Patent Grant
9055092
The present invention generally pertains to the field of monitoring computer networks. Specifically, the present invention pertains to the field of collecting and analyzing data in order to detect intrusions and understand network performance.
One of the key challenges in modern enterprise networks is monitoring network traffic across nodes that simultaneously play multiple diverse roles within the network. Examples of network nodes include an employee's personal computer, an email server, a web application server, a database server, and a file server. The applications running on these nodes use the IP Protocol layer to establish inter-nodal connections for communication and information transfer.
Each IP connection consists of a client (typically the node that initiates the connection) and a server. Networked nodes may concurrently act as both a client and a server, depending on the applications they run. For example, a personal computer node can act as a client, by browsing and downloading web pages, at the same time as it is acting as a server, by sending email attachments. A web application server can act as a client by requesting information from a database server, while it also performs its function as a server in responding to application requests by clients that connect with it. Furthermore, while nodes are acting as both a server and client, they are often members of one or more logical groups.
Traditional network monitoring solutions group network traffic according to whether a network node is a client or a server. Useful metrics such as total throughput for the node disregard whether the node is acting as a server or a client. Other metrics, such as total response time, requests served, and connections established require knowledge of whether the node is acting as a client or a server but are more meaningful when aggregated with other node metrics.
In addition, traditional device-centric methods of measuring network utilization require that device-specific diagnostics or parameters (e.g., the number of packets passing through a particular router) be read and then later aggregated at some central location to determine overall utilization of network resources. These processes are manually intensive and rapidly become unmanageable in the face of network reconfigurations and scaling. While such processes may ultimately provide indications of some parameters, for example bandwidth utilization, they often provide little or no information about the actual end-to-end performance for servers and clients passing traffic across the link of interest. Furthermore, information at the individual node level in a large network comprises information at thousands of nodes and is often too much to process in order to receive a real-time picture of the network. Therefore, a new method of grouping diagnostic information is required.
A method and system for grouping network diagnostic information is described. Client network traffic data and server network traffic data regarding a number of network nodes is collected and then grouped by IP address. The network nodes are divided into logical groupings and the network traffic data is presented in the aggregate for all IP addresses in each logical group. The logical groupings may be further divided by protocol, application, port and/or logical group-to-group. Each logical group can be further generalized as either a set of IP addresses (e.g., a business group) or a specific logical link between one set of IP addresses to another set of IP addresses (e.g., a business group link). Either or both of these “groups” may be divided in further logical sub-groups: for example, by protocol, application, port and in the case of business groups, group-to-group. The logical groups provide facilities for initial problem detection and identification while the logical sub-groups provide facilities for troubleshooting and problem isolation.
The present invention is illustrated by way of example, and not limitation, in the figures of the accompanying drawings in which:
In the following description, examples of logical groupings of network diagnostic information are set forth. By using logical groupings such as those described herein, network operators and others can measure network resource utilization independent of any nodes between information sources and destinations (i.e., the logical groupings permit monitoring of inter-group traffic); to understand the actual end-to-end performance between the sources and destinations. Such groupings also allow the traffic metrics to be stored in a meaningful and scaleable manner and in ways that are more immune to network topology changes (e.g., than traditional monitoring methods). These examples, however, are not meant to be restrictive. Thus, the true measure of the present invention should not be restricted to the examples set forth below, but rather should be consistent with the scope of the claims which follow that discussion.
In this embodiment, a firewall 10 surrounds a geographic collection of networked nodes and separates the internal network from the external network. A network traffic monitoring device 12 is shown at the firewall. However, as will be apparent to one skilled in the art, the network traffic monitoring device 12 may be located within the internal network, or on the external network or anywhere that allows the method of the present invention to be practiced. Note, entity 12 need not be “inline.” That is, traffic need not necessarily pass through entity 12 in order to pass from the server to the client and vice versa. The entity 12 can be a passive monitoring device, e.g., spanning a switch or router, whereby all the traffic is copied to a switch span port which passes traffic to entity 12.
As shown in
In the exemplary embodiment shown here, BG1 contains several internal network nodes N101, N102, N103, and N104 and external nodes N105, N106 and N107. Similarly, BG2 contains several internal network nodes N201, N202, N203, N204, N205, N206. A network node may be any computer or device on the network that communicates with other computers or devices on the network.
Each node may function as a client, server, or both. For example, node N103, is shown as a database which is connected to Node N104, a web application server, via a network link 2. In this configuration, it is typical for node N104 to function as a client of node 103 by requesting database results. However N104 is also depicted as connected to the external network via network link 28. In this configuration, it is typical for N104 to function as a server, which returns results in response to requests from the external network. Similarly, database node N103, which functions as a server to N104, is shown connected to node N107 via a network link 4. N107 may upload information to the database via link 4, whereby N107 is functioning as a server and N103 is functioning as a client. However, N107 is also shown connected to the external network via link 26. This link could indicate that N107 is browsing the Internet and functioning as a client.
Furthermore, network nodes need not be within the internal network in order to belong to a logical group. For example, traveling employees may connect to the logical group network via a Virtual Private Network (VPN) or via ordinary network transport protocols through an external network such as the Internet. As shown in
For vector metrics that have a direction, such as throughput, the method of the present invention allows the information to be presented from the viewpoint of a selected logical group. For example, if BG1 is the selected logical group for the viewpoint of network traffic data, outbound traffic will include traffic on virtual link 20 that is going to the external network and traffic on virtual link 30 that is going to BG2. The second component of the total output for BG1, namely the outbound traffic to BG2, is also a component of inbound traffic for BG2. So, if the selected logical group for the viewpoint of network traffic data is BG2, the total inbound traffic metric will include inbound network traffic over virtual link 40 as well as inbound traffic from BG1 over virtual link 30. In order to avoid storing the same metric twice (traffic from BG1 to BG2), the method of the present invention combines the selected logical group with vector metrics between logical groups in order to present the network traffic data properly. For example, all nodes within BG1 may default to BG1 as the viewpoint of choice when viewing network traffic data according to the method of the present invention. In an alternate embodiment, a user may select any logical group as the reference viewpoint when viewing network traffic data.
In addition to vector metrics, the present invention also accommodates scalar role-specific metrics; that is metrics that are associated with a logical group (as defined by one or more IP addresses), either when the logical group is acting as a server or as a client. For example, application response time can either be associated with the servers of a logical group, or, indirectly, with the clients of a logical group. If it is associated with the clients of a logical group, the metric describes the application response time of all the servers that the clients of the logical group are communicating with. So, the present invention allows information to be presented from the viewpoint of the roles of a selected logical group thereby providing for side-by-side comparisons of both vector and scalar role-specific metrics for a given logical grouping. In other words, this allows both utilization and performance information to be viewed side-by-side for a given logical group. Such capabilities are not found in conventional network monitoring systems.
The left column may present multiple logical groups. For example, logical groups may include designations such as Headquarters 300, the Internet 302, and Employees 304. Each logical group view of the network traffic data may be expanded into a more detailed view (for example, of associated logical sub-groups), which shows local traffic 306 and connected logical groups 308. Furthermore, the logical group view offers several further expanded views, or drilldown views of network traffic data. In one embodiment, network traffic data belonging to individual nodes associated with the group such as connected IP addresses 310 (for nodes outside of the logical group that are connected to the logical group) and member IP addresses 312 is available in an expanded view. In another embodiment, drilldown views of network traffic data by IP protocols 314 (e.g., GRE, TCP, UDP, etc.) are available. In yet another embodiment, further drill down views of network traffic data are available by application type 316, which are then divided into connected IP addresses (nodes external to the logical group) and member IP addresses (nodes internal to the logical group).
These additional drilldowns or logical sub-groupings are especially useful for troubleshooting purposes. For example, once the network diagnostic information is displayed for a given logical group, the next process in troubleshooting a problem is often to isolate any issues to further logical sub-groups. For example, having isolated a problem to a particular sub-group, a network technician might then go on to discover which application in that sub-group is experiencing the problem. Having isolated the application, the technician may then drilldown to determine which employee (i.e., Connected IP in the illustration shown in
The reference viewpoint of the originally selected logical group is passed down to the drill downs as well. For example, in
Each of the columns to the right of the logical group and drill down column include information and/or metrics to help diagnose problems within the network. It is important to note that all metrics available on a time-binned basis are presented as such. In other words, the method of the present invention does not wait until the connection has closed in order to report on network traffic. Examples of TCP-connection based performance metrics that are available for time-binned reporting (e.g., the amount or magnitude observed within the last minute, hour, day, week, and month) include connection set-up time, time to first byte, application response time, connection rate, connection request rate, round trip time, and packet loss.
Thus, the present invention provides a method for grouping network diagnostic information, in which client network traffic data and server network traffic data regarding a plurality of network nodes is collected and grouped by IP address into total network traffic data. Subsequently, the network nodes may be divided into one or more logical groups containing one or more IP addresses; and the total network traffic data for all IP addresses in each logical group presented in the aggregate. In various embodiments, this may include presenting total network traffic throughput for the logical group, presenting total incoming network traffic for the logical group, presenting total outgoing network traffic for the logical group, presenting total network traffic between logical groups, and/or presenting the total network traffic data from the perspective of a selected logical group.
Alternatively or in addition, a node-level view of the total network traffic data may be presented for each logical group. This may include presenting the total network traffic data from the perspective of a selected logical group. Likewise, an application-level view of the total network traffic data for all IP addresses in each logical group may be presented. This may include presenting the total network traffic data from the perspective of a selected logical group.
For various embodiments of the present invention, the total network traffic data may include time-binned statistics, region-independent application response time (e.g., as measured by subtracting a server-side latency from an overall client-server response time as discussed below). Often, the total network traffic data will be presented so as to include vector and scalar role-based metrics, for example in side-by-side fashion. Vector metrics are those having an associated direction, while scalar role-based metrics are associated with a logical group as defined by one or more IP addresses either when the logical group is acting as a server or as a client.
In some embodiments, the total network traffic data is presented by protocol, by port, by application, or by member-connected IP address per logical group. In this latter case, the total network traffic data is presented, for each logical group, by any or all of a logical group member IP address, a logical group connected IP address, a logical group member and then a logical group connected IP address, and a logical group connected and then a logical group member IP address. In still further embodiments, the total network traffic data may be presented by member-connected IP address per logical application sub-group, by member-connected IP address per logical group-to-group connection, by member-connected IP address per logical application group-to-group connection, by member-connected IP address per logical port sub-group, in a role-based fashion for the IP addresses, and/or from a standpoint of a common reference point.
The exemplary computer system 400 includes a processor 402 (e.g., a central processing unit (CPU) a graphics processing unit (GPU) or both), a main memory 404 and a static memory 406, which communicate with each other via a bus 408. The computer system 400 may further include a video display unit 410 (e.g., a liquid crystal display (LCD) or a cathode ray tube (CRT)). The computer system 400 also includes an alphanumeric input device 412 (e.g., a keyboard), a cursor control device 414 (e.g., a mouse), a disk drive unit 416, a signal generation device 418 (e.g., a speaker) and a network interface device 420.
The disk drive unit 416 includes a machine-readable medium 422 on which is stored one or more sets of instructions (e.g., software) 424 embodying any one or more of the methodologies or functions described herein. The software 424 may also reside, completely or at least partially, within the main memory 404 and/or within the processor 402 during execution thereof by the computer system 400, the main memory 404 and the processor 402 also constituting machine-readable media. The software 424 may further be transmitted or received over a network 426 via the network interface device 420.
While the machine-readable medium 422 is shown in an exemplary embodiment to be a single medium, the term “machine-readable medium” should be taken to include a single medium or multiple media (e.g., a centralized or distributed database, and/or associated caches and servers) that store the one or more sets of instructions. The term “machine-readable medium” shall also be taken to include any medium that is capable of storing, encoding or carrying a set of instructions for execution by the machine and that cause the machine to perform any one or more of the methodologies of the present invention. The term “machine-readable medium” shall accordingly be taken to include, but not be limited to, solid-state memories, optical and magnetic media, and carrier wave signals.
Thus, a method and system for grouping network diagnostic information is described. Although the present invention has been described with reference to specific exemplary embodiments, it will be evident that various modifications and changes may be made to these embodiments without departing from the broader spirit and scope of the invention. Accordingly, the true scope of the present invention should be measured only in terms of the claims, which now follow.
| Number | Name | Date | Kind |
|---|---|---|---|
| 5712981 | McKee et al. | Jan 1998 | A |
| 6691165 | Bruck et al. | Feb 2004 | B1 |
| 6801940 | Moran et al. | Oct 2004 | B1 |
| 8717911 | Bugenhagen et al. | May 2014 | B2 |
| 20030229613 | Zargham et al. | Dec 2003 | A1 |
| 20050286434 | McKee et al. | Dec 2005 | A1 |
| Entry |
|---|
| Marik Marshak et al., “Evaluating web user perceived latency using server side measurements” Computer Communications, 26 (2003) 872-887. |
| Mogul et al., ‘Internet standard subnetting procedure’, Network Working Group, Aug. 1985. |
| Number | Date | Country | |
|---|---|---|---|
| 20060069768 A1 | Mar 2006 | US |
