capture device, the distributed network data joint analysis platform, the distributed network data joint analysis client and the basic network switching device are connected through a standard network link interface; wherein: The distributed network data joint analysis client provides a user human-computer interaction interface, receives analysis parameters input by the user in the interface, and displays the data capture process status and joint analysis results to the user; The distributed network data joint analysis platform receives analysis parameters sent by the distributed network data joint analysis client, converts the analysis parameters into interface instructions including capture detection trigger conditions and publishes them to each distributed network data capture device, monitors the status of each distributed network data capture device, uploads the captured data of each distributed network data capture device, performs joint analysis, and notifies the distributed network data joint analysis client of the joint analysis results; There are several distributed network data capture devices, which are distributed in different locations of the network, and the number of configurations is not less than the number of levels of basic network switching equipment. According to the configuration parameters and instructions of the distributed network data joint analysis platform, the business data of a certain level of basic network switching equipment is captured; The specific process is as follows: (1) According to the number of levels of network switching equipment, a corresponding number of distributed network data capture devices are configured. Assuming that the number of levels of basic network equipment is N, and N is a positive integer, the number of distributed network data capture devices configured is at least N; (2) connecting the network interface of the distributed network data capture device to the corresponding basic network device, and configuring the port mirroring associated with the current level of business data to the network interface of the distributed network data capture device at this level; (3) Set the capture trigger conditions according to the service characteristics, set the minimum detection period T of the service data, and set the abnormal data bandwidth range [Sa, Sb] of the service; (4) The distributed network data joint analysis platform sends the capture trigger conditions to the distributed network data capture devices at all levels and configures the trigger delay Td; (5) The distributed network data joint analysis platform sends a capture start instruction to distributed network data capture devices at all levels to start the trigger capture function; (6) When any distributed network data capture device meets the trigger condition, it stops capturing after a delay of Td and notifies the distributed network data joint analysis platform that the capture is successful; (7) After the distributed network data joint analysis platform receives the information that any distributed network data capture device has successfully captured, it immediately stops the capture work of all other distributed network data capture devices and pushes the capture success information to the distributed network data joint analysis client; (8) After the distributed network data joint analysis client obtains the capture success information, it informs the user; if the user wants to analyze this capture, the data analysis command is started, and the distributed network data joint analysis platform uploads the data captured by the distributed network data capture devices at all levels and analyzes it; after the analysis is completed, the results are pushed to the user, and the user can quickly locate the cause of the abnormal problem based on the analysis results.

2. The method for detecting instantaneous service anomalies using a distributed network data joint analysis system according to claim 1, characterized in that: The distributed network data capture device includes a network interface module, a parameter configuration module, a time synchronization module, a data packet filtering module, a capture detection module, a capture data writing module, a data storage module, a capture data reading module, and a capture data uploading module; wherein: The network interface module is directly connected to the basic network switching device. The basic network device copies the port carrying the service data to this module by mirroring to obtain all data packets of the service port. The parameter configuration module obtains the parameters issued by the distributed network data joint analysis platform, including data packet filtering conditions, capture detection trigger conditions and time synchronization information; The time synchronization module receives the time synchronization information and sets the local time of the distributed network data capture device to keep it consistent with the time of the distributed network data joint analysis platform; The data packet filtering module filters the data sent by the network interface module according to the received data packet filtering conditions and sends it to the capture detection module; The capture detection module generates a detection trigger condition according to the received capture detection trigger condition, and once the detection trigger condition is met, after a certain period of time, notifies the capture data writing module to stop writing to the data storage module; The capture data writing module receives the write command sent by the capture detection module, and then continuously writes data into the data storage module; when receiving the stop write command sent by the capture detection module, the write operation is stopped; The data storage module receives the data writing instruction from the captured data writing module and saves the data; at the same time, it receives the reading instruction from the captured data reading module and retrieves the corresponding data; The captured data reading module receives the read-back command from the captured data uploading module, then takes out the business data and sends it to the captured data uploading module; The capture data upload module receives the capture readback instruction issued by the distributed network data joint analysis platform, notifies the capture data reading module to read the business data, and then frames the received business data into data packets and sends them back to the distributed network data joint analysis platform.

3. The method for detecting instantaneous service anomalies using a distributed network data joint analysis system according to claim 1, characterized in that: The analysis parameters set in the distributed network data joint analysis client include: business data source IP address, MAC address, destination IP address, MAC address, business anomaly detection cycle, and business anomaly bandwidth.

4. The method for detecting transient service anomalies using a distributed network data joint analysis system according to claim 1, characterized in that: The basic network switching device is a switch.

5. The method for detecting instantaneous service anomalies using a distributed network data joint analysis system according to claim 1, characterized in that: The network interface modules in the distributed network data capture device include Gigabit Ethernet interface, 10 Gigabit Ethernet interface and 100G fiber optic Ethernet interface.