Discovery of network software relationships

Abstract

Embodiments relate to systems and methods for automatic discovery of network software relationships. A set of managed machines, such as personal computers or servers, can be managed by a network management platform communicating with the machines via a secure channel. The network management platform can issue a set of discovery commands interrogating nodes of the network to report software or services which they consume. The network management platform can build a dependency map encoding the dependency or other relationships between nodes in the managed network based on their use of software resources. Network restorations can therefore be conducted in a rational order, maintaining stability in the restored nodes and network.

Description

FIELD

The present teachings relate to systems and methods for automatic discovery of network software relationships, and more particularly to platforms and techniques for the automatic scanning and discovery of software dependency relationships in managed networks.

BACKGROUND OF RELATED ART

Network management platforms exist which permit a systems administrator to connect to a set of targets in a network, and perform maintenance and diagnostic activities on that group of managed machines. The network management platform can identify individual targets via a public key or other security credential, and identify the software provisioning requirements, obtain a hardware inventory, or perform other management tasks on validated machines. In general, existing network management platforms can be hosted on a central server or other facility that permits connection to, identification, and management of the set of targets.

In terms of the maintenance of machines populated on a network as well as the maintenance of the network on a comprehensive basis, systems administrators from time to time may need to take stock of individual machines and their network relationships. The managed network may need to be interrogated to determine installation configurations and network interconnections, for instance, to prepare for the possible restoration of the network or individual machines in the event of a crash, intrusion, disaster, or other unforeseen hazard.

Existing network management platforms suffer from various limitations of design and function, however, in terms of their ability to manage any low-level or high-level restoration process. For one, in the event of the failure of multiple servers, clients, or other nodes or machines, existing platforms are not equipped to identify the software dependencies of one machine on another. For instance, where one server supports a Web page for commercial online customers whose accounts are stored in a second remote database, it may be necessary to restore the database server first, before the Web server, to ensure account data can be retrieved using that networked software or service. Besides crash recovery, the ability to identify software-related dependencies would also prove useful for security management purposes, if that capability were available.

For further instance, in the case of the corruption or loss of individual machines, the particular configuration of those machines can be difficult or impossible to restore using existing network management platforms. While individual “kickstart” files reflecting initial configuration states can be received from targets or other machines during installation, those files can be lost or damaged in existing network management platforms, which have no facility for managing such files. In further regards, existing network management platforms or provisioning platforms provide no integrated tool to attempt the restoration of an entire datacenter or network, including all individual nodes and taking software dependency into account. Thus, there is a need in the art for methods and systems that provide centralized management of restoration and related operations.

DESCRIPTION OF THE DRAWINGS

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments of the present teachings and together with the description, serve to explain the principles of the present teachings. In the figures:

FIG. 1 illustrates an overall network in which systems and methods for automatic discovery of network software relationships can be practiced, according to various embodiments of the present teachings;

FIG. 2 illustrates an overall network in which systems and methods for automatic discovery of network software relationships can be practiced to various embodiments of the present teachings, including issuance of discovery commands and other features;

FIG. 3 illustrates an illustrative hardware configuration which can support systems and methods for automatic discovery of network software relationships according to various embodiments of the present teachings;

FIG. 4 illustrates a flowchart for software dependency processing, according to various embodiments of the present teachings;

FIG. 5 illustrates a flowchart of generation of a reverse kickstart file, according to various embodiments of the present teachings; and

FIG. 6 illustrates a flowchart of generating a network-scale restoration in controlled physical order, according to various embodiments of the present teachings.

DESCRIPTION OF EMBODIMENTS

Embodiments of the present teachings relate to systems and methods for automatic discovery of network software relationships. More particularly, embodiments relate to platforms and techniques for automatically probing a managed network to discover software and/or service dependencies between nodes in the network and other nodes or resources. According to embodiments, the nodes of a managed network can be interrogated to discover those applications services which supply input to a subject node, which services support or interact with a given node, or which other software-based resources are required for the operation of a given node. A network management platform can interrogate the managed network and build a dependency map encoding those and other software relationships. According to embodiments, a systems administrator or other user can thereby efficiently restore a disabled network in proper logic sequence to allow software and services to resume normal operation.

Reference will now be made in detail to exemplary embodiments of the present teachings, which are illustrated in the accompanying drawings. Where possible the same reference numbers will be used throughout the drawings to refer to the same or like parts.

FIG. 1 illustrates an overall system 100 in which systems and methods for automatic discovery of network software relationships can be implemented, according to various embodiments of the present teachings. In embodiments as shown, network management platform 102 can communicate with managed network 116 via a secure channel 148. Secure channel 148 can for instance be or include the secure channel and related resources described in co-pending U.S. application Ser. No. 12/130,424, filed May 30, 2008, entitled “Systems and Methods for Remote Management of Networked Systems Using Secure Modular Platform,” assigned or under obligation of assignment to the same entity as this application, and which application is incorporated by reference herein. Secure channel 148 can be or include, for example, a secure socket layer (SSL) connection, a channel established using a public/private key infrastructure, and/or other connections, channels, or protocols. Managed network 116 can include, as shown, a set of supervisory hosts 112, a set of targets 180, and/or other machines, devices, clients, and/or other hardware, software, storage or other resources.

While secure channel 148 is illustratively shown as one channel to managed network 116 or devices therein, it will be understood that in embodiments, secure channel 148 can comprise multiple channels or connections. In embodiments, secure channel 148 can instead be replaced by a non-secure channel or connection. In general, network management platform 102 can communicate with the managed network 116 and its constituent machines and resources, which can for instance comprise personal computers, servers, network-enable devices, virtual machines, and/or other devices, and manage the security of those machines under the supervision of network management platform 102. The network management platform 102 can host a set of engines, logic, and/or other resources to interrogate managed network 116 and identify software dependencies, reverse kickstart files, and/or network-wide restoration data or policies to manage the tracking, backup, and restoration of managed network 116 and/or its constituent nodes, machines, and resources. Network management platform 102 can, in embodiments as shown, host engines or logic including dependency engine 186, reverse kickstart generator 188, and network reconstruction engine 190. Network management platform 102 can communicate with network store 104 to establish, maintain and store dependency map 162, set of reverse kickstart files 166, and network reconstruction data 168. Other logic and data stores can be used.

The network management platform 102 can also for instance establish a set of security policies for managed machines in its domain, including for example an access control list (ACL) and/or other policies, lists, or parameters for the maintenance and management of managed network 116. Security policies and other data can for instance be stored in network store 104.

Network management platform 102 can establish secure connection 148 or other connection to managed network 116, and conduct network scanning, backup and restoration operations, according to various embodiments described herein.

According to embodiments in one regard, network management platform 102 can operate to connect to a given one or more of managed network 116 to probe managed network 116 for software-based dependencies in two or more nodes in the network. As for example more particularly shown in FIG. 2, network management platform 102 can host a dependency engine 186 to interrogate managed network 116 and identify software-based dependencies amongst targets, hosts, and/or other nodes in managed network 116. Dependency engine 186 can for instance issue one or more discovery command 120 to cause set of targets 180, set of supervisory hosts 112, and/or other nodes in managed network 116 to report applications or other software or services with which those nodes communicate, for instance to receive input or initiate services. According to various embodiments, one or more discovery command 120 can probe for, request, and transmit various types of data to dependency engine 186 to generate dependency map 162 based on the responses from nodes in managed network 116. One or more discovery command 120 can, for instance, probe for reports on virtual hosts and guests, including what virtual process runs on what node or machine. One or more discovery command 120 can likewise probe for reports on which nodes or machines have which other machines in an SSH (Secure Shell protocol) authorized keys file, and/or which targets in set of targets 180 may have received authentication to communicate with/act as a local supervisory node to other low-level targets. One or more discovery command 120 can, for instance, probe for reports on what configuration files indicate that network services, such as, for instance, sendmail, dns, syslog, or other services, reference other nodes or machines. One or more discovery command 120 can, for yet further instance, probe for reports on what network connections are present in a given node or machine, analogous to a “netstat—a-n” command. Other commands, requests, reports, or other probes can be used.

After receiving responses to one or more discovery command 120, network management platform 102 can build dependency map 162 encoding the software-based dependencies between set of target 180, set of supervisory hosts 112, and/or other nodes in managed network 116. Dependency map 162 can, for example, be configured as an ordered list, a tree, a graph, a database, a spreadsheet, and/or other data structure or coding. Dependency map 162 can encode relationships between software, services, and/or nodes of managed network 116, for instance to record that “Software Application A running on node 44 requires Software Application B running on node 32,” or other orderings or relations. According to embodiments in further regards, network management platform 102 can thereby access dependency map 162 and reconstruct an ordered set of software installations and/or machine re-initializations to be carried out after an interruption in managed network 116, such as a physical disaster, security intrusion, or other event. Network management platform 102 can also operate on dependency map 162 to examine managed network 116 for security, configuration or issues, and/or carry out cloning operations for any nodes in the network, or the network itself. Because necessary relations between software, services and/or nodes can be effectively encoded in dependency map 162, software restoration in a logically necessary order can be generated at any needed time, enhancing the security and reliability of managed network 116. According to embodiments in further regards, network management platform 102 can operate to connect to managed network 116 and extract or build a reverse kickstart file 166, as also for instance illustrated in FIG. 2.

In embodiments as shown, network management platform 102 can issue one or more discovery commands 120 to locate any original kickstart files generated during an installation or initialization process for any node of managed network 116. The kickstart files which one or more discovery commands 120 seek out can be kickstart files and related data such as that described in co-pending U.S. patent application Ser. No. 11/763,315 and U.S. patent application Ser. No. 11/763,333, the disclosures of which are incorporated herein, in their entirety, by reference. According to various embodiments, network management platform 102 can locate any kickstart file copied by the installer and perform other operations using reverse kickstart generator 188, or other logic. If present, in embodiments “/root/anaconda.ks can be used as a base point to locate any kickstart files. Network management platform 102 can further access a package management tool or application programming interfaces (APIs), such as “rpm-va” or yum-based protocols, to determine what files that were originally present in installation packages have been user-modified. If user-modified files are found, those files can be embedded in reverse kickstart file 166, which in embodiments can be stored to “% post” or other locations. Network management platform 102 can further scan managed network, for instance using available storage APIs, to detect mounted file systems in a subject node and regenerate a new section of reverse kickstart file 166 that represents the storage section of the installation. This section can, in embodiments, replace the section in the original kickstart file whose image is being extracted. In embodiments, network mounted file systems can also be detected, and for instance stored to “% post” or other locations.

In embodiments, network management platform can likewise embed/include any authorized keys in the reverse kickstart file 166, as well as any files not controlled by a package management tool. In embodiments, instead of embedding/including said files in reverse kickstart file 166, that data can be stored to a backup system (such as, for example, bacula or rdiff-backup) and be made to include in the reverse kickstart file 166 the paths/commands needed to restore them. Once the reverse kickstart file 166 is generated, it can then be used stand alone or with a provisioning platform 170 to recreate the systems, in current configurations, from scratch, effectively automatically reverse engineering their configuration from their current state. The systems administrator can later adjust reverse kickstart 166 as needed to begin maintaining and updating captured systems via that mechanism, or use that resource to clone or restore more systems as needed, for instance for disaster recovery purposes.

According to embodiments in one regard, network management platform 102 can operate to connect to a given one or more of managed network 116 to generate a reconstruction of managed network 116 in the event of a significant or large-scale network malfunction, intrusion, or disaster. As for example also more particularly shown in FIG. 2, network management engine 102 can host network reconstruction engine 190 communicating with network reconstruction data 168 to perform an ordered restoration of a portion or all of managed network. In embodiments as shown, network reconstruction engine 190 can incorporate and/or access both reverse kickstart files and software dependency mappings as described herein, to perform a comprehensive or large-sale restoration of managed network 116 in a logically structured physical order. In embodiments, network management engine 102 via network reconstruction engine 190 can operate to generate a set of reconstruction commands 172 by accessing dependency map 162, reverse kickstart file 166, network reconstruction data 168, and/or other data or resources. In embodiments, the network reconstruction engine 190 can access dependency map 162, and identify a set of ordered nodes to restore to managed network 116. For each node, network reconstruction engine 190 can access a reverse kickstart file 166 or other source to re-provision the software resources of that node, or otherwise restore that node to managed network 116. Some or all of the restoration or re-provisioning of the nodes of managed network 116 can be performed by a provisioning platform 170, such as that described in the aforementioned co-pending U.S. patent application Ser. No. 11/763,315 and U.S. patent application Ser. No. 11/763,333, or other platforms or systems. In embodiments, files for restoration can be served from a backup server or system, if file sizes dictate. Once a network-wide or complete datacenter restoration has been performed in dependency or other order, network management platform 102 can execute one or more validation or debugging tests, using for example a secure connection and related resources such as those described in the aforementioned U.S. patent application Ser. No. 12/130,424, filed May 30, 2008, entitled “Systems and Methods for Remote Management of Networked Systems Using Secure Modular Platform.” Other network management platforms or systems can be used.

FIG. 3 illustrates an exemplary diagram of hardware and other resources that can be incorporated in a network management platform 102 configured to communicate with managed network 116 and/or other resources, according to embodiments. In embodiments as shown, the network management platform 102 can comprise a processor 124 communicating with memory 126, such as electronic random access memory, operating under control of or in conjunction with operating system 130. Operating system 130 can be, for example, a distribution of the Linux™ operating system, the Unix™ operating system, or other open-source or proprietary operating system or platform. Processor 124 also communicates with a network store 104, such as a database stored on a local hard drive. Processor 124 further communicates with network interface 128, such as an Ethernet or wireless data connection, which in turn communicates with one or more networks 110, such as the Internet or other public or private networks. Processor 124 also communicates with dependency engine 186, reverse kickstart generator 188, and network reconstruction engine 190, to execute control logic and perform management and restoration processes described herein. Other configurations of the network management platform 102, associated network connections, and other hardware and software resources are possible. While FIG. 3 illustrates network management platform 102 as a standalone system comprises a combination of hardware and software, network management platform 102 can also be implemented as a software application or program capable of being executed by a conventional computer platform. Likewise, network management platform 102 can also be implemented as a software module or program module capable of being incorporated in other software applications and programs. In either case, network management platform 102 can be implemented in any type of conventional proprietary or open-source computer language.

FIG. 4 illustrates a flowchart of processing to manage software dependency and network restoration, according to various embodiments. In 402, processing can begin. In 404, a secure connection 148 or other connection(s) can be established from network management platform 102 to managed network 116. In 406, one or more discovery command 120 can be issued from network management platform 102 to hosts, targets, and/or other nodes in managed network 116 to obtain a report of software and/or services upon which each node depends or requires. In 408, network management platform 102 can generate a dependency map 162 indicating the dependency order of relationship between each node and/or its software complement and other nodes and their software complement in managed network 116. Dependency map 162 can be generated, for example, in the form of an ordered list of nodes and/or applications or other software, a tree, a linked list, a graph, a spreadsheet, a database, and/or other data structure. In 410, network management platform 102 can initiate a software reconstruction, re-installation, recovery, or other provisioning or installation process on managed network 116. In 412, network management platform 102 can extract a map, graph, spreadsheet, or other representation or encoding of dependency relationships between nodes in managed network 116 and/or the software, services and/or other resources or processes installed, hosted, or accessed on nodes of managed network 116. In 414, network management platform 102 and/or other logic can identify an ordering or other relationship of the nodes of managed network 116 to be rebuilt or re-installed using dependency map 162 and/or other data. In 416, a re-provisioning or other software re-installation process can be executed on managed network 116, for instance, to provision or install operating systems, applications, data, services, and/or other software to nodes of managed network 118. In 418, a record of the re-installation activity can be generated and/or stored, as appropriate. In 420, as understood by persons skilled in the arts processing can repeat, return to a prior processing point, jump to a further processing point, or end.

FIG. 5 illustrates a flowchart of processing to manage the generation of a reverse kickstart or other reverse installation file, according to various embodiments. In 502, processing can begin. In 504, network management platform 102 can search managed network 116 to locate any kickstart or other installation file left over or copied by an installation process on nodes of managed network 116. In 506, a reverse kickstart file 166 can be created/initialized. In 508, network management platform 102 can use a package management tool and/or other tool or resource to inspect/identify any user-modified files located on the target machine or other node for which a reverse kickstart file 166 or other reverse installation file is being generated.

In 510, network management platform 102 can located a mounted file system(s) and regenerate a new section of the kickstart file representing a storage section of the installation. In 512, the storage part of the original installation file can be replaced, and/or added or inserted in reverse kickstart file 166. In 514, network management platform 102 can generate a record of mounted file systems on the node associated with reverse kickstart file 166. In 516, any authorized keys or other security credentials from the subject node can be embedded in reverse kickstart file 166, and/or the security data can be stored to a backup system in which the security data along with paths, commands, and/or keys needed to restore the node can be inserted. In 518, any files not detected or controlled by the package management tool used by network management platform 102 can be embedded reverse kickstart file 166, or those files can be stored to a backup system in which those files along with paths, commands, and/or keys needed to restore the node can be inserted. In 520, the network management platform 102 can access reverse kickstart file 166 and initiate a machine restoration process for one or more nodes in managed network 116, as appropriate. In 522, as understood by persons skilled in the art, processing can repeat, return to a prior processing point, jump to a further processing point, or end.

FIG. 6 illustrates a flowchart of processing to perform network-scale restoration, according to various embodiments. In 602, processing can begin. In 604, a recovery process for a managed network 116 and its components, and/or other datacenter entity can be initiated. In 606, a first node or nodes required for network recovery can be identified based on dependency map 166, for instance, by identifying the lowest/highest node in managed network 116 required by other nodes. In 608, a reverse kickstart file 166 and/or other backup files or data associated with the first required node(s) can be extracted from network store 104 or other location. In 610, network management platform 102 can build installation file(s) for the first required node(s) can be built from reverse kickstart file 166 and/or other data.

In 612, the re-provisioning or re-installation of the first required node(s) can be initiated using a provisioning platform 170 and/or other tools or resources. In embodiments, provisioning platform 170 and related processes can be or include those described in the aforementioned copending U.S. patent application Ser. No. 11/763,315 and U.S. patent application Ser. No. 11/763,333. In 614, the identification of a next required node(s) based on dependency map 162 can be begun/repeated, and the extraction of associated reverse kickstart file 166 can be begun/repeated, until the last node of managed network 116 is processed/re-installed. In 616, the re-installation of all nodes recovered in managed network 116 can be verified via network management platform 102 after all nodes have been processed. In 618, network management platform 102 can generate a record of re-installation or recovery activity, as appropriate. In 620, as understood by persons skilled in the art, processing can repeat, return to a prior processing point, jump to a further processing point, or end.

The foregoing description is illustrative, and variations in configuration and implementation may occur to persons skilled in the art. For example, while embodiments have been described in which a single network management platform 102 manages and maintains managed network 116, in embodiments, multiple engines, servers, or other entities can cooperate to perform network management functions. For further example, while embodiments have been described in which restoration, dependency, and/or other operations can be performed on a single host or target, in embodiments, multiple diagnostic targets or an entire group of managed machines can be operated on at one time or together. Other resources described as singular or integrated in some embodiments can in embodiments be plural or distributed, and resources described as embodiments as multiple or distributed can in embodiments be combined. The scope of the present teachings is accordingly intended to be limited only by the following claims

Claims

1. A method comprising: issuing a discovery command to a plurality of nodes of a network, the discovery command requesting information about dependencies between software installed on the plurality of nodes;receiving responses to the discovery command from the plurality of nodes of the network, the responses comprising the information about the dependencies;generating, based on the responses, a dependency map representing a plurality of dependency relationships between the software installed on the plurality of nodes of the network;determining an ordered set comprising: software installations for each software installed on the plurality of nodes, comprising a first software and second software, wherein installing the first software and the second software comprises retrieving a first reverse automatic installation file and a second reverse automatic installation file, anda machine re-initialization order to be carried out in view of the dependency map after an interruption in the managed network;saving the ordered set as network reconstruction data;installing, by a processor and according to the ordered set in the network reconstruction data, the first software on a first node of the plurality of nodes and the second software on a second node of the plurality of nodes; andre-initializing, after an interruption, machines in the network in a sequential order according to the software installations and machine re-initialization order determined in the ordered set.

2. The method of claim 1, further comprising storing a record of the installation of the first software and the second software.

3. The method of claim 1, wherein the dependency map comprises at least one of a tree, a linked list, a table, or a database.

4. The method of claim 1, wherein the plurality of dependency relationships comprises a representation of a first plurality of software applications whose output comprises input to a second plurality of software applications.

5. The method of claim 1, wherein the plurality of dependency relationships comprises a representation of a first plurality of software applications whose network-enabled services are accessed by a second plurality of software applications.

6. The method of claim 1, wherein installing the first software and the second software comprises retrieving a first reverse kickstart file and a second reverse kickstart file.

7. A system comprising: a memory to store a dependency map representing a plurality of dependency relationships between software installed on a plurality of nodes of a network; anda processor coupled to the memory, the processor to: issue a discovery command to the plurality of nodes of the network, the discovery command requesting information about dependencies between software installed on the plurality of nodes;receiving responses to the discovery command from the plurality of nodes of the network, the responses comprising the information about the dependencies;generate, based on the responses, the dependency map;determine an ordered set comprising: software installations for each software installed on the plurality of nodes, comprising a first software and second software, wherein the processor is to install the first software and the second software by retrieving a first reverse automatic installation file and a second reverse automatic installation file from the memory, anda machine re-initialization order to be carried out in view of the dependency map after an interruption in the managed network;save the ordered set as network reconstruction data;install, according to the ordered set in the network reconstruction data, the first software on a first node of the plurality of nodes and the second software on a second node of the plurality of nodes; andre-initialize, after an interruption, machines in the network in a sequential order according to the software installations and machine re-initialization order determined in the ordered set.

8. The system of claim 7, wherein the processor is further to store a record of the installation of the first software and the second software.

9. The system of claim 7, wherein the dependency map comprises at least one of a tree, a linked list, a table, or a database.

10. The system of claim 7, wherein the plurality of dependency relationships comprises a representation of a first plurality of software applications whose output comprises input to a second plurality of software applications.

11. The system of claim 7, wherein the plurality of dependency relationships comprises a representation of a first plurality of software applications whose network-enabled services are accessed by a second plurality of software applications.

12. The system of claim 7, wherein the processor is to install the first software and the second software by retrieving a first reverse kickstart file and a second reverse kickstart file from the memory.

13. A non-transitory computer readable storage medium having instructions encoded thereon which, when executed by a processor, cause the processor to: issue a discovery command to a plurality of nodes of a network, the discovery command requesting information about dependencies between software installed on the plurality of nodes;receive responses to the discovery command from the plurality of nodes of the network, the responses comprising the information about the dependencies;generate, in view of the responses, a dependency map representing a plurality of dependency relationships between the software installed on the plurality of nodes of the network;determine an ordered set comprising: software installations for each software installed on the plurality of nodes, comprising a first software and second software, wherein installing the first software and the second software comprises retrieving a first reverse automatic installation file and a second reverse automatic installation file, anda machine re-initialization order to be carried out in view of the dependency map after an interruption in the managed network;saving the ordered set as network reconstruction data;install, by the processor and according to the ordered set in the network reconstruction data, the first software on a first node of the plurality of nodes and the second software on a second node of the plurality of nodes; andre-initialize, after an interruption, machines in the network in a sequential order according to the software installations and machine re-initialization order determined in the ordered set.

14. The non-transitory computer readable medium of claim 13, wherein the operations to further cause the processing device to store a record of the installation of the first software and the second software.

15. The non-transitory computer readable medium of claim 13, wherein the dependency map comprises at least one of a tree, a linked list, a table, or a database.

16. The non-transitory computer readable medium of claim 13, wherein the plurality of dependency relationships comprises a representation of a first plurality of software applications whose output comprises input to a second plurality of software applications.

17. The non-transitory computer readable medium of claim 13, wherein the plurality of dependency relationships comprises a representation of a first plurality of software applications whose network-enabled services are accessed by a second plurality of software applications.

18. The non-transitory computer readable medium of claim 13, wherein to install the first software and the second software the processing device to retrieve a first reverse kickstart file and a second reverse kickstart file.