t|||||||||||||ll||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
`
`US 20020161883A1
`
`(19) United States
`(12) Patent Application Publication (10) Pub. No.: US 2002/0161883 A1
`Matheny et al.
`(43) Pub. Date:
`Oct. 31, 2002
`
`(54) SYSTEM AND METHOD FOR COLLECTING,
`AGGREGATING, AND COALESCING
`NETWORK DISCOVERY DATA
`
`(22) Filed;
`
`Apr, 30, 2001
`
`Publication Classi?cation
`
`(76) Inventors; David Matheny, South Jordan, UT
`(US); Dale Taylor, Lindon, UT (US);
`Richard R. Winterton, Provo, UT
`(US)
`
`Correspondence Address;
`FISH & RICHARDSON, PC
`4350 LA JOLLA VILLAGE DRIVE
`SUITE 500
`SAN DIEGO, CA 92122 (US)
`
`(21) Appl. No.:
`
`09/846,521
`
`(51) Int. Cl.7 ................................................. .. G06F 15/173
`(52) US. Cl. .......................................... .. 709/224; 709/223
`
`(57)
`
`ABSTRACT
`
`In an embodiment, a network management system is capable
`of performing a coordinated discovery operation on devices
`in a network using a number of discovery and aggregator
`agents. Data collected and generated by the agents may be
`coalesced into a database ?le. New agents may be added to
`the system as plug-ins by registering an XML ?le with a
`network manager.
`
`[
`
`112
`
`106
`Discovery J
`Agent
`
`104
`f
`
`I
`A
`Discovery v
`
`t
`v
`
`Network
`Manager
`
`V
`
`Aggregator
`K Agent
`108
`
`A
`
`f- 120
`
`V
`
`V
`
`Network
`Device
`
`\110
`
`V
`
`V
`
`Network
`Device
`
`\110
`
`Network
`Device
`
`\110
`
`001
`
`ServiceNow, Inc.'s Exhibit 1003
`
`

`

`Patent Application Publication Oct. 31, 2002 Sheet 1 0f 5
`
`US 2002/0161883 A1
`
`100 f
`
`102 \
`
`1 12
`
`A
`Discovery v
`
`V
`
`Aggregator
`K Agent
`108
`
`106
`Discovery J
`Agent
`
`K 104
`
`>
`
`Network
`Manager
`
`A
`
`/- 120
`
`I
`
`V
`
`Network
`Device
`
`\110
`
`V
`
`Network
`Device
`
`K110
`
`if
`
`Network
`Device
`
`K110
`
`V
`
`Network
`Device
`
`K110
`
`FIG. 1
`
`002
`
`ServiceNow, Inc.'s Exhibit 1003
`
`

`

`Patent Application Publication Oct. 31, 2002 Sheet 2 0f 5
`
`US 2002/0161883 A1
`
`200 N
`
`i
`
`202 \ Store XML file
`describing agent
`in Agent Directory
`
`t
`204 \ Request
`Discovery
`l
`
`206
`\ Generate matrix
`of capabilities
`
`208
`
`Plug-in new
`agent?
`
`FIG. 2
`
`003
`
`ServiceNow, Inc.'s Exhibit 1003
`
`

`

`Patent Application Publication
`
`Oct. 31, 2002 Sheet 3 0f 5
`
`US 2002/0161883 A1
`
`300 N
`
`Receive discovery request
`l
`Identify command file containing
`address ranges or subnet for
`discovery
`
`Call registered discovery agent
`with filename of command file
`iv
`Create discovery directory
`
`306 /
`
`/ 308
`
`.
`
`7
`Collect dlscovery data
`l
`Create discovery file for network A 312
`device
`
`r‘ 310
`
`314
`
`Last device identified in
`command file?
`
`no
`
`Perform aggregation operation J 400
`it
`Perform coalescing operation
`
`500
`
`Copy discovery document into
`database
`l
`Remove created directory and
`command file
`
`Last discovery agent called?
`
`FIG. 3
`
`004
`
`ServiceNow, Inc.'s Exhibit 1003
`
`

`

`Patent Application Publication Oct. 31, 2002 Sheet 4 0f 5
`
`US 2002/0161883 A1
`
`400 x
`
`(
`
`Begin
`
`)
`
`402 \
`
`7
`Identify Aggregator Agent
`
`404 —\
`
`l
`
`Ful?ll any appropriate dependency
`requirements
`
`i
`_
`406 \
`Call identified Aggregator Agent
`with full path of discovery directory
`
`l
`Aggregate data
`
`i
`
`Coalesce data
`
`408 —\
`
`0 50x
`
`410
`
`Last Aggregator Agent?
`
`FIG. 4
`
`005
`
`ServiceNow, Inc.'s Exhibit 1003
`
`

`

`Patent Application Publication
`
`Oct. 31, 2002 Sheet 5 0f 5
`
`US 2002/0161883 A1
`
`500 x
`
`(
`
`Begin
`
`)
`
`502 \ Create discovery document
`
`V
`
`A
`
`504 \ Copy data for discovered machine
`
`V
`
`form generated discovery files into
`discovery document
`
`7
`Compare new data for discovered
`machine to data in discovery
`document
`
`508 \
`
`‘'
`Eliminate duplicate data
`
`510
`
`_\ Create valid key for discovered
`machine
`
`.
`
`.
`
`Last discovery file?
`
`FIG. 5
`
`006
`
`ServiceNow, Inc.'s Exhibit 1003
`
`

`

`US 2002/0161883 A1
`
`Oct. 31, 2002
`
`SYSTEM AND METHOD FOR COLLECTING,
`AGGREGATING, AND COALESCING NETWORK
`DISCOVERY DATA
`
`BACKGROUND
`
`[0001] Managers of computer networks may use a variety
`of network management tools to manage a network. These
`tools may be used to determine the operational status of
`equipment and transmission facilities and to obtain noti?
`cation of faults and threshold conditions, e.g., network traf?c
`bottlenecks. Such tools enable network managers to better
`manage large and complex networks and facilitate con?gu
`ration changes.
`[0002] One type of network management tool is the dis
`covery agent. These agents gather information from devices
`in the network. This information may be used to evaluate
`network performance and possible faults, as well provide
`information needed to recon?gure the network. A network
`manager may wish to incorporate new discovery agents into
`its eXisting set as they become available or as the network
`eXpands and evolves. Also, as new network devices are
`created and incorporated into the network, new ways to
`discover the new devices may be necessary. However,
`incorporating the new agents into the network management
`system may require recon?guration and modi?cation of the
`new agent and/or the system. Furthermore, the new agents
`may gather information already provided by eXisting agents,
`thereby generating redundant data that adds to the complex
`ity of processing the discovery information.
`
`BRIEF DESCRIPTION OF THE DRAWINGS
`
`[0003] FIG. 1 is a block diagram of a network manage
`ment system according to an embodiment.
`
`[0004] FIG. 2 is a ?owchart illustrating a registration
`operation according to an embodiment.
`
`[0005] FIG. 3 is a ?owchart illustrating a discovery opera
`tion according to an embodiment.
`
`[0006] FIG. 4 is a ?owchart illustrating an aggregation
`operation according to an embodiment.
`
`[0007] FIG. 5 is a ?owchart illustrating a coalescing
`operation according to an embodiment.
`
`DETAILED DESCRIPTION
`
`[0008] Anetworked computer system (“network”) 100, as
`shown in FIG. 1, includes a network management system
`102 capable of performing a coordinated network discovery
`operation using a number of discovery agents according to
`an embodiment. The network management system 102
`includes a network manager 104, which utiliZes a number of
`different network discovery agents 106 and aggregator
`agents 108 to perform discovery operations. The network
`management system 102 accommodates the incorporation of
`new discovery agents and aggregator agents into the system
`over time.
`
`[0009] The network 100 includes a variety of intercon
`nected network devices 110, which may include, for
`eXample, network servers, as well as workstations, ?le
`servers, web servers, hubs, routers, etc., and software objects
`on such machines. The network 100 may be organiZed as a
`Local Area Network (LAN), Wide Area Network (WAN),
`
`intranet, etc., and may include a number of smaller subnets.
`The network may operate over a variety of communications
`media 120, e.g., twisted pair cable, coaXial cable, ?ber optic
`cable, wireless transceiver links, etc., and communication
`protocols, e.g., Transmission Control Protocol/Internet Pro
`tocol (TCP/IP), Ethernet, Synchronous Optical Network
`(SONET), etc.
`
`[0010] The network manager 104 may be a host computer
`that includes software for initiating and coordinating net
`work discovery operations on devices in the network using
`a number of different agents. More than one network man
`ager may be allocated to the network 100.
`
`[0011] The discovery agents 106 collect information from
`targeted network devices 110 during a discovery operation
`(described below), for eXample, by polling devices in a
`certain range of addresses or in a particular subnet. Different
`discovery agents may perform discovery operations using
`different techniques, and may collect different types of data.
`Discovery agents may be on the same computer as the
`discovery manager, or may reside on a remote machine that
`uses a local module to communicate with the discovery
`manager. The discovery agents used in the system may use
`a number of different management protocols including, for
`eXample, Simple Network Management Protocol (SNMP),
`Remote Monitoring (RMON), Internet Control Message
`Protocol (ICMP), and other eXisting standardiZed and pro
`priety management protocols, as well as management pro
`tocols that may become available in the future. Compliant
`devices may have a corresponding agent. For eXample,
`SNMP-compliant device include an SNMP agent.
`
`[0012] SNMP (documented in the Institute of Electrical
`and Electronics Engineers, Inc., (IEEE) Request for Com
`ments (RFC) 1098, April 1989) is an application-level
`protocol that is part of the TCP/IP protocol suite. An SNMP
`discovery agent sends messages, called protocol data units
`(PDUs), to different parts of the network. SNMP-complaint
`devices store data about themselves in Management Infor
`mation Bases (MIBs) and return this data to the SNMP
`requester. The MIBs may include low-level attributes (vari
`ables) of the con?guration or operation of the managed
`device.
`
`[0013] RMON (documented in IEEE RFC 1757, February
`1995) is another management protocol. Unlike SNMP,
`RMON de?nes a number of different MIB types, and
`therefore may return a more detailed set of data than an
`SNMP discovery agent.
`
`[0014] ICMP (documented in IEEE RFC 792, September
`1981), an extension of the Internet Protocol (IP), supports
`packets containing error, control, and information message.
`The Packet Internet Groper (PING) is an ICMP utility, which
`is used to test an Internet connection. PING tests the
`connection by sending a single data packet and listening for
`a single packet in reply from an ICMP-compliant device.
`
`[0015] Aggregator agents 108 process the data collected
`by the discovery agents to generate additional information
`that may be useful to the network manager 104. This
`additional information may include higher-level attributes
`about a device that are computed from the data collected by
`a discovery agent, e.g., MIB variables. Aggregator agents
`may have certain dependencies that must be ful?lled before
`they are called in order to generate useful information. For
`
`007
`
`ServiceNow, Inc.'s Exhibit 1003
`
`

`

`US 2002/0161883 A1
`
`Oct. 31, 2002
`
`example, some aggregator agents may only operate on
`certain variables. Also, some aggregator agents may only
`operate on data collected by a particular discovery agent or
`set of agents.
`
`generated as a valid XML ?le With a tag line folloWed by the
`tag <node>. The terminating </node> may be omitted in
`order to enable aggregator agents, called later in the opera
`tion 300, to append to the ?le.
`
`[0016] In an embodiment, the Extensible Markup Lan
`guage (XML) may be used as the communication language
`betWeen agents and the discovery manager during the dis
`covery operation. When the different components commu
`nicate, they do so by passing XML ?les to each other.
`Although XML is described, other standard markup lan
`guages that accommodate customiZed tags for different
`attributes may be used, for example, the Standard General
`iZed Markup Language (SGML).
`
`[0017] FIG. 2 is a ?oWchart that describes a registration
`operation 200 according to an embodiment. An agent reg
`isters With the discovery manager by placing an XML ?le in
`an agent directory in the discovery database 112 (block 202).
`The registration ?le includes tags describing the attributes of
`the agent, calls that the agent supports, and any requirements
`the agent has, including dependencies. Each time a discov
`ery operation 300 (described beloW) is requested, the net
`Work manager 104 reads through the agent directory and
`generates a matrix of available capabilities and existing
`dependencies (block 206) by parsing the registration ?les of
`the various agents.
`
`[0018] Agents may be registered during installation or
`during an upgrade. The netWork system may be upgraded by
`adding neW agents as “plug-ins” (block 208). The registra
`tion operation for the plug-in agents may be performed on
`the ?y, Without modifying the existing agents or discovery
`methodology. As plug-ins, the agents may be called as
`executables or loadable modules.
`
`[0019] FIG. 3 is a ?oWchart that describes a discovery
`operation 300 according to an embodiment. The discovery
`operation 300 is initiated by the netWork manager 104
`receiving a discovery request (block 302). The discovery
`request may include requested data types and designate an
`address range(s) or subnet(s) for discovery. The discovery
`request may be compared to the available capabilities
`de?ned by the matrix derived from the registration ?les in
`the agent directory. The netWork manager 104 loops through
`?les in a command directory, searching for XML ?les that
`match the address ranges or subnets identi?ed for discovery
`(block 304). These ?les may include a high-level tag named
`<task> for easy recognition. The netWork manager may then
`create a command ?le for each identi?ed discovery agent.
`
`[0020] As each ?le is found, a registered discovery agent
`may be called With the full path and ?lename of the ?le as
`a parameter (block 306). When the discovery agent begins to
`run, it may create a discovery directory in the same folder
`that contains the command ?le using a request identi?er in
`the command ?le (block 308). For example, if the command
`?le is “/somepath/command/cmd.xml” With a <requestid>
`tag With the value “1234567”, the discovery agent creates
`the directory “/somepath/command/1234567”.
`
`[0021] The discovery agent 106 collects data from each
`device that it discovers (block 310), and places the collected
`data in a ?le created for that device in the discovery
`directory (block 312). The discovery agent may name the
`?les in such a Way as to avoid duplicates, for example, by
`using the device’s IP address as a ?lename. Each ?le may be
`
`[0022] The discovery agent may also create a relationship
`?le, Which indicates hoW the discovered nodes relate to each
`other. This ?le may be named using the request identi?er
`from the original command ?le. In the example above, a
`relationship ?le may have the name “/somepath/command/
`1234567/1234567xml”.
`[0023] When the discovery agent completes collecting
`data from all available devices in the designated address
`range or subnet (block 314), an aggregator operation 400
`according to an embodiment is performed. As shoWn in
`FIG. 4, the netWork manager 104 identi?es a registered
`aggregator agent (block 402) and ful?lls any dependencies
`(block 404) before calling the aggregator agent. The aggre
`gator agents may be called With the full path of the discovery
`directory created by the discovery agent (block 406) and
`aggregate the data in the ?les (block 408).
`[0024] FIG. 5 is a ?oWchart describing a coalescing
`operation 500 according to an embodiment. After the ?rst
`aggregator agent completes, the netWork manager 104 may
`create a discovery document and begin to coalesce the
`discovered and aggregated information (block 502). The
`discovery document may be created With a top-level node
`named, for example, <i-discover>. The netWork manager
`104 loops through the information in all of the discovery
`?les for the discovered devices generated and appended by
`the various discovery and aggregator agents and copies the
`data into the discovery document (block 504).
`[0025] In the process of copying the data, neW data may be
`compared to previously copied data (block 506) and any
`duplicate data may be eliminated (408). The same device
`may be discovered multiple times, by different discovery
`agents. As described above, ?les corresponding to a particu
`lar device may have the same name, e.g., the IP address.
`Identical information for the same device, as identi?ed by
`the ?le name, may be eliminated.
`
`[0026] The netWork manager 104 may create a valid key
`for each discovered device (block 510). The netWork man
`ager may generate the key by parsing a special precedence
`?le that contains speci?c instructions for de?ning a valid
`node for generating the key. After data from all of the
`discovery ?les has been coalesced, this precedence ?le may
`also be appended to the discovery document.
`
`[0027] After the ?rst aggregator agent 108 completes, the
`netWork manager 104 loops through the other aggregator
`agents, repeating the aggregator operation until all registered
`aggregator agents have been called.
`
`[0028] After the aggregator operation 400 has been per
`formed on all ?les in the discovery directory created by the
`last-called discovery agent 106, the discovery document
`may be copied to the discovery database (block 316), and the
`discovery directory and command ?les generated for the
`discovery operation 300 removed (block 318). The netWork
`manager 104 continues to loop through the other discovery
`agents 106 (block 320), repeating the operation (blocks
`306-318) until all registered discovery agents have been
`called and all discovered and aggregated information has
`been coalesced into the discovery document.
`
`008
`
`ServiceNow, Inc.'s Exhibit 1003
`
`

`

`US 2002/0161883 A1
`
`Oct. 31, 2002
`
`[0029] In an embodiment, discovery agents 106 may be
`given different priorities based on the accuracy of the data
`they collect. For example, SNMP may be assigned a higher
`priority than PING. SNMP and PING may both return the
`same data from a discovered device, e.g., the hostname. In
`the event of a discrepancy during the coalescing operation,
`the data collected by the agent With the higher priority
`(SNMP) Would be copied into the discovery document, and
`the data collected by the loWer priority agent (PING) dis
`carded.
`
`[0030] A number of embodiments have been described.
`Nevertheless, it Will be understood that various modi?ca
`tions may be made Without departing from the spirit and
`scope of the invention. For example, steps of the various
`operations may be performed in a different order and still
`achieve desirable results. Accordingly, other embodiments
`are Within the scope of the folloWing claims.
`
`1. A method comprising:
`
`receiving discovery data collected from a netWork device
`by tWo or more discovery agents;
`aggregating said discovery data;
`coalescing the discovery data in a discovery document,
`said discovery data including tWo or more duplicate
`data entries; and
`
`removing all but one of the duplicate data entries from the
`discovery document.
`2. The method of claim 1, further comprising registering
`agents in an agent directory.
`3. The method of claim 2, Wherein said receiving discov
`ery data comprises calling said tWo or more discovery agents
`from the agent directory.
`4. The method of claim 2, Wherein said aggregating said
`data comprises calling tWo or more aggregator agents listed
`in an agent directory.
`5. The method of claim 2, Wherein the agent directory
`comprises a plurality of Extensible Markup Language
`(XML) ?les.
`6. The method of claim 1, Wherein said removing all but
`one of the duplicate data entries comprises:
`
`identifying tWo or more agents responsible for generating
`the tWo or more duplicate data entries, each agent
`having a priority value;
`
`comparing the priority values of the tWo or more agents;
`
`identifying a ?rst agent having a highest priority, said ?rst
`agent responsible for generating a ?rst duplicate data
`entry in the tWo or more duplicate data entries; and
`
`removing all but the ?rst duplicate date entry.
`7. The method of claim 6, Wherein the tWo or more agents
`are discovery agents.
`8. The method of claim 6, Wherein the tWo or more agents
`are aggregator agents.
`9. The method of claim 1, Wherein the discovery docu
`ment is an Extensible Markup Language (XML) ?le.
`10. The method of claim 1, Wherein said receiving dis
`covery data comprises receiving discovery data collected
`from tWo or more netWork device by said tWo or more
`discovery agents.
`
`11. The method of claim 10, further comprising:
`
`storing the discovery document in a discovery database;
`and
`
`generating a key for each discovered netWork device in
`the discovery document.
`12. A method comprising:
`
`registering a ?rst plurality of agents;
`performing a ?rst discovery operation including
`
`collecting and aggregating data from a ?rst plurality of
`netWork devices With said ?rst plurality of agents,
`and
`
`coalescing said data in a discovery document;
`
`registering an additional agent; and
`
`performing a second discovery operation With a second
`plurality of agents including said ?rst plurality of
`agents and said additional agent.
`13. The method of claim 12, Wherein said ?rst plurality of
`agents include a plurality of discovery agents and a plurality
`of aggregator agents.
`14. The method of claim 12, Wherein said registering the
`additional agent comprises storing an agent ?le describing
`attributes of the additional agent in an agent directory.
`15. The method of claim 14, Wherein the agent ?le
`comprises an Extensible Markup Language (XML) ?le.
`16. An article, including instructions residing on a
`machine-readable medium, the instructions causing a
`machine to:
`
`receive discovery data collected from a netWork device by
`tWo or more discovery agents;
`aggregate said discovery data;
`coalesce the discovery data in a discovery document, said
`discovery data including tWo or more duplicate data
`entries; and
`
`remove all but one of the duplicate data entries from the
`discovery document.
`17. The article of claim 16, further comprising instruc
`tions that cause the machine to register agents in an agent
`directory.
`18. The article of claim 17, Wherein the instructions that
`cause the machine to receive discovery data comprise
`instructions that cause the machine to calling said tWo or
`more discovery agents from the agent directory.
`19. The article of claim 17, Wherein the instructions that
`cause the machine to aggregate said data comprise instruc
`tions that cause the machine to call tWo or more aggregator
`agents listed in an agent directory.
`20. The article of claim 17, Wherein the agent directory
`comprises a plurality of Extensible Markup Language
`(XML) ?les.
`21. The article of claim 16, Wherein the instructions that
`cause the machine to remove the second duplicate data entry
`comprise instructions that cause the machine to:
`
`identify tWo or more agents responsible for generating the
`tWo or more duplicate data entries, each agent having
`a priority value;
`
`compare the priority values of the tWo or more agents;
`
`009
`
`ServiceNow, Inc.'s Exhibit 1003
`
`

`

`US 2002/0161883 A1
`
`Oct. 31, 2002
`
`identify a ?rst agent having a highest priority, said ?rst
`agent responsible for generating a ?rst duplicate data
`entry in the tWo or more duplicate data entries; and
`
`remove all but the ?rst duplicate date entry.
`22. The article of claim 21, Wherein the tWo or more
`agents are discovery agents.
`23. The article of claim 21, Wherein the tWo or more
`agents are aggregator agents.
`24. The article of claim 16, Wherein the discovery docu
`ment is an Extensible Markup Language (XML) ?le.
`25. The article of claim 16, Wherein the instructions that
`cause the machine to receive discovery data comprise
`instructions that cause the machine to receive discovery data
`collected from tWo or more netWork device by said tWo or
`more discovery agents.
`26. The article of claim 25, further comprising instruc
`tions that cause the machine to:
`
`store the discovery document in a discovery database; and
`
`generate a key for each discovered netWork device in the
`discovery document.
`27. An article, including instructions residing on a
`machine-readable medium, the instructions causing a
`machine to:
`
`register a ?rst plurality of agents;
`perform a ?rst discovery operation including
`collect and aggregating data from a ?rst plurality of
`netWork devices With said ?rst plurality of agents,
`and
`
`coalesce said data in a discovery document;
`
`register an additional agent; and
`
`perform a second discovery operation With a second
`plurality of agents including said ?rst plurality of
`agents and said additional agent.
`28. The article of claim 27, Wherein said ?rst plurality of
`agents include a plurality of discovery agents and a plurality
`of aggregator agents.
`29. The article of claim 28, Wherein the instructions that
`cause the machine to register the additional agent comprise
`instructions that cause the machine to store an agent ?le
`describing attributes of the additional agent in an agent
`directory.
`30. The article of claim 29, Wherein the agent ?le com
`prises an Extensible Markup Language (XML) ?le.
`
`*
`
`*
`
`*
`
`*
`
`*
`
`010
`
`ServiceNow, Inc.'s Exhibit 1003
`
`