Method of and system for computer system state checks

Abstract

A system for and method of system state analysis of a computational system. The method is comprised of capturing selective state information of a computational system configured to operated with one or more guest machines running on a virtual machine layer and configured to output state information. The state information is then analyzed to for compliance checking. The system for system state analysis is comprised of a storage system, computation hardware configured to run the guest machines and the virtual machine layer, guest machines, a virtual machine layer configured to output guest machine state information, a system state snapshot server configured to control the virtual machine layer for the capture of state information.

Description

FIELD OF THE INVENTION

The invention relates to methods of and systems for minimally invasive state checks of electronic computational systems including but not limited to computer systems and network systems for the purposes not limited to compliance, configuration, or security checks.

BACKGROUND OF THE INVENTION

The state of a system can include but is not limited to, the existence of files on a storage system, file permissions, file content, file permissions, registry key existence, registry key permissions, registry values, system software versions, kernel software versions, system variables, network configuration, hardware configuration, data structures within the operating system kernel, hardware state, processor states, database configuration, database content, and user permission levels. For security purposes, system auditing, or system compliance checking, there is a need to quickly and with minimal impact capture and analyze the state of a system.

The present state of the art uses individual software agents installed on each processing system to capture and analyze the system state. However, as discussed below there are several disadvantages to this configuration.

The system shown in FIG. 1A shows one embodiment of a prior art system 100 utilizing a software agent 120A-120C to provide a state check of the processing system. The embodiment of the computational system 100 shown includes system software 110, kernel software 130, the computational hardware 140 on which the system software 110 and kernel software 130 runs, and a storage system 150. The system software 110 includes application software for performing task specific operations. The kernel software 130 typically includes an operating system such an Unix, Linx; Microsoft NT, XP, Windows Vista®, and Windows Server® operating systems. The hardware 150 can be based on any commonly found CPU, memory, networking hardware, and other supporting hardware. The agents can be located within the system software, with the kernel, or with the hardware. Usually only one agent is found at either the system, kernel, or hardware level within a system.

In one configuration, an agent 120A located with the system software, is used to capture and analyze the state of the system 100. The agent 120A runs as a separate process(es) or application concurrently with the other system software 110. Disadvantages of a system software based agent 120A is that the agent 120A consumes large amounts of system resources such as CPU processing cycles, storage system bandwidth, and storage system space for taking state snapshots. Further, if the agent runs in parallel with the system software, the system configuration could change while capturing the system state and thus the agent can only report the system state at points within a potentially large time window. A further disadvantage of a system software based agent is that such an agent is limited in the scope of system state information that can be analyzed. A system software based agent 120A does not have access to the operating system kernel data structures and thus has limited ability to analyze the state of the kernel, or evaluate the kernel for viruses or root kits.

In a second configuration, a kernel based agent 120B is used to capture and analyze the state of the system 100. This configuration enables the agent to check a broader scope of state information including checks on the kernel data structures. However, this configuration has the same drawbacks as the first configuration. Specifically, the kernel based agent 120B consumes a significant amount of CPU cycles, storage system bandwidth, and storage space.

In a third configuration, a hardware based agent 120C is used to capture and analyze the state information of the system. A hardware board agent 120C is connected to the hardware 140. When a state check is to be performed, the hardware board agent 120C stops the processor and examines files, registries, system software states, and examines kernel data structures. Typically, the computational hardware 140 is stopped during the state snapshot capture and analysis. This implementation has the disadvantage of stopping the CPU and thus the computing system 100 unavailable during the state check. Further, the extra hardware required has the additional disadvantage that the hardware based agent board 120C can reduce the reliability of the system, and increase the power usage. Further, the cost varies with the number of server cards given that an agent board 120C is required for each server card. Further, a hardware board agent can introduce security issues of physically having to gain access to the hardware.

FIG. 1B illustrates a typical prior art data center 100B where many of processing units 110A-110n can be grouped. Each of the processing unit 110A-110n can be a stand alone server or a server blade, each having its own kernel (operating system). Each processor unit 110A-110n has an agent 120A-120n running on the processor. Such a configuration of servers has two drawbacks. First, each of the agents consumes a large amount of computational processing power for each processor. Secondly, the installation, maintenance, and update of hundreds of agents 120A-120n can be costly and time consuming.

FIG. 1C illustrates another prior art configuration 100C of processing units running guest machines 160A-160n. A guest machine 160 typically includes application software and a kernel that interfaces with the computational hardware and system resources through the virtual machine layer 170. The virtual machine layer 170 makes it appear to each guest machine that it is exclusively running on the hardware 180. Again each guest machine 160A-160n has agent software 120A-120n, each of which demands from the hardware 180 computation processing power to execute, and the time and expense to install, maintain, and upgrade. The drain on system resources, processing power and time for maintenance increases linearly with each additional guest machine. This configuration does not provide economies of scale with an increasing number of guest machines.

What is needed is a means to analyze the state of an electronic system with minimal impact to a performance of the system, a state analysis solution that is easy to maintain, and does not reduce the reliability of a system.

SUMMARY OF THE INVENTION

The invention provides a new, less intrusive, and easier to maintain system for capturing and analyzing the state of a computational processing system. The computational system is typically a computational system such as a server farm or data center and can have a number of guest computational machines running on a virtual machine layer. A state snapshot server can execute on a guest machine or can execute on separate hardware coupled to communicate with the virtual machine layer directly or through a network. Further the state snapshot server can work in conjunction with a configuration management server for compliance and security checking.

Within this document references are made to state information, state data, or a state snapshots which are considered to be equivalent. State snapshot information is comprised of persistent and non-persistent state information. Generally, non-persistent state information is associated with runtime information. State snapshot information includes file related information and physical memory information. The file related information includes but is not limited to file/directory existence, content, version, permissions and other attributes, registry key existence and permissions, registry value and existence, versions of the operating system, operating system components, attributes of system passwords such as the password length, and the age of the password, database configuration, schema structure and table values. Physical memory snapshot includes associated user memory and associated kernel memory. Snapshot state information found in kernel memory includes application runtimes, kernel runtime data structures, open network ports, network sockets connected with particular hosts, users logged into the system, virtual hardware configuration state information, and processor state information or a combination thereof. User memory state information includes state information on what programs a user is running and the state of user data structures that can indicated the presence of mal-ware. System security checks can involve the analysis of a combination of the above state information. If the type of system state analysis is a compliance check, then typically kernel checks, device state and processor state information is not included.

Further, the analysis the of the state information can be performed in terms of a policy. The policy is an information structure containing parameters, characteristics, ranges, and specific data relating to system state information that is captured and analyzed. Preferably the policy information is configurable and stored on the State Snapshot Server or any storage device accessible by the State Snapshot Server. For example, the policy can be a file stored on a disk directly attached to the State Snapshot Server or a disk on a network through which the State Snap Server can communicate. Also contemplated by the invention is the policy information being coded within the State Snapshot Server application. The policy specifies attributes by which to analyze of a piece of the state snapshot information. For example, the policy information can set ranges for state information such as but not limited to registry values ranges, database value ranges, environment or system variable ranges, and minimum password lengths. System characteristics can include whether a files is writable, and who has permission to change a file. Further, the policy for runtime information can include which processes should be running, who can be logged in, what times a user can be logged in, what communication channels are active, which network ports are open, what host can be connected to the network sockets, and the analysis of kernel data structures to verify that the kernel data structures are not corrupted or infected by a root-kit virus. Specific data relating to system state can include but are not limited to the password names, names of allowed hosts, allowed or not allowed users, allowed communication protocols.

The specification refers to triggers for taking and analyzing snapshots of system state data. Triggers include changes in the configuration change information. Configuration change information is described in U.S. patent application Ser. No. 12/008,274, filed Jan. 9, 2008, entitled “METHOD AND APPARATUS FOR PROCESS ENFORCED CONFIGURATION MANAGEMENT” that is incorporated by reference in its entirety. Further, the triggers can include host content change requests. Host content change request are described in pending U.S. patent application Ser. No. 11/346,741, filed Feb. 2, 2006, entitled “ENFORCING ALIGNMENT OF APPROVED CHANGES AND DEPLOYED CHANGES IN THE SOFTWARE CHANGE LIFE-CYCLE” that is incorporated by reference in its entirety. Further, the triggers can be based on a specified user logging on or off a system, a process stop/start, or a reboot event on a Guest Machine, a State Snapshot Server, or a Configuration Management Server.

A first aspect of the invention is for a method of computational system state analysis. The method involves the steps of capturing selective state information of a computational system having one or more guest machines running on a virtual machine layer configured to output the state information. In a subsequent step, the state information is analyzed. The guest machines can have the same or different kernels (operating systems) running each of the guest machines.

In one embodiment, the state information is selectively specified on a guest machine basis. The capture of this state information can be for one guest machine, multiple guest machines or all guest machines.

In another embodiment, the analysis of the state information is according to a policy. The analysis can include processing according to the policy persistent state information, non-persistent information, or a combination thereof. Further, the processing of non-persistent state information includes, kernel state data, user state data and a combination thereof. The analysis of the kernel data structures can include verifying that root kit has not been configured into the system, the absence of a virus, or a given configuration of the kernel and the analysis of the physical user data snapshot can check fro mal-ware or whether the programs being run conform with the policy.

In another embodiment, the state information is stored on a storage system. The storage system can be directly coupled to the hardware running the virtual machine layer and guest machines or can be a distributed storage system. The coupling of the storage system can be directly tied to the computational hardware executing the virtual machine layer and guest machines or can be coupled using a network interface including ethernet, optical, or SAN (Storage Area Networks).

In a further embodiment, the method includes the step of controlling the virtual machine layer to selectively capture system state information. The capture of state information includes selective capture of information on a single guest machine or the capture of state information from a specified plurality of guest machines. The control of the virtual machine layer that generates snapshots can originate from a process or a software program running on a guest machine or by a process or software program running on a different computational hardware. The coupling of the control process with the virtual machine layer can be made directly or indirectly, such as through a network. Also, control over other communication links such as a direct communication link is contemplated. Alternatively, special purpose software or networking software can be configured to communicate with the virtual machine layer through the kernel running on the virtual machine.

In another embodiment, the method can include configuration management steps used in the maintenance and upgrading of an electronic computational system. The method can include the creating or opening configuration change ticket. The capture of the system state information and the analysis of the state information can be triggered in response to the creating or opening the configuration change ticket. Preferably, in the configuration change that is specified in the configuration change ticket is implemented before the capture and analysis of the state information for the one or more guest machines.

In another embodiment the capturing and processing of a state snapshot can be invoked by a number of triggers. These triggers can include a computer scheduled event, a change to the policy, or event triggers. The policy change can include a change to a file containing policy information or an application that is given an indication that the policy is changed. The triggers are described above and in the incorporated references.

In a further embodiment, the issuing of the change ticket includes using change process rules in the configuration management of a computational system. The analysis of the system state can utilize the change process rules in that analysis. The analysis can use the change process rules to extend the configuration, compliance, or security analysis to include when configuration change was made and by whom. Alternatively, the analysis can be used to determine any conflicts between the configuration, compliance, or security requirements and a configuration change. In one configuration, the analysis of the system state information is performed on the same computational hardware which executes the virtual machine layer and guest machines. In another embodiment, separate computational hardware is used for the state snapshot server. Further, in one embodiment, the configuration management system executes on one of the guest machines. In another embodiment, the configuration management system executes on hardware separate from the hardware executing the virtual machine layer and the guest machines.

Another aspect of the invention, is for a system for analyzing the state of a computational system. The system includes one or more guest machines, a storage system, and a virtual machine layer interfaced to the one or more guest machines. The virtual machine layer is configured to generate system state information for the one or more guest machines. A state snapshot server is configured to control the virtual machine layer. A first computation hardware unit is configured to execute the one or more guest machine and the virtual machine layer, and is coupled to the storage system. The storage system can be coupled through a network including but not limited to ethernet or a storage area network. Preferably, the virtual machine layer is configured to take fast system snapshots of state data without effecting the performance of any of the guest machines. The first computational hardware unit can comprise a single computational processing unit or a plurality of processing units. The processing units can be but is not limited to a single processor, multiple processing cores, a server blade, an array of processors or a combination thereof.

Further the analysis of the state information data can be analyzed according to a policy. A trigger can be used to invoke a state snapshot and analysis. The triggers are described above and in the incorporated references.

In one embodiment, the state snapshot server is configured to execute within one of the guest machines. In a different embodiment, the state snapshot server is configured to execute on a second computational hardware unit. The second computational hardware unit can be located away-from or near the first computational hardware unit. The communication channel between the first and second computational hardware unit can be through a wired or wireless network or through a dedicated communication channel using standard or proprietary communication protocols. Further, the taking and process of a state snapshot can be invoked by a number triggers. These triggers can include a computer scheduled event, a change to the policy, or event triggers. The policy change can include a change to a file containing policy information or an application that gives an indication that the policy is changed. The triggers are described above and in the incorporated references.

In another embodiment, the system further comprises a configuration management server. The server is configured to communicate control and configuration information with the state snapshot server. The configuration management server can issue a configuration change ticket. In response to the issuing the change ticket, the configuration management server can configure the state snapshot server to perform a system snapshot and analysis of the system state. The analysis can be performed according to a policy. In one embodiment the system state snapshot is taken before the configuration change indicated in the issued configuration ticket. In another embodiment, the state snapshot and analysis is performed after the configuration change is performed. The analysis of the system snapshot can include utilizing the change process rules information communicated by the change management server to the state snapshot server. Further, all or part of the state analysis results can be incorporated into the change ticket. The configuration management server can either read the information from an accessible file, request the information from the state snapshot server, or the state snapshot server can send the data to the configuration management server.

In another aspect of the present invention, the invention embodies one or more machine readable storage devices having processor readable code embodied on the storage devices for programming computational hardware to perform a method of system state analysis of a computational system. The code comprises the steps of capturing selective state information of a computational system having one or more guest machines running on a virtual machine layer configured to output the state information and analyzing the state information. The machine readable storage devices can include but is not limited to CD-ROMs, DVDs, hard disks, solid-state storage devices, tape, floppies, or other magnetic media. In one embodiment the state information is selectively captured from one or more guest machines. In another embodiment, the captured system state data is process according to a policy. In another embodiment the processor readable code is embodied on the storage devices is configured to read a system state snapshot from a storage system and then analyze the snapshot information.

BRIEF DESCRIPTION OF THE DRAWINGS

The invention is better understood by reading the following detailed description of exemplary embodiments in conjunction with the accompanying drawings.

FIG. 1A illustrates different configurations of a prior art processing unit with an agent that performs a system state analysis. The agent is shown operating selectively at a system level, a kernel level, or at a hardware level.

FIG. 1B illustrates a prior art server farm or processing center with multiple processing units, wherein each processing unit has a system analysis agent.

FIG. 1C illustrates a prior art processing system with multiple guest machines running on a virtual machine layer where each guest machine has an agent for performing system state analysis.

FIG. 2 illustrates one embodiment of the current invention having a plurality of guest machines running on a virtual machine layer, a state snapshot server, and a configuration management server.

FIG. 3 is a block diagram illustrating a method of taking system state snapshots and performing state analysis on a system state snapshot.

FIG. 4 illustrates a block diagram of an exemplary computing device configured to implement a state snapshot server.

DETAILED DESCRIPTION OF THE INVENTION

The following description of the invention is provided as an enabling teaching of the invention in its best, currently known embodiment. Those skilled in the relevant art will recognize that many changes can be made to the embodiment described, while still obtaining the beneficial results of the present invention. It will also be apparent that some of the desired benefits of the present invention can be obtained by selecting some of the features of the present invention without utilizing other features. Accordingly, those who work in the art will recognize that many modifications and adaptations to the present inventions are possible and can even be desirable in certain circumstances, and are a part of the present invention. Thus, the following description is provided as illustrative of the principles of the present invention and not in limitation thereof, since the scope of the present invention is defined by the claims.

The illustrative embodiment of the invention provides means for analyzing the system state of a processing system for purposes including but not limited to configuration management, compliance or security checking, security enforcement or a combination thereof. Processing systems that could benefit from the invention include but are not limited server farms, data centers, and processing clusters. Also contemplated by the invention is the use of the invention within networking equipment. Further, the methods of and system for system state snapshotting and analysis can be integrated with configuration management systems to provide control of configuration changes, post change compliance security checking or a combination thereof. After an authorized configuration management change is implemented, a compliance check can be performed to validate that a desired system configuration is made. Any standard configuration management system can be used. Also, a process enforced configuration management system can be utilized. Detail of such a system can be found in U.S. patent application Ser. No. 12/008,274, filed: Jan. 9, 2008, entitled “METHOD AND APPARATUS FOR PROCESS ENFORCED CONFIGURATION MANAGEMENT” which is hereby incorporated by reference in its entirety.

The invention comprises a state snapshot server which controls a virtual machine layer in the taking of system state snapshots and the analysis of the snapshot data.

The invention provides a number of benefits. First, using the virtual machine layer to take system state snapshots is typically much quicker than taking a snapshot from the a guest machine. Many of the current implementations of virtual machine layer utilize incremental file systems where the state of the files, including the content can quickly be stored away. Secondly, the analysis of the system state can be performed with minimal impact to the performance of each guest machine. Further, a system state snapshot taken by the virtual machine layer can include persistent and non-persistent state information where the non-persistent state information includes physical memory, both user and kernel, state information. As described above, the state snapshot information can be analyzed according to a policy. This includes the analysis of non-persistent system snapshot data that includes kernel and user physical memory. The analysis of the physical memory can include but is not limited to validating kernel data structures have not been corrupted by a root kit or a virus, that the kernel is properly configured, and that the kernel has the required components installed. Further the user data memory can be analyzed to determine whether the programs being run conform with the policy and whether any of the programs are corrupted by mal-ware. A major benefit of the current invention is that only one snapshot server has to be installed, maintained, and upgraded as oppose one for each guest machine.

An illustrative embodiment of the present invention is shown in FIG. 2. The system 200 is illustrates one embodiment of a system 200 that provides system state snapshot analysis and can further provide system configuration management. The system 200 includes computational hardware 230. A virtual machine layer 220 executes on the computational hardware 230. One or more guest machines 210a-210n can interoperate with the computational hardware 230, interfacing through the virtual machine layer 220. The computational hardware 230 are coupled to the storage systems 240, configuration management server 260/260′, and a system state snapshot server 250/250′. The computational hardware 230 can include network hardware 235 to interface with a storage system 240 or the other network based processing components including but not limited to the configuration management server 260 and the state snapshot server 250. The computational hardware 230 can include a single core processor, a multi-core processor, a server a server blade having multiple multi-processors, or a racks of server blades within a data center providing distributed computational resources.

The virtual machine layer 220 is preferably a software component that runs on the computational hardware 230. The virtual machine layer 220 is to allows multiple kernels 215A-215n, also referred to as operating systems, to share the same computational hardware and its associated resources including the storage systems 240 and the network communications 235. The operating system 215A-215n on each guest machine can be the same operating system, different versions/configurations of the same operating system, a different operating system 215 or a combination thereof. Each operating system is able to run concurrently on the same computational hardware 230 while the virtual machine layer 220 makes it appear to each operating system 215 that it the only operating system 215 controlling the hardware 230. The advantage of a system configured with a virtual machine layer 220 is that different software applications can use either a different operating system 215, a different version of the operating system 215, or different configuration of an operating system 215. For example, an accounting software package may require Unix while the engineering tools require two different versions of version of Linx, while Microsoft Windows® Server is used for general file sharing services. Instead of requiring a separate dedicated computational hardware 230 for each of the of different operating systems 215 and applications, the computational hardware 230 can be easily shared. Two commonly used virualization machines are VMware®, Widows Virtual Server by Microsoft®, or Microsoft Virtual PC®. The virtual machine layer 220 has an interface for the storing a snapshot of the state of file system and a kernel state each guest machine 210A-210n. These snapshots can be taken by a state snapshot server 250/250′, specifying for each guest machine the system state snapshot to be taken. Further, the virtual machine layer 220 can implement an incremental file system. The incremental file system operates by saving and tracking incremental changes to a file. Thus, a snapshot of the file does not require the storage of an entire file but only the storage of references to the incremental changes up to the time when a snapshot is taken. Saving this incremental information takes significantly less time than storing a copy of an entire file for state analysis.

The computational hardware 230 can range from a single processor to a distributed data center utilizing racks of server blades. Thus, the virtual machine layer 220 abstracts the computational hardware 230 allowing the computational resource to be more effectively shared between different applications requiring different operating system environments. The computational hardware 230 is coupled to the networking hardware 235 for interfacing with storage devices or to other processing systems including but not limited to configuration management servers 260 or state snapshot servers 250. The interface is not limited to a single network or a specific type of network. The connection 242 to the storage system 240 can be over ethernet or storage area networks.

The guest machines 210A-210n include system software including applications, and kernel software 215A-215n (the operating system). The kernel software 215 accesses hardware resources through the virtual machine layer 220 which makes it appear to each kernel software 215 of each guest machine 240 as if it is the only operating system controlling the hardware resources. Each guest machine 210 can access the network 242 or storage systems 240 attached to the computational hardware 230. The state snapshot server 250′ and the configuration management server 260′ can execute as an application or process on one of the guest machines 210.

The storage system 240 can be any commonly found media designed to interface with computational systems such as data server. The storage system 240 can include disk drives, solid state storage, tape drives or any other magnetic media. Preferably, the interface with the storage system 240 has a high bandwidth. A SAN or storage area network 242 can be used to couple the storage system 240 with the computational hardware 230. The guest machines 210 can access the storage system through the network 242. The storage system 240 can be partitioned such that each guest machine 210 only has access to a part of the storage system 240. The state snapshot server 250 may also be coupled to the storage system over a SAN or through another type of network 242. State snapshots or information referenced by the snapshot for an incremental file systems is stored on the storage system or can be directly used by the state snapshot server 250/250′.

The state snapshot server 250/250′ controls the virtual machine layer 220 to produce system state snapshots of the guest machines 210A-n. Further references to the guest machines 210A-n can include any of the guest machines and can be referenced the numerical indicator 210. The state snapshot server 250/250′ is shown as either a separate snapshot server 250 running on different computational hardware than the guest machines 210, or can be a state snapshot server 250′ running on a guest machine 210. Further references to the state snapshot server 250, 250′ will refer to either configuration unless stated otherwise and will only use the numerical indicator 250. The state snapshot server 250 can also interface with the configuration management server 260/260′ or through user interface. Further references to the configuration management server 260/260′ will refer to either configuration unless stated otherwise and will only be indicated by the numerical indicator 260. A user interface can provide control for taking system state data, storing and analyzing. Communication with the storage system can be over network including but not limited to ethernet, SANs or over a dedicated communication channel. Communication with the virtual machine layer 220 can be over interprocess communication protocols such as but not limited to sockets for the case were the snapshot sever is running on a guest machine 210, or though a library or kernel call.

The configuration management server 260 is used to provide a process enforced system for making system configuration changes. The U.S. patent application Ser. No. 12/008,274, filed: Jan. 9, 2008, entitled “METHOD AND APPARATUS FOR PROCESS ENFORCED CONFIGURATION MANAGEMENT” which is hereby incorporated by reference in its entirety describes in detail a process enforced configuration management system. The configuration management system uses change process rules to control when and if a configuration change is made. The configuration management system 260 can execute on a server with separate hardware and communicate with the state snapshot server 250 over a network. Alternatively, the configuration management server 260′ executes on a guest machine. Communication with the state snapshot server can be through but not limited to common interprocess or network communications including but not limited to socket communication.

FIG. 3 is an illustrative embodiment of the process 300 for system state analysis. The process involves the controlling a virtual machine layer to take a system state snapshot and analyze the state data. The process begins at A. In step 310, associated with one embodiment of the invention, a configuration management system generates a configuration change ticket. The configuration change ticket will indicate changes to be made to a computational system 200. In the step 320, associated with one embodiment of the invention, the configuration change specified in the configuration change ticket is made. Upon completion of the configuration change, or in response an operator input, the step 330 is performed. In step 330, the virtual machine layer is controlled to take a snapshot of the state information for one or more guest machines 210. While step 330, taking a snapshot of state information, is described as occurring after a step 320, the configuration change, the invention contemplates these two step being performed in the opposite sequence.

FIG. 4 illustrates a block diagram of an exemplary computing device 400 configured as a System State Snapshot Sever implementing a method of taking a system state snapshot and analysis of a computational system. The snapshot server 400 can control a virtual machine layer 220 executing on computational hardware and either directly analyze the state information or access and analyze state snapshots stored on a storage system 412. The storage of the snapshots and the system state snapshot server code can be stored on separate devices or on the same device. For example, the state snapshot code could be stored on a tape local hard drive or CD-ROM and the state snapshots could be stored on a disk farm. Further, the computing device 400 is configured to communicate with other computational systems such as a configuration management server 260 or though a human interface such as a graphical user interface. The communication can be through a network, direct communication through a dedicated communication link, or through an operating system communication channel such as a socket. For example, if a computing device 400 is running on a guest machine, it can connect over a socket interface or other communication interface with a virtual machine layer. The computing device requests that a system state snapshot of one or more of the guest machines be generated. The snapshot information is then stored on a storage system 240 or is communicated directly to the state snapshot server 250 device for analysis. Additionally, the invention contemplates working with incremental file systems. Only the references to a sequence of incremental files need to be stored to form a snapshot of a file. In general, a hardware structure suitable for implementing the computing device 400 includes a network interface 402, a memory 404, a processor 406, I/O device(s) 408, a bus 410 and a storage device 412. The choice of processor is not critical as long as a suitable processor with sufficient speed is chosen. The memory 404 can to be any conventional computer memory known in the art. The storage device 412 can include a hard drive, tape, CDROM, CDRW, DVD, DVDRW, flash memory card or any other storage device. The computing device 400 can include one or more network interfaces 402. An example of a network interface includes a network card coupled to an Ethernet or other type of LAN. The I/O device(s) 408 can include one or more of the following: keyboard, mouse, monitor, display, printer, modem, touchscreen, button interface and other devices including remote systems. The state snapshot server application(s) 430 control the virtual machine layer 220 and analyze state snapshots that can be stored in the storage device 412 and memory 404. The state snapshotserver applications 430 are processed as applications which are typically processed data. More or fewer components shown in FIG. 4 can be included in the computing device 400. Additional processors, either distributed or not distributed, and additional storage can be incorporated.

In some embodiments, the state snapshot server application(s) 430 includes several applications and/or modules. In some embodiments, the state snapshot server application(s) 430 include an analysis module 432, a control module 434 and an configuration management interface and control module 436.

To analyze system state snapshots, the state snapshot server 250 controls a virtual machine layer 220 to take a snapshot of one or more of the guest machines. The snapshot is then also analyzed according to a policy. The policy can include analysis parameter for persistent state information and non-persistent state information. The non-persistent state information can include user and kernel physical memory. In some embodiments, via an external interface, a configuration management server 260 sends control messages to the state snapshot server 250 to take and analyze the system state snapshot. These control messages can be the result of a trigger. The configuration management server 260 can send information related to change process rules that can be used in the analysis of the system state snapshots. For example, the configuration management server 260 can have rules on who and when a configuration change can be made. For security or compliance reasons, there can be restrictions related to these parameters. This allows for configuration, compliance, and security verification to extend beyond just state information. This invention now enables the incorporation of the additional information of “who” and “when” a configuration change was made. For example, the verification can now include a check of who made a configuration change, the time that it was made, and that the change was properly authorized as specified in a configuration change ticket and at the time specified by the configuration change ticket.

In operation, the state snapshot server 250 controls the taking and analyzing of state snapshots of guest machines 210 through the control of the virtual machine layer 220 in a manner that has minimal effect on the guest machine 210 performance. As discussed above, the state snapshot sever 250′ can executed on one of the guest machines 210 or on a separate piece of computational hardware. The state snapshot server 250 receives a control input for the taking and analyzing of one or more snapshots from one or more of the guest machines 210. The source of the control can be a user input from a terminal, a graphics display device, or can be scheduled by the operating system to execute a system state snapshot at a specified time. Further, the system state snapshot server 250 can be configured to be controlled by another process, application or server such as a configuration management server 260. In response to the issuance of a configuration change ticket, the configuration management server 260 can change the configuration of one or more of the guest machines 210, associated files systems or the computational hardware 230 configuration. Subsequently an indication is communicated to the state snapshot server 250 to take and analyze one or more system state snapshots. Preferably, the state snapshot server 250 interfaces with a virtual machine layer 230 that implements incremental file storage. Thus, as described above, the system state can be saved with a minimal copying of files. Thus, the time to take and store a snapshot is minimized and computation processing time and storage bandwidth is minimized.

The state snapshot server 250 can communicate with the virtual machine layer 220 over any standard computer communication method. Communication with the virtual machine can be over network sockets, signals, or shared memory communication. A dedicated program or library can be added to the guest machine 210 for the state snapshot application or process to communicate with the virtual machine layer 220 for the configuration where the state snapshot server 250′ is running on a guest machine. The state snapshot server 250 indicates to the virtual machine layer 220 which state snapshots are to be taken. The state snapshot server 250 can also specify the scope of the snapshot data to be taken. For example, snapshots of only file data can be taken. In another snapshot, the snapshots could include kernel data to be analyzed.

The virtual machine layer 230 can communicate back to the state snapshot server upon completion of the state snapshot. The snapshot is then analyzed by the state snapshot server 250. The analysis can include checking the state of the system against a specified state configuration. The analysis check includes analyzing the system snapshot data according to a policy. As described above, the system snapshot data can include persistent and non-persistent data including physical memory data that include user and kernel state information. Further, the analysis can utilize information provided by the configuration management server 260. This information can include process change rules from which conflicts between the configuration change, configuration change rules and a compliance configuration can be compared. The results of the analysis can be stored on a storage system 240, displayed on a graphical display device, communicated to the configuration management server 260, or a combination thereof.

Claims

1. A method of system state analysis of a computational system comprising: specifying a scope of state information to be captured based on a selected analysis to be performed on captured state information, wherein the scope of the state information is to include at least kernel memory information if the selected analysis is a first type of analysis, wherein the scope of the state information is to exclude kernel memory information if the selected analysis is a second type of analysis, and wherein the scope of the state information is specified for one or more selected guest machines of a plurality of guest machines running on a virtual machine layer of a computational system;controlling the virtual machine layer to capture selective state information of the one or more selected guest machines running on the virtual machine layer, wherein the selective state information is within the specified scope of the state information; andanalyzing the captured selective state information according to the selected analysis, wherein the selected analysis includes a policy containing an attribute, and wherein the attribute is used to analyze a portion of the captured selective state information, wherein the second type of analysis is a compliance check to validate that the computational system includes a desired system configuration, and wherein the first type of analysis is to: validate kernel data structures in the kernel memory of a kernel have not been corrupted;validate the kernel is properly configured; andvalidate the kernel includes necessary components.

2. The method of claim 1, wherein the attribute of the policy comprises one of a parameter, a system characteristic, a range, or specific system-related data to be compared to at least a portion of the captured selective state information.

3. The method of claim 1, further comprising storing the captured selective state information within a storage system.

4. The method of claim 3, wherein the storage system is a distributed disk storage system.

5. The method of claim 1, wherein the controlling of the virtual machine layer is through a process executing on a single one of the guest machines.

6. The method of claim 1, wherein the controlling of the virtual machine layer is through a process executing on a second computational system.

7. The method of claim 1, further comprising issuing a configuration change ticket, wherein the controlling the virtual machine layer and the analyzing the captured selective state information are performed in response to the issuing of the change ticket.

8. The method of claim 7, wherein a configuration change specified in the configuration change ticket is implemented before the selective state information for the one or more guest machines is captured.

9. The method of claim 8, wherein the issuing of a change ticket includes change process rules and the analysis of the captured selective state information includes the change process rules.

10. The method of claim 1, wherein the controlling the virtual machine layer is triggered by at least one of a scheduled system state analysis, a change to a policy, and event triggers.

11. A system for analyzing the state of a system comprising: one or more guest machines;a storage system;a virtual machine layer interfaced to the one or more guest machines, wherein the virtual machine layer is configured to generate state information for the one or more guest machines;a system state snapshot server configured to: specify a scope of state information to be captured based on a selected analysis to be performed on captured state information, wherein the scope of the state information is to include at least kernel memory information if the selected analysis is a first type of analysis, wherein the scope of the state information is to exclude kernel memory information if the selected analysis is a second type of analysis, and wherein the scope of the state information is specified for at least one selected guest machine of the one or more guest machines;control the virtual machine layer to capture selective state information for the at least one selected guest machine;analyze the captured selective state information; andproduce a state analysis result, wherein the captured selective state information is within the specified scope of state information, wherein the captured selective state information is analyzed according to the selected analysis, and wherein the selected analysis includes a policy containing an attribute that is used to analyze a portion of the captured selective state information; anda first computational hardware unit configured to execute the one or more guest machines and the virtual machine layer and coupled to the storage system, wherein the second type of analysis is a compliance check to validate that the computational system includes a desired system configuration, and wherein the first type of analysis is to: validate kernel data structures in the kernel memory of a kernel have not been corrupted;validate the kernel is properly configured; andvalidate the kernel includes necessary components.

12. The system of claim 11, wherein the state snapshot server is configured to execute within one of the guest machines.

13. The system of claim 11, wherein the attribute of the policy comprises one of a parameter, a system characteristic, a range, or specific system-related data to be compared to at least a portion of the captured selective state information.

14. The system of claim 11, further comprising a second computational hardware unit, wherein the second computational hardware unit is configured to execute the state snapshot server.

15. The system of claim 14, wherein the storage system and the first computational hardware communicate over a storage area network.

16. The system of claim 11, wherein the system state snapshot server is configured to control the virtual machine layer to generate the selective state information upon the occurrence of at least one of a scheduled system state analysis, a change to a policy, and event triggers.

17. The system of claim 11, further comprising a configuration management server configured to communicate with the state snapshot server, wherein the configuration management server is configured to issue a configuration change ticket, and wherein in response to the issuing the configuration change ticket, the state snapshot server is configured to capture the selective state information from the at least one guest machine and analyze the captured selective state information.

18. The system of claim 17, wherein the system configuration change ticket indicates a system configuration change, and wherein the system configuration change is implemented before the state snapshot server captures the selective state information from the at least one guest machine and analyzes the captured selective state information.

19. The system of claim 17, wherein the configuration management system is configured with change process rules.

20. The system of claim 17, wherein the configuration management server uses the state analysis results in an analysis of configuration management changes.

21. The system of claim 20, wherein the configuration management ticket is modified to include at least part of the state analysis results.

22. One or more processor readable storage devices having processor readable code embodied on the processor readable devices for programming one or more processors to perform operations comprising: specifying a scope of state information to be captured based on a selected analysis to be performed on captured state information, wherein the scope of the state information is to include at least kernel memory information if the selected analysis is a first type of analysis, wherein the scope of the state information is to exclude kernel memory information if the selected analysis is a second type of analysis, and wherein the scope of the state information is specified for one or more selected guest machines of a plurality of guest machines running on a virtual machine layer of a computational system;controlling the virtual machine layer to capture selective state information of the one or more selected guest machines running on the virtual machine layer, wherein the selective state information is within the specified scope of state information; andanalyzing the captured selective state information according to the selected analysis, wherein the selected analysis includes a policy containing an attribute that is used to analyze a portion of the captured selective state information, wherein the second type of analysis is a compliance check to validate that the computational system includes a desired system configuration, and wherein the first type of analysis is to: validate kernel data structures in the kernel memory of a kernel have not been corrupted;validate the kernel is properly configured; andvalidate the kernel includes necessary components.

23. The one or more processor readable storage devices of claim 22, wherein the attribute of the policy comprises one of a parameter, a system characteristic, a range, or specific system-related data to be compared to at least a portion of the captured selective state information.

24. The one or more processor readable storage devices of claim 22, the operations further comprising storing the captured selective state information within a storage system.

25. The one or more processor readable storage devices of claim 22, wherein if the first type of analysis includes a system security check, the scope of the state information is to include a combination of the kernel memory information, user memory information, and file-related information.