Patent Grant
10063441
The present invention relates to a distributed computing system and in particular, it relates to calculating latency in the distributed computing system.
In the past few years, grid computing has shown a tremendous growth in the computer industry. Grid computing is a form of a distributed computing system in which numerous users share computing resources. The grid computing expedites selection, aggregation, and sharing of the computing resources across multiple hops (or nodes) over a diverse geographic area and across multiple networks. The sharing may be based on availability, capability, and cost of the computing resources available in the grid as well as on a user's quality of service requirements and the user's functional demands. Grid computing is growing in popularity to address the growing demands of processing power and capability.
In order to meet the growing demand of processing and a high performance in the distributed computing system, it is increasingly becoming critical to minimize latency between the multiple nodes. Latency in the network increases the possibilities of having bottlenecks at different hops in the grid due to factors, external to the system like network timeouts, CPU/memory bottlenecks and so forth. Latency also increases the packet loss in the network. Therefore, minimizing latency in the distributed system can result in aggregated and improved efficiency of computing, data, and storage resources.
In common approach, systems try to minimize latency in a packet-to-packet transmission. Such approach is expensive in terms of both space and time. In a conventional approach, systems attempt to minimize latency by synchronizing data packets only on source and destination clocks. Accordingly, these systems do not provide the functionality necessary to optimize and minimize system latency between remotely operating computing devices. In another conventional approach, a data stream is modified to include a time stamp synchronized to a common clock. This method is also intrusive and can delay distribution of the data stream.
In light of the above discussion, there is a need for a method and system, which overcomes all the above stated problems.
The above-mentioned shortcomings, disadvantages and problems are addressed herein which will be understood by reading and understanding the following specification.
In embodiments, the present invention provides a system for calculating latency in a distributed data flow environment. The system includes a plurality of hop servers and a central registry server. The plurality of hops servers is configured to receive a plurality of data packets across a time window, bucket each of the data packets of the plurality of data packets received across the time window into one or more buckets and transmit a count of the plurality of the data packets in the one or more buckets and a window identifier associated with the time window to the central registry server. Each data packet of the plurality of data packets includes a header. The central registry server is configured to receive a plurality of counts and a plurality of window identifiers, to receive a latency query and to calculate latency based on the plurality of counts and the plurality of window identifiers.
In an embodiment, the plurality of hop servers includes one or more data agents. Each data agent of the one or more data agents is configured to parse the header of the each data packet of the plurality of data packets.
In an embodiment, the header of each data packet of the plurality of data packets includes a timestamp at which the each data packet of the plurality of data packets originated.
In another aspect, the present invention provides a method for transmitting data packet information to calculate latency over a distributed data flow environment. The method includes receiving a plurality of data packets across a time window, bucketing each of the data packets of the plurality of data packets received across the time window into one or more buckets each of the data packet of the plurality of data packets, and transmitting a count of the plurality of the data packets in the one or more buckets and one or more identifiers of the time window to a central registry server. Each of the data packets of the plurality of data packets includes a header.
In an embodiment, the header of each data packet of the plurality of data packets includes a timestamp at which each data packet of the plurality of data packets originated.
In yet another aspect, the present invention provides a method of calculating latency in a distributed data flow environment. The method includes receiving a plurality of counts and one or more identifiers of a time window from a plurality of hop servers, receiving a latency query, and calculating latency. The plurality of counts indicates a plurality of data packets in one or more buckets. Each of the one or buckets comprises a plurality of data packets grouped together based on a header of each of the data packets from the plurality of the data packets. Latency is calculated based on the one or more buckets and the one or more identifiers of the time window.
In an embodiment, the header of each data packet of the plurality of data packets includes a timestamp at which the each data packet of the plurality of data packets originated.
In the following detailed description, reference is made to the accompanying drawings that form a part hereof, and in which is shown by way of illustration specific embodiments, which may be practiced. These embodiments are described in sufficient detail to enable those skilled in the art to practice the embodiments, and it is to be understood that other embodiments may be utilized and that logical, mechanical, electrical and other changes may be made without departing from the scope of the embodiments. The following detailed description is, therefore, not to be taken in a limiting sense.
The distributed computing system 100 includes a plurality of application servers 130. The plurality of the application servers 130 includes an application server 132, an application server 134 and an application server 136, as shown in
In an embodiment, the application server 132 of the plurality of the application servers 130 is an advertisement server. The advertisement server is a web server used in online marketing to deliver advertisements (referred to as “ads” from herein) to a user. The advertisement server generates ads and distributes the ads using multiple data streams to multiple locations. Additionally, the advertisement server performs tasks such as counting a number of impressions for the ads, etc. The number of impressions for the ads is calculated based on the number of clicks that the ads receive.
Functions and capabilities of the application server 132 are same as the functions and the capabilities of an application server 134 and an application server 136.
In a context of the present invention, a plurality of event generators includes a mobile device 112, a mobile device 114, and a mobile device 116. The mobile device 112 communicates with the application server 132 in the distributed computing system 100. A plurality of the data packets generated on the mobile device 112 are transmitted from the mobile device 112 to the application server 132. Examples of the mobile device 112 include but may not be limited to, a cell phone, a smart phone, a personal digital assistant (PDA), a wireless email terminal, a laptop, a desktop computer, and a tablet computer.
In a context of the present invention, the plurality of the data packets received by the application server 132 is transmitted to other application servers of the plurality of the application servers 130 in a pipeline. The pipeline is a set of data processing units connected in series (or parallel), where an output of one unit is an input of a next unit. In a context of the present invention, the plurality of the application servers 130 is in one or more pipelines. Each of the plurality of the application servers 130 processes the plurality of the data packets received from a previous application server of the plurality of the application servers 130.
In a context of the present invention, the application server 132 includes a data agent. The data agent collects one or more data packets of the plurality of the data packets coming from the various nodes of the distributed computing system 100. The data agent processes the collected data and transmits the collected plurality of the data packets to the application server 134. Similarly, a data agent of the application server 134 collects the plurality of the data packets, processes the plurality of the data packets and transmits the plurality of the data packets to the application server 136.
In the embodiment discussed above, the advertisement server receives the impression and a clickstream from the mobile device 112. The clickstream is a record of the user's activity on the Internet. In an embodiment, the clickstream is recorded on a web browser of the mobile device 112. The advertisement server generates reports based on the impression and the clickstream for the ads. The reports determine return on investment (ROI) for an advertiser. A data analysis system is used in conjunction with the clickstream to build business intelligence The examples of the data analysis systems include but are not limited to data mining, online analytical processing (OLAP), multidimensional online analytical processing (MOLAP), relational online analytical processing (ROLAP), and a column-oriented database management system (DBMS).
In the embodiment discussed above, the mobile device 112 receives the ads from the advertisement server. In an embodiment, a user associated with the mobile device 112 views the ads while browsing a website. In another embodiment, the user views the ads on an application installed on the mobile device 112. The user, based on interest, clicks on the ads. The clickstream is a record of user information like user's Internet protocol (IP) address, a time stamp of the click, information of the mobile device 112, website information, application information, and so forth. The advertisement server processes the impression and clickstream. On processing the impression and clickstream, the advertisement server transmits the processed impression and the processed clickstream to a next advertisement server in the pipeline. In a context of the present invention, functions and capabilities of the mobile device 112 are same as the functions and the capabilities of a mobile device 114 and a mobile device 116.
In an embodiment, the plurality of the application servers 130 in the grid is used for video processing. The grid includes a plurality of the video filters. Each of the plurality of the video filters performs a processing task. The application server 132 is a denoising filter. The denoising filter receives raw video frames from the mobile device 112. The denoising filter removes the noise from the raw video frames. On removing the noise from the raw video frames, the denoising filter transmits denoised video frames to a resizing filter. The resizing filter upscales and downscales the denoised video frames. The resizing filter transmits resized video frames to a correcting filter. The correcting filter corrects attributes such as brightness, color, contrast, etc. The correcting filter transmits the corrected video to the mobile device 112. The corrected video is the processed video.
In a context of the present invention, the plurality of application servers 130 is connected to a central registry server 120. In an embodiment, the central registry server 120 includes a database to store information flowing in the pipeline. In a context of the present invention, information related to the plurality of the data packets collected by the data agents are transmitted to the central registry server 120.
In an embodiment, the application server 132 is configured to transmit the information related to the plurality of the data packets in a constant periodicity. The periodicity can be pre-determined according to a need of a Service Level Agreement (SLA).
At step 210, the flowchart 200 initiates. At step 220, the application server 132 receives the plurality of the data packets across a time window. A window identifier identifies the time window. Each of the data packets of the plurality of the data packets includes a header. The header refers to supplement data placed at beginning of a block of the data packets. In an embodiment, the application server 132 includes the data agent. The data agent is coupled to the application server 132 to parse the header. The supplement data includes the time stamp at which each data packet of the plurality of the data packets originated. In an embodiment, the window identifier is the time period between receiving the plurality of the data packets by the application server 132 and the transmitting the plurality of the data packets to the application server 134.
In the embodiment mentioned above, the advertisement server receives the clickstream from the mobile device 112 in the time window. The clickstream includes additional information like the timestamp, user information, etc.
At step 230, the application server 132 buckets data packets of the plurality of data packets based on the header received across the time window. The time stamp of each of the data packets of the plurality of data packets is used to bucket the plurality of the data packets. In an example, the plurality of the data packets originated on the mobile device 132 in a time-period of 1 pm to 2 pm is grouped in a bucket. Similarly, a plurality of data packets originated on the mobile device 132 in a time-period of 2 pm to 3 pm is grouped in another bucket.
The application server 132 maintains a count of the data packets in one or more buckets. At step 240, the application server 132 transmits the count of the data packets in the one or more buckets and the window identifier to the central registry 120. In a context of the present invention, the application server 132 transmits the plurality of the data packets to the application server 134 after processing the plurality of the data packets. Accordingly, the application server 134 buckets data packets of the plurality of the data packets based on the header. On bucketing, the application server 134 transmits a count of the data packets in one or more buckets and a window identifier to the central registry server 120.
In an example, assume the following set of information about the application server 132, application server 134 and the plurality of the data packets.
As per the information provided above, application server 132 receives a data packet D1, a data packet D2, a data packet D3, and a data packet D4. The data packets originate on the mobile device 112. The origin times of the data packet D1 is 1:00 pm, the data packet D2 is 1:30 pm, the data packet D3 is 1:40 pm, and the data packet D4 is 2:00 pm. The application server 132 receives the data packet D1 at 1:10 pm, the data packet D2 at 1:40, the data packet D3 at 1:45 pm, and the data packet D4 at 2:15. The application server 132 buckets the data packets in two buckets, a bucket A and a bucket B, based on the origin time of the data packets. The bucket A includes the data packets originated between 1:00 pm to 1:30 pm. Similarly, the bucket B includes the data packets originated between 1:30 pm to 2:00 pm. On receiving the data packets, the application server 132 processes the data packets and transmits the data packets to the application server 134. The application server 132 transmits the data packet D1 at 1:20 pm, the data packet D2 at 2:00 pm, the data packet D3 at 1:55 pm, and the data packet D4 at 2:30. Additionally, the application server 132 transmits a count of the data packets in the buckets A and B to central registry server 120. As per the information provided, the application server 134 receives the data packet D1 at 1:30 pm, the data packet D2 at 2:10 pm, the data packet at D3 at 2:15 pm, and the data packet 2:45 pm. Accordingly, the application server 134 buckets the data packets into the two buckets, the bucket A and the bucket B, based on the origin time of the data packets. The application server 134 transmits a count of the data packets in the bucket A and the bucket B to the central registry 120.
In an embodiment, application server 132 transmits the count of the data packets in the one or more buckets and the window identifier to the central registry server 120 in a constant periodicity. The central registry server 120 calculates latency based on the received count of the data packets in the one or more buckets and the window identifier. The calculation of latency is further discussed in a description of the
At step 330, the central registry server 120 receives a latency query. In a context of the present invention, latency is calculated for the transmission of the plurality of data packets in the pipeline between the plurality of the application servers 130.
At step 340, the central registry server 120 calculates the latency based on the average delay based on the plurality of the counts received from the plurality of the application servers 130. In an embodiment, central registry server 120 computes the packet loss based on the plurality of the counts received from the plurality of the application servers 130.
As per the above discussed example, assume the following information about the central registry 120, the application server 132, the application server 134, the bucket A and the bucket B. The bucket A includes the data packets originating at 1 pm to 1:30 pm and the bucket B includes the data packets originating in 1:30 pm to 2 pm.
As per the information provided above, the central registry server 120 receives the count of the data packets in the bucket A and the bucket B. Additionally, the central registry server 120 receives a window identifier for a time window 1 pm to 2 pm and a window identifier for a time window 2 pm to 3 pm. The central registry server 120 receives the latency query to calculate latency across the time window 1 pm to 2 pm and the time window 2 pm to 3 pm.
The central registry server 120 receives, from the application server 132, a count of 2 data packets belonging to the bucket A and a count of 1 data packet belonging to the bucket B across the time window 1 pm to 2 pm. Moreover, the central registry server 120 receives, from the application server 134, a count of 1 data packet belonging to the bucket A and a count of 0 data packets belonging to the bucket B. Similarly, the central registry server 120 receives, from the application server 132, a count of 0 data packets belonging to the bucket A and a count of 1 data packet belonging to the bucket B across the time 2 pm to 3 pm. Furthermore, the central registry server 120 receives, from the application server 134, a count of 1 data packet belonging to the bucket A and a count of 2 data packets belonging to the bucket B.
In this example, maximum latency is calculated by weight arithmetic mean technique. The maximum time for a data packet to reach the application server 134 from the application server 132 in the 1 pm to 2 pm time window is 59 minutes for bucket A and 29 minutes for bucket B. In the time window 1 pm to 2 pm, one data packet (D1) belonging to bucket A reaches the application server 134 from the application server 132. The maximum time the 1 data packet D1 can take is 59 minutes. In the time window 2 pm to 3 pm, 1 packet (D2) from the bucket A and two packets from the bucket B reach the application server 134 from the application server 132. These two data packets from bucket B reach the applications server 132 in the time window 2 pm to 3 pm. The maximum time these two packets would have taken to reach Application server 134 is 1 hour 59 minutes. Similarly, 1 packet from the bucket A, moved from the application server 132 to the application server 134 in the time window 2 pm to 3 pm. The maximum time this packet can take is 59 minutes. Accordingly, the weight average maximum latency is calculated as:
It can be understood by one skilled in the art that although not shown, in the above example, latency can be expressed in other unit of time such as, seconds, minutes, hours, days and so forth.
At step 350, the flowchart 300 terminates.
The application server 400 includes a computer server 405, which is operational with numerous other general purpose or special purpose computing system environments or configurations. Examples of well-known computing systems, environments, and/or configurations that may be suitable for use with computer server 405 include, but are not limited to, personal computer systems, server computer systems, thin clients, thick clients, handheld or laptop devices, multiprocessor systems, microprocessor-based systems, set top boxes, programmable consumer electronics, network PCs, minicomputer systems, mainframe computer systems, and the like.
In
The one or more buses represents one or more of any of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, and a processor or local bus using any of a variety of bus architectures. By way of example, and not limitation, such architectures include Industry Standard Architecture (ISA) bus, Micro Channel Architecture (MCA) bus, Enhanced ISA (EISA) bus, Video Electronics Standards Association (VESA) local bus, and Peripheral Component Interconnects (PCI) bus.
Computer server 405 typically includes a variety of computer system readable media. Such media may be any available media that is accessible by computer server 405, and includes both volatile and non-volatile media, removable and non-removable media. In an embodiment, the system memory 455 includes computer system readable media in the form of volatile memory, such as random access memory (RAM) 460 and cache memory 470. Computer server 405 may further include other removable/non-removable, non-volatile computer system storage media. In an embodiment, the system memory 455 includes a storage system 480.
Computer server 405 can communicate with one or more external devices 450 and a display 410, via input-output (I/O) interfaces 440. In addition, computer server 405 can communicate with one or more networks such as a local area network (LAN), a general wide area network (WAN), and/or a public network (for example, the Internet) via the network adapter 420.
It can be understood by one skilled in the art that although not shown, other hardware and/or software components can be used in conjunction with the computer server 405. Examples, include, but are not limited to microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data archival storage systems, etc.
In an embodiment, configuration and capabilities of the application server 132, the application server 134, the application server 136 and the central registry server 120 are same as configuration and capabilities of the application server 400.
As will be appreciated by one skilled in the art, aspects can be embodied as a system, method or computer program product. Accordingly, aspects of the present invention can take the form of an entirely hardware embodiment, an entirely software embodiment (including firmware, resident software, micro-code, etc.) or an embodiment combining software and hardware aspects that may all generally be referred to herein as a “circuit,” “module” or “system.” Furthermore, aspects of the present invention can take the form of a computer program product embodied in one or more computer readable medium(s) having computer readable program code embodied thereon.
Any combination of one or more computer readable medium(s) can be utilized. The computer readable medium can be a computer readable storage medium. A computer readable storage medium can be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of the computer readable storage medium would include the following: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the context of the present invention, a computer readable storage medium can be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.
Computer program code for carrying out operations for aspects of the present invention can be written in any combination of one or more programming languages, including an object oriented programming language and conventional procedural programming languages.
The present invention is useful in monitoring the overall health of distributed system. It helps to detect any bottleneck in the pipeline. Further, the present invention is useful in computing the packet loss in the distributed system. The present invention helps the system engineers in debugging a distributed system by quantifying that which hop of the system has potential issues. Furthermore, the present invention is also very useful to determine whether the SLA requirements are met or not.
In an example, the present invention is useful in reporting metrics like 99% of all packets that were generated in the mobile device, reached the final hop in ‘x’ minutes. Thereby, the present invention helps to fix a 99 percentile latency. The present invention is also useful in reporting a metric stating that, in the last month how many times the SLA (here the 99 percentile latency) was met or not.
This written description uses examples to describe the subject matter herein, including the best mode, and also to enable any person skilled in the art to make and use the subject matter. The patentable scope of the subject matter is defined by the claims, and may include other examples that occur to those skilled in the art. Such other examples are intended to be within the scope of the claims if they have structural elements that do not differ from the literal language of the claims, or if they include equivalent structural elements with insubstantial differences from the literal language of the claims.
| Number | Date | Country | Kind |
|---|---|---|---|
| 105/CHE/2014 | Jan 2014 | IN | national |
| Number | Name | Date | Kind |
|---|---|---|---|
| 7463892 | Eiger et al. | Dec 2008 | B2 |
| 8274905 | Edwards et al. | Sep 2012 | B2 |
| 9477691 | Reiner et al. | Oct 2016 | B1 |
| 20070064715 | Lloyd | Mar 2007 | A1 |
| 20110264822 | Ferguson | Oct 2011 | A1 |
| 20120173662 | Hickey et al. | Jul 2012 | A1 |
| Number | Date | Country | |
|---|---|---|---|
| 20150195170 A1 | Jul 2015 | US |
