US20040064543A1 - Software application domain and storage domain management process and method - Google Patents

Abstract

A system monitoring process for monitoring data flow includes an application level agent process that collects configuration data from a software application, the configuration data represents configuration dependencies between the software application and a storage system, and a storage system level agent process that collects configuration data from the storage system, the configuration data represents the configuration dependencies between the software application and the storage system.

Description

BACKGROUND
• This invention relates to managing software application domains and storage domains.[0001]
• Data storage systems receive, store, and provide access for data retrieval by software applications such as database applications or other types of software. For determining internal layouts of the storage systems, representative maps can be displayed to provide insight of the hardwire interconnections within the storage system. The representative maps may be shown as a graphical tree structure where tree branches represent the interconnections of the storage system.[0002]
SUMMARY
• According to an aspect of this invention, a system monitoring process for monitoring data flow includes an application level agent process that collects configuration data from a software application, the configuration data represents configuration dependencies between the software application and a storage system, and a storage system level agent process that collects configuration data from the storage system, the configuration data represents the configuration dependencies between the software application and the storage system.[0003]
• According to a further aspect of the invention, a method for providing monitoring of data flow in resources of an enterprise includes collecting configuration data from a software application that is a first resource of the enterprise, the configuration data represents configuration dependencies between the software application and a storage system that is a second resource of the enterprise, collecting configuration data from the storage system, the configuration data represents the configuration dependencies between the software application and the storage system, mapping the configuration dependencies from the software application and the storage system, and disseminating the mapped configuration dependencies to a user.[0004]
• One or more of the following features may also be included.[0005]
• An operating system level agent process may collect configuration data from an operating system, the configuration data represents configuration dependencies between the operating system and the storage system. The storage system level agent process may collect configuration data that represents configuration dependencies between components of the storage system. The configuration dependencies between the software application and the storage system may include a software data path. The configuration dependencies between components of the storage system may include a physical data path. The configuration dependencies between components of the storage system may include a logical data path. A mapping process for may map the configuration dependencies between the software application and the storage system, and a visualization process may disseminate the mapped configuration dependencies to a user. The storage system may include a host server. The storage system may include a switch. The storage system may include a storage device. The storage system level agent process may collect configuration data from a table stored in the storage system.[0006]
• One or more advantages can be provided from the above. By monitoring data paths of software applications and storage systems respectively included in a software application domain and storage domain, complete maps of the data paths may be created that include both the software application domain and the storage domain along with the components that reside in both domains. Further, by stitching and mapping the components of both the application and storage domain, the components may be included in a comprehensive risk analysis, performance evaluation, and modeling and simulation for system upgrading. Thus by monitoring the components included in the application and storage domain, characterization measures that include system performance, availability, recoverability, cost and security may be computed for the complete combination of the two domains or a portion of the two domains. Additionally, by constraining access to particular components and data paths in the storage domain, data storage efficiency may be increased while still providing the appropriate storage capacity.[0007]
DESCRIPTION OF DRAWINGS
• FIG. 1 is a block diagram depicting an arrangement between software application and storage domains, and a domain management process.[0008]
• FIG. 2 is a block diagram depicting an illustrative example of the arrangement in FIG. 1.[0009]
• FIG. 3 is a block diagram of domain management process, agent, and a software application.[0010]
• FIG. 4 is a block diagram of information passed to a stitching process and information processed by the stitching process.[0011]
• FIG. 5 is a flow chart of a stitching process.[0012]
• FIG. 6 is a block diagram of another example of a software application domain and a storage domain, and a domain management process.[0013]
• FIG. 7 is a flow chart of a characterization process.[0014]
• FIG. 8([0015]a)-(c) represent five characterization measures.
• FIG. 9 represents five characterization measures as a function of time.[0016]
• FIG. 10 is a flow chart for computing characterization measures.[0017]
• FIG. 11 is a flow chart for assigning downtime business costs.[0018]
• FIG. 12 is a flow chart for a downtime business cost risk analysis process.[0019]
• FIG. 13 is a flow chart for a single point of failure risk analysis process.[0020]
• FIG. 14 is a flow chart for a data path redundancy risk analysis process.[0021]
• FIG. 15 is a flow chart for a configuration compatibility risk analysis process.[0022]
• FIG. 16 is a flow chart for a server binding process for constraining a storage system.[0023]
• FIG. 17 is a flow chart for a device zoning process for constraining a storage system.[0024]
• FIG. 18 is a flow chart for a masking process for constraining a storage system.[0025]
• FIG. 19 is a representation of a graphical user interface of a domain management process.[0026]
DETAILED DESCRIPTION
• Referring to FIG. 1, a[0027]domain management process10 provides end-to-end enterprise-wide data management between asoftware application domain12 and astorage domain14. Thesoftware application domain12 includes one or more software applications and an operating system, which are individually in communication with thedomain management process10 and are stored and executed on components included in thestorage domain14. Thedomain management process10 is a dedicated management station and serves multiple purposes and can be clustered for added redundancy and load balancing as well as for higher availability of thestorage domain14. Since thedomain management process10 is in communication with both domains, the management process collects configuration data from bothdomains12,14 to provide a complete representation of software data paths between thesoftware application domain12 and thestorage domain14 along with physical and logical data paths between the components of the storage domain. Physical data paths represent the physical connections between two components in thestorage domain14. For example a server, located in thestorage domain14, can be connected to a storage device, also located in the storage domain, by a network connection that provides ten potential physical data paths between the devices. However, only one or two of the ten paths may be software configured to actively pass data between the server and the storage device. These one or two software configured paths are logical data paths and are typically a subset of the physical data paths. By collecting the configuration data from both domains, thedomain management process10 examines global and local configuration dependencies across thesoftware application domain12 and thestorage domain14. In some arrangements, an operating system is included in thesoftware application domain12 and thedomain management process10 collects configuration information from the operating system to examine such dependencies among thesoftware application domain12, thestorage domain14, and the operating system. Also in some arrangements, switches and their associated switch ports that can be included in thestorage domain14 are not included in a logical data path.
• The[0028]domain management process10 provides a variety of functions and capabilities, such as proactively configuring and monitoring the software application andstorage domains12,14. Thedomain management process10 provides a unified virtualization configuration while monitoring interfaces of components of thestorage domain14 and their interactions with thesoftware application domain12. Thedomain management process10 also provides functionality for actively isolating and avoiding faults within the application andstorage domains12,14. Theprocess10 also provides a common interface to each application included in thesoftware application domain12 and the components of thestorage domain14. In addition, thedomain management process10 generates and provides a user with various activity and performance reports of the two domains interacting along with reports of the individual performance of the domains and the included components.
• [0029]Application domain12 includes one or more software applications e.g., mail servers and information management software such as relational database products. Thestorage domain14 includes components such as host servers and switches along with one or more storage subsystems that include network-attached storage (NAS) devices (e.g., a hard drives), which are typically assigned a permanent network address, and storage area network (SAN) subsystems that interconnect data storage devices and are typically part of an overall network of computing resources for an enterprise. The storage domain may also include the NAS subsystems, SAN subsystems, and other storage devices individually or in combination.
• In some arrangements, fibre channel (FC) technology may be used for transmitting data between the[0030]domains12,14 at high data rates which is especially suited for connecting computer servers to shared storage devices and for interconnecting storage controllers and drives. Some arrangements may also use fibre channel over an internet protocol (FCIP or FC/IP, also known as fibre channel tunneling or storage tunneling) that enables transmissions of fibre channel information by tunneling data between storage area network facilities over IP networks to facilitate data sharing over a geographically distributed enterprise. Thestorage domain14 may also include FC or IP virtualization engines, host-based storage volume managers, FC or IP switches (with or without virtualization), tape devices, or other similar storage associated mechanisms.
• Referring to FIG. 2, the[0031]domain management process10 is contained within thesoftware application domain12 and monitors the configuration of the software application domain, thestorage domain14 and the data paths16a-fbetween and within the domains. In this particular example, thesoftware application domain12 includes an email application18 (e.g., MS Exchange) and asales database application20 for a corporation enterprise. Thesoftware application domain12 also includes anoperating system22 that manages operations of ahost server24 that executes theemail application18 and thedatabase application20. In this particular example the host server is located in thestorage domain14.
• In order to map the layout of the data paths[0032]16a-fbetween and within bothdomains12,14, information including configuration data is collected by agents26a-dfrom each of theapplications18,20 and theoperating system22 in the software application domain. For example, oneagent26bconnects to theemail application18 and anotheragent26aconnects to thedatabase application20. Eachagent26a,26bprovides data that represents the association between the respective application and connectedsoftware data paths16a,16b,16c. Therespective agents26a,26balso provide information representing the associations between data structures within the portion of the application that the agent is assigned. In this particular example, oneagent26bis assigned to the mailbox portion of theemail application18 that includesmailboxes28 for all the employees of the corporation. One particular subset of themailboxes28 may be thecorporate executive mailboxes30 that contains and provides quick access to the email messages associated with the executives. Also in this particular example, the corporate executive email subset may also be stored in agroup storage32 that may store other subsets and is passed from theemail application18 to thestorage domain14 in order to save the contents (e.g., messages) of the mailboxes on a storage device included in the domain.
• The[0033]agent26bis in communication with theemail application18 and collects information that describes the associations between theemail structures28,30,32 and the associations between the application and thesoftware data path16ato thehost server24 in thestorage domain14. This collected association information is transmitted to thedomain management process10. Thedomain management process10 uses this association information along with information from other agents and storage devices to map the relationships between themailbox structures28,30,32, theapplication18, and the twodomains12,14. Similarly theagent26ais in communication with the corporationsales database application20, and collects configuration information from the database application associated with sales application data structures and collects information associated with connections between thesoftware data path16c, the application, and thehost server24. Thedomain management process10 uses this information to map the associations between the twodomains12,14.
• The[0034]operating system22 provides commands to transfer data between thehost server24 and other portions of thestorage domain14 and has information that is collected by anagent26c. Thisparticular agent26cgathers host-generic data from theoperating system22 and submits the information to thedomain management process10 for processing along with the information from theother agents26a,26b. Besides describing thesoftware data path16bbetweenoperating system22 and thehost server24, the information also contains management data attributes such as a host file system configuration for both SAN and NAS file systems, host server usage statistics, host server adapter information (Fibre Channel, ISCSI and SCSI), along with the particular version of theoperating system22.
• Information describing the associations between the components included in the[0035]storage domain14 is collected and transferred to thedomain management process10 for processing with the information collected from thesoftware application12 domain. The components of thestorage domain14 provide a storage hierarchy and structure for storing the data passed from theemail application18 and thedatabase application20 to thestorage domain14 for storage. In typical operations, data passes over thesoftware data paths16a,16cfrom therespective applications18,20 to thehost server24 and thehost server24 determines the storage location and the particular data path for transferring the data. For storing, the data is transmitted from thehost server24 to aswitch34 that directs the data, in this particular example, to either a network-attached storage (NAS)hard drive36 or aSAN38 that includes two redundant array of independent disks (RAID)40,42 and atape storage system44. In general,switch34 provides connectivity for thehost server24 and the storage devices and the switch may range from entry-level to enterprise-level depending on the business requirements. In some arrangements two or more interconnected switches may be used to produce a SAN fabric that provides optimum SAN performance, scalability, and availability. In addition,SAN38 may support disk mirroring, backup and restore, archival and retrieval of archived data, data migration from one storage device to another, and the sharing of data among different servers on a network. Also, in some arrangements SANs can incorporate subnetworks with NAS systems. SANs, such asSAN38 enable more efficient use of storage capacity by consolidating widely distributed disk space onto fewer storage arrays that are accessible by heterogeneous servers across the SAN. This improves storage resource utilization by allowing excess storage capacity from one application to be used for another. TheSAN38 can provide better return on capital expenditure for storage and servers, greater configuration flexibility, and reduced management costs through centralization.
• As mentioned, agents[0036]26a-cgather and transmit configuration data from therespective software applications18,20 and theoperating system22 to thedomain management process10. Anagent26dis also assigned to thehost server24 for gathering associations between the software data paths16a-c, which connect thehost server24 to theapplication domain12, and respectivelogical data path16d(that is also a physical data path), which connects the host server to thecomponents34,36,38 included in thestorage domain14. Similar to the other agents26a-c, after collecting the configuration information, the information is passed to thedomain management process10 for processing to determine associations with the information gathered from the other agents26a-c. However, some components of thestorage domain14 such as theswitch34, theNAS storage36, and theSAN38 are not agent assignable for passing configuration information, such as thelogical data paths16e,16f(that are also physical data paths) to thedomain management process10. To provide configuration information for thesestorage domain components34,36,38, each component includes a respective table46,48,50 that contains information for transferring to thedomain management process10. The respective tables46,48,50 contain information such as identification numbers for therespective storage components34,36,38 and their interconnections. For example, the table46 associated with theswitch34 provides the world wide name (WWN) assigned to the switch (i.e. WWN33) and the world wide name associated with the component connected to each port of theswitch34. In this particular example,port152 is connected to WWN30 (i.e., host server24),port254 is connected to WWN35 (i.e., NAS drive36) andport356 is connected to WWN32 (i.e., SAN38). World wide names are 64-bit identifiers that are used to uniquely identify ports on a network during discovery and network initialization. With this WWN assignment information of theswitch34 and each WWN connected to each respective switch port, thedomain management process10 can map the associations of theswitch34 without collecting information from an agent. Also, since the table48 included in the NAS drive36 provides the WWN of the drive (WWN35), along with the WWN of the connected devices, and the table50 included in theSAN38 provides the WWN of the SAN (WWN32), along with connections, thedomain management process10 can associate the WWN's provided by the switch table46, the NAS drive table48 and the SAN table50.
• In some arrangements the[0037]storage domain14 can contain one or more sub-domains that include the components of the storage domain. For example, thehost server24 can be included on in a host domain (not shown) or theswitch34 can be included in a switch domain (also not shown). Further, theNAS drive36 and theSAN38 can be included, for example, in a subsystem domain (not shown).
• Also in some arrangements the[0038]domain management process10 may include reception processes (not shown) that are in communication with respective agents26a-dto which the reception processes are assigned. In some arrangements the reception processes can also be assigned to particular sub-domains (e.g., host domain, switch domain, subsystem domain, etc.) to communicate with one or more agents in the respectively assigned sub-domain. These reception processes can also be assigned to the components in thestorage domain14 that are not agent-assignable (e.g., switch34, NAS drive36, SAN38). Once assigned to one or more of these components, the reception processes can receive information (e.g., tables46,48,50) from the components. Thedomain management process10 may also produce and maintain one or more tables, from the information collected, to identify the components included in thestorage domain14. The tables may also maintain a listing of the connections and associations between the components in the storage domain or sub-domains.
• In some arrangements the[0039]domain management process10 may store a portion of the information collected from the agents26a-dand thestorage domain components24,34,36,38 in an information repository (not shown) that may reside in thedomain management process10 or on one or more of the components in thestorage domain14. This information can also be stored in one or more tables and retrieved from the tables to determine the software, physical, and logical data paths among components of thesoftware application domain12 and thestorage domain14.
• In some arrangements host bus adapters (HBAs) connect the[0040]host server24 to theswitch34 and each HBA is often assigned a unique WWN by the manufacturer that is encoded in non-volatile memory of a HBA controller included with each HBA. HBAs typically include hardware and drivers that interface with operating systems to represent fibre channel storage as devices in the operating system. HBAs also often negotiate with one or more connected switches along with other devices attached to a network and provide processing capabilities that minimize CPU overhead on thehost server24.
• Referring to FIG. 3, the[0041]agent26bassociated with the mailboxes of theemail application18 is shown in communication with the application and thedomain management process10. In general theagent26bis a program that gathers information from theemail application18 mailboxes or performs other services typically as a background type of process. Theagent26balso monitors different aspects of the entity under its control, assesses the status of the entity based on information collected, and uses one or more predefined management policies or rules in order to determine which action(s) to perform based on the policies. Typically, theagent26buses parameters provided by the user and gathers information from the assigned application and presents the information on a periodic basis to thedomain management process10. As mentioned theagent26bis responsible for one type of application object (i.e., the email mailboxes). Often agents are implemented in Java and are supported by the code of the application to which the agent is assigned.
• As mentioned the[0042]email application18 includesobjects28,30,32. Once theagent26bis assigned to the email application18 acollection process60 directs theagent26bto collect data from theemail application18. In particular, thecollection process60 directs acollector62, included in theagent26b, to determine what configuration data should be collected from theemail application18 while avoiding unnecessary data collection. Once thecollection process60 determines the data sought, thecollector62 is directed to transmit arequest64 to theemail application18 and receive aresponse66 that contains the data. In some arrangements,collector62 examines a boolean parameter present in theemail application18, which changes logic level when data is present for collection. Typically thecollector62 has a short execution period so that execution is completed within the predetermined cycle time of thecollection process60.
• After the[0043]application response66 is received by thecollector62, information is passed to anassessment process68 that is responsible for assessing aspects of the assignedemail application18, recording the assessment results in a database, and passing assessment results on to anassessor70. Most of the business logic resides in theassessment process68. In some arrangements business logic prioritizes events associated with the assigned application based on business importance. For example if storage is limited, the business logic in theassessment process68 may determine that theexecutive mailboxes30 are to be stored in the limited storage while the mailboxes for all theemployees28 is not to be stored because they are less important to the business. In some other arrangements the business logic may be based on other principals such as design principals. For example, theassessment process68 may determine to only store the messages of theexecutive mailbox30 because the large volume of messages for all theemployees28 may create potential storage problems. Theassessment process68 is controlled by a hierarchy of thresholds and switches in a policy system (e.g. business logic) and the assessment process can use various implementations to assess the particular responses from theemail application18. Once theassessment process68 has determined if the assigned application is to be reconfigured, assessment results are stored in anassessor70, which is included in theagent26b. Theassessor70 then provides the assessment results to anapplication management process72. Theapplication management process72 uses the assessment results to direct anapplication manager74 to alter the configuration of theapplication18. For example, theapplication manager74 can alter the configuration of theemail application18 such that the email messages from theexecutive mailboxes30 are stored in a different storage device. In another example, theapplication management process72 may direct the application manager to alert (e.g., send an email message, send a page, etc.) a system administrator of theapplication18 status. In some arrangements, theassessment process68 is constrained to execute relatively quickly, use inputs only provided by thecollector62, and store assessment results in theassessor70.
• In another arrangement, the domain management process[0044]10 (shown in FIG. 2) can include a policy manager process (not shown) that defines the policies and rules that are used to determine if an event associated with an application in communication with thedomain management process10 has occurred. An event is detected if one of the policies or rules has been violated or possibly not violated. Policy filters (also not shown) may be included in the policy manager process to determine which components of thedomains12,14 (shown in FIG. 2) should have the policies and rules applied. These policies and rules can be stored in a database in an accessible format such as an XML file. The policy manager process may also execute the policies and rules on a pre-determined schedule, on user demand, on receipt of an event, or other similar process.
• Once an event has been detected and possible action taken, the policy manager process can report the event to the user. These events can be generated based on a component of one of the[0045]domains12,14 and typically the event is stored in an event repository. For example an event can be to notify the administrator to replace a component due to failure of a subsystem drive, fan, or I/O card. In another example, an event can alert the system administrator that a port on a particular component has failed. Thedomain management process10 may also be the basis to generate the events. For example, thedomain management process10 may determine that a certain resource or component has reached a dangerous level (e.g., the storage capacity of a storage device is nearly filled) or the process may determine that a particular agent is not functioning correctly. Also, once an event has been detected, the event may be stored and its association with other events tracked over a period of time.
• The[0046]domain management process10 is responsible for execution control of all agents, including theagent26b, and periodically cycles through the agents. Thedomain management process10 is also responsible for calling thecollection process60 relatively soon after calling the agent'sassessment process68. Because thecollection process60 may check a database to determine what data may have already been collected, it can be more efficient for thedomain management process10 to invoke all the collection processes for a managed application instance before calling any of the assessment processes. Thedomain management process10 also receives reconfiguration requests from theassessor70 so that theapplication management process72 is executed to direct theapplication manager74 at the appropriate times and in the appropriate priority order. Some application managers may run for hours through many collection and assessment cycles. Also among the responsibilities of the domain management processes10, responsibilities include: invoking each agent's collector once each cycle, invoking each agent's assessor once for each cycle, invoking each agent's collector before its assessor, scheduling each application manager as requested, not overloading instances with application managers, not missing collection or assessment because of running application managers, maintaining regular collect/assess cycle timing, and running application managers in the order defined by the priorities established by the assessors.
• The[0047]assessor70 identifies issues from the assessment results and schedules application managers to deal with them. To identify the issues, theassessor70 may be organized with conditionals to determine if particular parametric actions should be taken. Theassessor70 also decides which alternative actions to take, when to take them, and with what parameters. In some arrangements, theassessor70 may include a process to diagnosis the issues and a problem solver process to define the course of action. However, as shown in the figure, theassessor70 may be a single entity and may be used in solving problems such as performance problems with no obvious cause. In some arrangements, the reconfiguring by theapplication manager74 may include scripted sequences of actions, however, for more difficult problems the script may be generated by theassessor70.
• Since this[0048]particular agent26bis assigned to themailboxes28,30,32 of theemail application18, the agent would typically collect the current free memory space from the mailboxes and alert an email administrator if the memory space is nearing capacity. Other separate agents may also be associated with theemail application18 and, for example, alert the administrator that a public folder store is nearing capacity, or when a particular service is being executed by theemail application18. One or more agents may also scan the mailboxes and public folder stores for corrupt messages or files and prompt the email administer to delete the corrupted items or allow them to remain. An agent associated with theemail application18 may also help collect vital information from one or more servers, such as CPU utilization, memory utilization, and whether the server is online or down. An agent may also aid in the evaluating of storage performance. For example, an agent may monitor the rate at which storage is growing, or measure the time period until an overall email store reaches capacity, or even measure the time period for data to transfer from the host server24 (shown in FIG. 2) to a storage device in the storage subsystem.
• Referring to FIG. 4, the[0049]information80 collected by the agents26a-d(shown in FIG. 2) and included in the tables46,48,50 (also shown in FIG. 2) of the components of thestorage domain14 is represented. For example, themailbox hierarchy82 collected by theagent26bassociates with the mailboxes of theemail application18. Also, aname84 identifying the executing sales data base was collected by theagent26aassociated with the corporation salesdata base application20. The O/S version86 was collected by theagent26cassociated with theoperating system22 and theagent26d assigned to thehost server24 provides the label (i.e., “C:\”), also known as a mount point, and world wide name88 (WWN30) of the host server along with the WWN (i.e., WWN33) of the device connected to the host server. The respective tables46,48,50 from theswitch34, theNAS drive36, and the SAN38 (each shown in FIG. 2), provide labels and the worldwide numbers90,92,94 assigned to the three storage domain components along with the respective WWNs of connected components.
• This collected information is transmitted to a[0050]stitching process81, which may reside in thedomain management process10. Thestitching process81 determines the software data paths between the software application domain12 (shown in FIG. 2) and the storage domain14 (also shown in FIG. 2) and the software data paths within the application domain. Thestitching process81 also determines the physical and logical data paths extending within the storage domain. Since theemail application18 and the sales database application20 (also shown in FIG. 2) are executed on thehost server24, software data paths connect the applications and the server. Also, since thestorage domain components24,34,36,38 (shown in FIG. 2) store the WWNs of connected components, physical data paths between the storage domain components, which are also logical data paths in this example, may be determined. For example, theinformation90 associated with the switch provides thatswitch port1 is connected to a device withWWN30,switch port2 is connected to a device withWWN35, and switchport3 is connected to a device withWWN32. So, from this information, and the labels associated with the information for the other components in the storage domain, thestitching process81 can determine thatswitch port1 is connected to host server “C:\”24,switch port2 is connected to theNAS drive36, and switchport3 is connected to theSAN38.
• Once the[0051]information80 is received by thestitching process81, the information is processed into amap96 of the software, physical, and logical data paths between each of the components in the software application and storage domains. By stitching the appropriate interconnections of both domains, themap96 may be displayed by thedomain management process10 to a user for analyzing the interconnections and, for example, to determine replacement component positions, insertion positions for new components, or re-arranging the currently displayed data paths. As shown the software data paths98a-care represented with dashed-lines between the applications, the operating system, and the host server “C:\”. Logical data paths99a-care represented with respective solid-lines and extend between the host server, the brocade switch, the NAS disk drive, and the SAN. By stitching amap96 of the data paths of both domains the user is provided a complete layout of the applications and storage system components and may use the stitched information to model and analyze the addition of new components to the domains or the replacement of current components and measure the performance of the altered domains.
• In some arrangements, the[0052]stitching process81 may receive information associated with software, logical, or physical data paths to an unknown device. In such as case, a generic device (not shown) is assigned, thestitching process81, to the particular data path.
• Referring to FIG. 5, a[0053]process100 for stitching data paths between the software application domain12 (shown in FIG. 2) and the storage domain14 (also shown in FIG. 2) is shown. Theprocess100 starts102 and retrieves104 information that was collected by agents assigned to software applications and an operating system that are present in thesoftware application domain12. This information can include an identifier for each application and the operating system and also identifies the server that executes the applications and operating system. Thestitching process100 retrieves106 information that was collected from the components of thestorage domain14. As shown in FIG. 2, these components may include a host server, a switch, and one or more storage mechanisms such as a NAS drive or a SAN. The retrieved information can include the WWN of each component along with any lists that report WWNs of connected components. After retrieving the storedinformation104,106, thestitching process100 determines108 software data paths between the applications, the operating system, and the components of the storage domain. After determining108 the software data paths, thestitching process100 determines110 the physical and logical data paths between the storage domain components by comparing the WWN information collected from the components. Once the physical and logical data paths have been determined108,110, the stitching process maps112 the software, physical, and logical data paths among the applications, operating system, and storage domain components. After mapping112 the data paths, thestitching process100displays114 the map to a user for analyzing the data paths between the software application domain and the storage domain. Once the map is displayed130 to the user, thestitching process100 stops116.
• Referring to FIG. 6, in another example of the two domains, which includes more storage components, the corporation[0054]sales database application20 is now executed on a secondhost server #2120, and anagent26eis assigned tohost server #2 and provides thedomain management process10 with information regarding the data paths between the database application and the second host server along with the data paths between the second host server and the other components in thestorage domain14. Bothhost server #124 andhost server #2120 have respectivephysical data paths16d,16g,16h,16ithat respectively lead to switch #134 and a second switch (#2)122. Also switch #134 andswitch #2122 are connected to theNAS drive36 and theSAN38 by respectivephysical data paths16e,16f,16j,16k. All of thesedata paths16d,16g,16h,16i,16e,16f,16j,16kallow bothhost servers24,120 to pass data across eitherswitch #134 orswitch #2122 to either the NAS drive36 or the twoRAIDs40,42 or thetape drive44 of theSAN38. Similar to FIG. 2, each of thephysical data paths16d-kare configured to be logical data paths. Also similar to FIG. 2, the software, logical, and physical data paths may be determined by thedomain management process10 from the information provided by theagents26a,26b,26d,26eand the information provided byswitch #134,switch #2122, theNAS drive36, and theSAN38. Once the respective information is collected, the data paths can be stitched by the stitching process100 (shown FIG. 5) that is included in thedomain management process10 and a map containing the components of thesoftware application domain12 and thestorage domain14 may be displayed to the user.
• Once the map is displayed, the user can select a particular entity in either domain, a particular data path, a portion of one or both domains, or both entire domains for analysis. In this particular example, a data path from the[0055]email application18 to theRAID #040 is shown (by bold lines) to have been selected. Thesoftware data path16aextends from theemail application18 to thehost server #124 that executes the application. Alogical data path16gthen extends from theserver24 to switch #2122 and the selected path then extends across thelogical data path16kto theSAN38 and terminates at theRAID #040. By selecting this particular path a user can have the selected data path characterized, by thedomain management process10, while taking into account both theemail application18 in thesoftware application domain12 that initiates the selected data path and the components (i.e.,host #124,switch #2202,SAN38, andRAID #040) of thestorage domain14 that are included along the user selected data path.
• To analyze the selected data path, the[0056]domain management process10 includes a characterization process for computing characterization measures from the information passed by theagents26a,26b,26d,26eand thestorage components34,36,38,202 without agents. For example, a performance measure may be calculated to determine the selected data path throughput (i.e., amount of data moved in a time period in mega-bits per second) from theemail application18 to theRAID #040. In some arrangements performance may also be a measure of the total effectiveness of the selected data path, including throughput, individual response time, and availability.
• Another characterization measure that may be calculated is the availability of the selected data path. Availability can be defined as the accessibility of input and output ports and is used to describe products and services that ensure that data continues to be available at a required level of performance in situations ranging from normal through “disastrous”. In general, availability is achieved through redundancy involving where data is stored and how it can be reached. Often availability can be measured in terms of how often data is available. For example, a data path may provide 99.999 per cent availability, or in another example, availability may be measured by the amount of data flow through the data path at a particular time (e.g., 3200 megabytes per second).[0057]
• By measuring recoverability of the selected data path, the connected storage device is characterized for the capability to retrieve stored data. In this particular example, the[0058]RAID #040 recoverability may be determined by thedomain management process10 from a manufacturer error rate that is provided by theRAID #0. Cost may also be measured from the information passed to thedomain management process10 regarding the selected data path. Often when an application malfunctions or is shut down, a downtime cost in dollars per hour is associated to the application. For example, if theemail application18 for the corporation malfunctions and is brought off line, the corporation may experience a loss of $50.00 per hour due to this application downtime. Or in another example, if thesales database application20 malfunctions and is placed off line, the corporation may experience a loss of $500.00 per hour. Also, these downtime costs in dollars per hour cost for a malfunctioning application can vary as a function of time. For example, during the holiday season the downtime cost of thesale database application20 may greatly increase, for example, to $10,000 per hour while the downtime cost of corporation email application may decrease to $20.00 per hour due to the number of employees on vacation. So, downtime cost, which is calculated by thedomain management process10 for the selected data path, may vary over a particular time of year.
• Another characterization measure that may be calculated by the[0059]domain management process10 for a selected data path is the security level. In this particular example, information passed fromSAN38 to thedomain management process10 contains a measure of the susceptibility that data stored onRAID #040 may be improperly retrieved without proper authorization. In characterizing the security level, access to thedomains12,14 or portion of the domains may be dependent upon the role of the domain management process user (e.g., the company CIO, different administrators, help desk employees). The access to thedomains12,14 or portions of the domains may also be limited by the organization to which the user is a member. In some arrangements, besides limiting access based on users, access can also be based on thedomains12,14, portions of the domains, or components of either domain (e.g., a particular application in the software application domain, a host server in the storage domain, etc.). Access may be limited based on a component, such as, only allowing a user to view a component on a display or allowing access such that user can access the characteristics of the component or execute an analysis on the component, or other similar control function.
• So the[0060]domain management process10 may compute, for example, a performance, availability, recoverability, cost, and security measure to characterize the selected data path. These characterization measures may also be calculated for any other similarly selected data path, or portion of thesoftware application domain12, thestorage domain14, or a combination of both domains.
• Referring to FIG. 7 a[0061]characterization process130 that may be included in the domain management process10 (shown in FIG. 6) is shown. Thecharacterization process130 starts132 when a user selects134 a portion of the software application domain12 (shown in FIG. 6), the storage domain14 (also shown in FIG.6), or a combination of the two domains. After theuser selection134, theprocess130 accesses136 the appropriate information that was collected from the agents and storage components of the software application andstorage domains12,14 associated with the portion selected by the user. Referring briefly to FIG. 6 as a demonstrative example, the information accessed by theprocess130, for this particular user selection, would relate to theemail application18,host server #124,switch #2122, and theSAN38. Returning to FIG. 7, once the appropriate information has been accessed136, theprocess130 computes138 the performance measure, theavailability measure140, therecoverability measure142, thecost measure144, and thesecurity level measure146 associated with the user selection. After the characterization measures have been computed, theprocess130reports148 the computed measures to the user prior to stopping150.
• Referring to FIGS.[0062]8A-8C three sets of horizontal bar charts are shown that may be provided by thedomain management process10 to the user to represent the performance, availability, recoverability, cost, and security measures calculated by the process130 (shown in FIG. 7) for the selected data path shown in FIG. 6. For each of the three sets of bar charts, the range of each bar chart increases as each chart extends from left to right and each chart is scaled in the appropriate units associated to the particular measure. For example, the performance measure may be scaled in units of megabytes per second while the cost measure may be scaled to units of dollars/hour for downtime.
• Referring to FIG. 8A, after each of the five characterization measures are computed and displayed on the respective bar chart[0063]200a-e, the user can alter the computed value of one characterization measure and the management domain process10 (shown in FIG. 2) will re-compute the values for the other four characterization measures based on the altered value. In this particular example the user has increased the performance measure from onevalue202 to ahigher value204. Based on thishigher value204 and the information associated with the selected data path (shown in FIG. 6), themanagement domain process10 re-calculates the values for the other four measures and displays the values on therespective bar charts200b-e. For example, for the selected data path to operate with a performance at thehigher value204, theminimum availability measure206 is re-computed and displayed on theavailability bar chart200b. Correspondingly, the re-calculatedminimum recoverability measure208,minimum cost measure210, and theminimum security measure212 are displayed on therespective bar charts200c-e. However, the re-computed values displayed the fourbar charts200b-eare minimum values and thehigher performance value204 may also be attained by operating the selected data path with higher values then the reported on the fourbar charts200b-e.
• Referring to FIG. 8B, after the domain management process[0064]10 (shown in FIG. 6) calculates the characterization measures associated with the selected data path (also shown in FIG. 6), the user can simulate an insertion, into the selected data path, of one or more applications in the software application domain12 (shown in FIG. 6) or one or more components in the storage domain14 (also shown in FIG. 6). For example, as shown in the figure each bar chart220a-edisplays the respectivecalculated values222,224,226,228,230. In this particular example, the user uses thedomain management process10 to simulate insertingRAID #142 into the selected data path (shown in FIG. 6). Due to this simulated insertion, the performance of the selected data path, along with the other characterization measures, increase with respect to measures of the selected data path containing onlyRAID #040 (also shown in FIG. 5). Correspondingly, thedomain management process10 re-calculates the characterization measures to include the information associated with the newly insertedRAID #142 and the recalculated measures are displayed on the respective bar charts220a-e. In this particular example, each of the characterization measure values increase and are shown on the bar charts by respectivesolid bars232,234,236,238,240. Also, in this particular example the insertedRAID #142 was previously included in the storage domain14 (as shown in FIG. 6), however in other examples the simulated components may not be previously included in either domain.
• Referring to FIG. 8C, the domain management process[0065]10 (shown in FIG. 6) may also be used to simulate replacing components included in either domain with different components to alter a selected data path. For example, referring briefly to FIG. 6, thedomain management process10 may simulate replacingswitch #2122 with a lower performance switch. Once the user selects such a simulated switch replacement, thedomain management process10 recalculates the characterization measures and displays the new characterization measure values on the bar charts in comparison to the measures computed prior to the component replacement. Prior to replacing the switch, each of the characterization measure values252,254,256,258,260 are calculated and are shown with hashed bars on respective bar charts250a-e. After replacing the switch, the five characterization measures are recalculated and thenew values262,264,266,268,270 are displayed with solid bars on the respective bar charts250a-e. In this particular example, since the switch is replaced by a simulated low performance switch the five characterization measures decrease on the respective bar charts250a-e. Also in this particular example, a component in the storage domain14 (shown in FIG. 6) was replaced by a simulation, however in other examples simulations may replace one or more components in either the software application domain12 (also shown in FIG. 5) and/or the storage domain. Also, referring to FIGS.8A-C, each of the functions may be performed individually or in combination.
• Referring to FIG. 9, the domain management process[0066]10 (shown in FIG.6) can collect information from the software application andstorage domains12,14 (also shown in FIG. 6) to calculate the performance, availability, recovery, cost, and the security measures over a particular time period. In this particular example, each of the characterization measures are calculated over a calendar year and displayed on x-y line charts280a-efor the selected data path shown in FIG. 6. By displaying the characterization measures as a function of time, a user can determine peak activity periods along with slow periods and possibly adjust storage components and application assets to compensate for the activities during particular months. For example, as shown in the figure, theperformance measure280aof the selected data path, including theemail application18, has peak periods prior to the December holiday season and a slow period during the summer months. So, a storage system administrator may decide to use a high performance storage device for email mailbox storage for the peak usage time periods of the calendar year and switch in a lower performance storage device during the summer months when email traffic is in the lull period. Also, for example, by displaying the characterization measures the administrator can determine the proper time of year for general maintenance and upgrading of equipment in the particular data path selected by the administrator for analysis.
• Referring to FIG. 10, a[0067]characterization simulation process290 that may reside in the domain management process10 (shown in FIG. 6) is shown. Thesimulation process290 starts292 when a user selects294 a portion of the software application domain12 (shown in FIG. 6) or a portion of the storage domain14 (shown in FIG. 6) or a combination of the two domains. As previously mentioned, for example, this selection may include a particular data path from an application to a storage device, or a particular component in thestorage domain14, or the entire enterprise that contains both domains. After theuser selection294 is made, theprocess290 accesses296 and retrieves the appropriate information collected from the agents and storage domain components associated with the user selection. Upon retrieving296 the collected information, theprocess290 computes298 the five characterization measures (i.e., performance, availability, recoverability, cost, and security measures) based on the retrieved information for the user selection. After computing, the characterization measures are then displayed to the user for analysis. After computing and displaying298 the five characterization measures, theprocess290 determines300 if the user has changed one of the displayed characterization measures, such as was described with respect to FIG. 8A. If the user had changed one of the characterization measures, theprocess290 then re-computes302 the other four characterization measures and displays the re-computed measures to the user. If the user has not altered one of the characterization measures, theprocess290 determines304 if the user has inserted a simulated component into the prior user selection. If a simulated component has been inserted by the user, theprocess290re-computes302 the characterization measures based on the previously retrieved information and the insertion of the simulated component and displays the re-computed measures to the user. If the user has not inserted a simulated component into the selection, theprocess290 determines306 if the user has replaced a component or a portion of the user selection. If the user has replaced a component or portion of the selection, theprocess290recomputes302 the characterization measures based on the previously retrieved information and the replacement component and displays the re-computed measures to the user. If the user has not replaced a component in the selection, theprocess290 asks308 the user if the process should stop. If the user decides to end, theprocess290 is stops310. But if the user decides to continue the process, theprocess290 returns to computing and displaying298 the characterization measures based on the current information and continues from this point.
• Referring to FIG. 11, the information collected from both the application domain and the storage domain may also be used in risk analysis of the particular data path or other portion of the two domains selected by the user. By performing a risk analysis on the selection, potential risky configurations or operation environments may be detected. Again, briefly referring to FIG. 6, one possible selection is shown by the bold data path from the[0068]email application18 throughhost server #124,switch #2122, and ending inRAID #040 of theSAN38 for data storage. Various types of risk analysis may be performed on this selected data path. For example, a risk analysis may be based on downtime business cost of the selected data path. To perform this particular risk analysis, one or more downtime business costs are assigned tostorage domain14 components and applications in thesoftware application domain12.
• Returning to FIG. 11, an[0069]assignment process320 is shown that may be included in the domain management process10 (shown in FIG. 6) and that allows the user to assign one or more downtime business costs. Theassignment process320 starts322 by determining324 if the user is assigning downtime business costs to the host servers in the storage domain. Also referring briefly to FIG. 6, for this particular example, the user can assign the same or different downtime business costs to thehost server #124 andhost server #2120. If the user is assigning downtime costs to the host servers, theprocess320 assigns326 one or more downtime costs, provided by the user, to the respective servers and then stops344. If the user is not assigning a downtime cost to the server, theassignment process320 determines328 if the user is assigning business costs to the applications in the application domain12 (shown in FIG. 6). For example, the user may assign a downtime business cost of $50/hour to the email application18 (shown in FIG. 6) and a downtime cost of $500/hour to the sales database application20 (also shown in FIG. 6). If the user is assigning a downtime costs to one or more applications, theprocess320 assigns330 the costs to the applications and then stops344. If the user is not assigning a downtime business cost to the one or more applications, theprocess320 determines340 if the user is assigning a default downtime cost to all of the applications in the software application domain. For example, the user may assign a default downtime value of $10,000/hour to all the applications in the software application domain. If the user is assigning a default downtime business cost for all of the applications, theprocess320 assigns342 the default downtime business cost to each of the application and then stops344. Note that the business costs assigned to the servers and the business costs assigned to the applications are mutually exclusive and if a particular server is assigned a cost, any cost assigned to an application on the server is ignored for the risk analysis. Also, although the downtime cost uses here are described in units of dollars per hour, other units may be assigned and used in the risk analysis.
• Referring to FIG. 12, a downtime business[0070]cost risk analysis350 process that may be included in the domain management process10 (shown in FIG. 5) is shown. Therisk analysis350 processes starts352 by the user selecting354 a portion of the software application domain12 (shown in FIG. 6) and the storage domain14 (also shown in FIG. 6) for execution of the risk analysis. After theuser selection354 is made, theprocess350 accesses and retrieves356 the downtime business cost(s) assigned to the one or more host servers that are connected to any portion of the user selection. Upon retrieving356 the downtime business cost(s) of the connected host servers, therisk analysis process350 accesses and retrieves358 the downtime business cost(s) assigned to the one or more application(s) present on any data path included in any portion of the user selection. After retrieving358 the cost(s) for the one or more applications, therisk analysis process350 selects360 one of the elements (i.e., a switch, a SAN, etc.) included in the user selection and initializes a cost variable associated with the element to zero. Next, theprocess350 determines362 if the element is in any data path with a host server. If the element is in a data path with a host server, theprocess350 adds364 (i.e., aggregates) the downtime business cost of all the host servers on one or more data paths with the element to the element cost variable. After summing these one or more downtime business costs, theprocess350 determines366 if an application is in any data path with the selected element and if so, is there no downtime business cost assigned to the particular host server executing the application. If an application is in a data path with the element and no downtime business cost is assigned to the host server executing the application, then theprocess350 adds368 the downtime business cost, assigned to each application that meets this condition, to the element cost variable. If no application meets this condition, theprocess350 determines370 if another element is included in the user selection and has not been analyzed for downtime business cost. If there is another element included in the user selection, theprocess350 selects372 the next element, initializes the cost variable associated with the element to zero, and returns to determine362 if there are host servers connect to that particular element. If all the elements included in the user selection have been analyzed for downtime business cost, theprocess350 reports each element cost variable to the user for disseminating the cost analysis and stops376. For example a report may be generated by the process to provide the user with the particular costs used in the analysis, warnings associated with analysis (e.g., some servers or applications were not assigned downtime business cost), a list of the components prioritized by the calculated cumulative risk value ($/h), a list of the components prioritized by the calculated cumulative risk value and by application type, and a list of the applications associated with each component.
• Referring to FIG. 13, another type of risk analysis is a single point of failure analysis that monitors the capability of each data path originating from an application, in the software application domain[0071]12 (shown in FIG. 6), and identifies any non-functioning data paths or malfunctioning storage domain14 (also shown in FIG. 6) components associated with the application that will cause the application to malfunction due to an inability to access data. A single point offailure analysis process380, which may be included in the domain management process10 (also shown in FIG. 6), is shown. The single point offailure analysis process380 starts382 by the user selecting384 a portion of the software application domain12 (shown in FIG. 6) and the storage domain14 (also shown in FIG. 6) for execution of the risk analysis. After theuser selection384 is made, theprocess380 determines386 if one or more host servers are included in the user selection. If no host servers are included in the user selection, then the single point offailure process380 stops388. But, if one or more host servers are present in the user selection, theprocess380 determines390 if one or more applications are present on the one or more host servers included in the user selection. If an application does not reside on one or more of the host servers in the user selection, theprocess380 simulates392 the presence of an application on each host server where an application does not reside. After the application has been simulated392 or the process determines that each host server in the user selection has one or more applications, theprocess380 determines394 if a single point of failure is associated with each application that is present or is simulated on each of the host servers included in the user selection. If theprocess380 determines that there is one or more single points of failure each application, the process reports396 each of the single points of failure to the user and then stops388. If no single points of failure are associated with each application, then the process reports the absence of any single points of failure to the user and the process stops388. In some arrangements theprocess380 may also report any components in the user selection associated with any single points of failure.
• Another type of risk analysis may incorporate the functionality of the single[0072]failure point process380 and the downtime business cost analysis350 (shown in FIG. 12). In this particular risk analysis the components of a user selection are first prioritized based by downtime business cost. Second, the components are further prioritized based on whether multiple data paths are used for protecting against the event of data path failure. For example, if two switches are each included in a data path associated to the application, the downtime business cost assigned to each switch is reduced by 50%. So if the assigned application downtime business cost is $10,000, then the cost assigned to each switch is $5,000. In some arrangements the downtime business cost may be further reduced if the storage components are clustered. Clustering is the use of multiple storage devices, or other computing devices, and redundant interconnections, to form a single highly-available system to provide load balancing as well as high availability. So if a storage component is included in a cluster, the downtime business cost can be further reduced, for example, by another 50%. Also similar to the previous two risk analysis, reports may be generated to provide the user with the results from the analysis.
• Still another risk analysis process may analyze SAN redundancy to warn a user against potential dangerous SAN fabric design. Referring to FIG. 14 a SAN redundancy[0073]risk analysis process400, which may be included in the domain management process10 (also shown in FIG. 6), is shown. The SAN redundancyrisk analysis process400 starts402 by the user selecting404 a portion of the software application domain12 (shown in FIG. 6) and the storage domain14 (also shown in FIG. 6) for execution of the risk analysis. After theuser selection404 is made, theprocess400 determines406 if the user selection includes one or more data paths that extend from an application to a storage device. If the user selection does not include a data path that extends from an application to a storage device the SAN redundancyrisk analysis process400 stops408. If one or more data paths in the user selection extend from an application to a storage device, theprocess400 selects410 one the of the data paths that extends from an application to a storage device and then simulates412 a malfunction of a switch included in the selected data path. After simulating412 the switch failure, theprocess400 determines414 if there is one or more redundant paths for data to travel due to simulated switch failure. If theprocess400 determines that there is no redundant path for data to pass, the selected data path is considered failed and the downtime business cost of the application is added416 to the cost variable associated with the particular switch with the simulated malfunction. If the process determines that a redundant data path is present or after the downtime application cost is added to the switch cost variable, theprocess400 determines418 if another data path included in the user selection extends from an application to a storage device. If another data path is present theprocess400 returns to select another one of the data paths that extend from an application to a storage device. If there is no unanalyzed data path remaining, which extends from an application to a storage device, theprocess400reports420 to the user the results of the redundancy analysis and stops408. Typically the reporting420 may include the data paths included in the user selection that passed and failed the redundancy analysis and the cost variable of the switches included in the analysis along with a determination if one or more of the switch are also a single point of failure.
• In still another risk analysis the compatibility of components (i.e., switches, SANs, NAS drives, etc.) included in a user selection may be analyzed by one or more rules stored in a rule database. One possible rule may compare the configuration of each component, included in the user selection, against configuration information in the rule database. For example, each component may be analyzed to determine if the appropriate vender-suggested firmware is executed by comparing, for example, component type, vendor identification, model number, firmware revision level, and/or other similar configuration parameter. Also, components may be reduced into subcomponents for this analysis. For example, a host server component may be broken down to a host bus adapter (HBA), a volume manager, and a file system. Each component may have multiple configurations stored in the rule database since different firmware revisions could be certified with this release.[0074]
• Another rule that may be used in the compatibility analysis is the latest firmware rule. For analyzing using this rule, each component configuration, included in the user selection, is compared to the associated configuration information in the rule database to determine if the latest known firmware is executed on each component. Also, in some arrangements a rule may be used to detect switches in the same fabric (i.e., one or more fibre channel switches in some networked topology) containing different firmware. Still another rule may specifically flag any elements that meet the predetermined conditions of this rule. After one or more rules are applied to the components of the user selection, a report can be generated to provide the user with the analysis results for each component included in the user selection along with warnings associated with components that failed one or more of the applied rules, such as improper or outdated firmware executing on a component.[0075]
• Referring to FIG. 15 a component[0076]compatibility analysis process430, which may be included in the domain management process10 (shown in FIG. 6), is shown. The componentcompatibility analysis process430 starts432 by the user selecting434 a portion of the software application domain12 (shown in FIG. 6) and the storage domain14 (also shown in FIG. 6) for execution of the analysis. After theuser selection434 is made, theprocess430 accesses436 and retrieves one or more rules associated with the user selection from a rules database. After the one or more rules have been retrieved436, theprocess430 analyzes438 each component against the one or more rules that correspond to the component. After analyzing438 each component to the one or more corresponding rules, theprocess430reports440 the results of applying the one or more rules to the components of the user selection. After reporting440 to the user the componentcompatibility analysis process430 stops442.
• Referring back to FIG. 2, access to components included in the[0077]storage domain14 can be constrained by one or more constraint methods to assist theoperating system22. For example, three constraint methods can be applied to thestorage domain14 to assist theoperating system22. Some operating systems have been designed for directly attached storage devices and not for distributed storage networks such as a SAN. Based on this design, operating systems may assume a particular order of the store devices present on a bus. But, while this assumption may be valid for directly attached storage devices where the order persists after a system reboot, it may not be valid for distributed storage networks connected to a SCSI bus. For example, the order of the storage devices in an SAN on a SCSI bus may depend on numerous factors (e.g., WWN of the individual devices) that are outside the control of the operating system. By constraining the address of each storage device, the operating may not use the invalid assumption. In particular, by binding the WWN of each storage device to a particular bus address (i.e., a SCSI ID address) the operating system will not be alerted that the storage device was a previously mounted volume on another SCSI target ID address. Also, in some arrangements thedomain management process10 stores binding information in one or more tables. For example, a table can contain information associated with a combination of an HBA port, HBA SCSI Bus ID and a HBA logical unit number (LUN) number for mapping a particular combination of storage domain components.
• Referring to FIG. 16 a persistent[0078]binding process450, which may be included in the domain management process10 (shown in FIG. 2), is shown. The persistentbinding process450 starts452 by the user selecting454 a host server that is connected to the storage device to be bound. After selecting454 the server host, theprocess450 has the user select456 a host bus adapter (HBA) that is used by the host server to communicate to the storage device of interest. After selecting456 the HBA, theprocess450 has the user select theWWN458 of the particular storage device to be bound and select460 the bus target identification (e.g., SCSI target identification) that is to be bound to the WWN of the storage device. After selecting458 the WWN and selecting460 the bus target identification, the persistentbinding process450binds462 the WWN of the selected storage device to the selected bus target identification and the process ends464.
• Referring to FIG. 2, the second of the three constraint methods provides security and isolation functionality to the[0079]SAN38 by grouping components into a single group so that components within the group can sense the other group members but components outside the group cannot sense any of the group members, even if the outside components are connected to one or more of the group members. Typically this zoning method is accomplished either by hardware zoning or software zoning depending upon the type of component being added to the group. For example, some switch ports may only be zoned by hardware, however some fabric channel switch software allows grouping by software zones. Also, only certain entities such as switch ports, storage device port assigned a WWN, or a logical unit number (LUN) which is a unique identifier that is used on a bus (e.g., SCSI bus) to enable the bus to differentiate among up to eight separate devices may be zoned.
• Referring to FIG. 17 a[0080]zoning process470, which may be included in the domain management process10 (shown in FIG. 2), is shown. Thezoning process470 starts472 by the user selecting474 a fabric included in the storage domain14 (also shown in FIG. 2). After the fabric is selected474, theprocess470 has the user select476 one or more switches included in the selected fabric for zone membership. The user shall have the ability to drill down to the switch port level and select a port that is a zone-able entity. After the one or more switches have been selected476, theprocess470 has the user select478 one or more storage devices for zone membership. In this case the user shall have the ability to drill down to the WWN of the particular storage device and the LUN of the WWN of the zone-able storage device. Once the one or more switches and storage devices have been selected476,478, thezoning process470 creates480 a zone to include the selected members so that the members can be sensed by other members but not by components outside the newly created zone. After creating480 the zone, thezoning process470 stops482.
• Referring to FIG. 2, a third constraint method that can be placed on the[0081]storage domain14 is by controlling access to one or more of the storage devices included theSAN38. By controlling access, thehost server24 can be restricted from accessing and subsequently storing and retrieving data on thestorage device40,42,44 included in theSAN38. For example, by restricting the host server access fromRAID #142 andtape storage #044, all of the data marked for storage through thehost server24 is placed onRAID #040 and theother storage devices42,44 can be used by another host server (not shown).
• Referring to FIG. 18 a[0082]masking process490, which may be included in the domain management process10 (shown in FIG. 2), is shown. Themasking process490 starts492 by the user selecting494 a host server in order to have its access restricted to one or more particular storage devices in a SAN. After the user has selected494 the host server, theprocess490 has the user select496 a particular SAN which includes the one or more storage devices that are to have access restricted from the selected host server. After selecting496 the SAN, theprocess490 has the user select498 one or more storage devices in the SAN that the selected host server is to be restricted from. After the one or more storage devices have been selected498, theprocess490 restricts500 access of selected server host to the one or more selected storage devices and then stops502. By employing all three of these constraint methods on the storage domain14 (shown in FIG. 2), data storing by one or more host servers may be tailored for quick and efficient access to particular storage devices included in a SAN.
• Referring to FIG. 19, a[0083]graphical user interface510 included in thedomain management process10 is shown displaying graphical representations of the components included in the software application andstorage domains12,14 (also shown in FIG. 1) after stitching collected information. An email software application, which is included in the application domain12 (also shown in FIG. 1), is graphically represented with anapplication level512 that includes some components of the email software application that are positioned on an upperhierarchical level514 that includes amailbox store516, apublic folder store518, and other assorted email message stores.Respective data paths520,522 are shown connecting the components of the upperhierarchical level514 to components in a lower hierarchical level524 of the application. For example, onedata path520 displays the connection between themailbox store516 and afirst storage group526 and adifferent data path522 shows the connection between thepublic folder store514 and thefirst storage group520. As mentioned, typically this information is provided todomain management process10 by the particular agent assigned to the mailboxes of the email application. The lower hierarchical level524 of theapplication level512 also includesother storage groups528,530,532 for grouping other upper level components. Each of thestorage groups526,528,530,532 is connected to arespective data path534,536,538,540 that provides a graphical path to a particular storage label (e.g., “C:”, “E:”)542,544 of ahost server546 in ahost level548 of thestorage domain14. For example, as shown in the figure, themailboxes store516 is placed in thefirst storage group526 that is stored on the “C:” drive542 of thehost server546 that is located in thehost level548 of the storage domain14 (shown in FIG. 1). Continuing down the graphical tree structure of the levels included in the application andstorage domains12,14, thehost server546 passes data over adata path548 to aswitch level550 that is also included in thestorage domain14 and includes a graphical representation of one or more switches (not shown) present in the storage domain along with the associated data paths. Below the switch level550 a subsystem level (not shown) includes graphical representations of the storage devices that are connected by data paths to theswitch level550. For example, graphical representations of one or more SAN or NAS drives may be included in the subsystem level to represent the storage devices of thestorage domain14 which provide the actual data storage locations for themailbox store512 and the other assorted stores.
• Although some implementations have been described, other implementations are also within the scope of the following claims.[0084]
• The domain management process[0085]10 (shown in FIG. 1) described herein is not limited to the software embodiment described above; it may find applicability in any computing or processing environment. Thedomain management process10 may be implemented in hardware, software, or a combination of the two. For example, thedomain management process10 may be implemented using circuitry, such as one or more of programmable logic (e.g., an ASIC), logic gates, a processor, and a memory.
• The[0086]domain management process10 may be implemented in computer programs executing on programmable computers that each includes a processor and a storage medium readable by the processor (including volatile and non-volatile memory and/or storage elements). Each such program may be implemented in a high-level procedural or object-oriented programming language to communicate with a computer system. However, the programs can be implemented in assembly or machine language. The language may be a compiled or an interpreted language.
• Each computer program may be stored on an article of manufacture, such as a storage medium (e.g., CD-ROM, hard disk, or magnetic diskette) or device (e.g., computer peripheral), that is readable by a general or special purpose programmable computer for configuring and operating the computer when the storage medium or device is read by the computer to perform the functions of the[0087]domain management process10. Thedomain management process10 may also be implemented as a machine-readable storage medium, configured with a computer program, where, upon execution, instructions in the computer program cause a machine to operate to perform the functions of the domain management process described above.
• Embodiments of the[0088]domain management process10 may be used in a variety of applications. Although thedomain management process10 is not limited in this respect, the domain management process may be implemented with memory devices in microcontrollers, general purpose microprocessors, digital signal processors (DSPs), reduced instruction-set computing (RISC), and complex instruction-set computing (CISC), among other electronic components.
• Embodiments of the[0089]domain management process10 may also be implemented using integrated circuit blocks referred to as core memory, cache memory, or other types of memory that store electronic instructions to be executed by a microprocessor or store data that may be used in arithmetic operations.
• A number of embodiments of the invention have been described. Nevertheless, it will be understood that various modifications may be made without departing from the spirit and scope of the invention.[0090]

Claims (12)

What is claimed is:

1. A system monitoring process for monitoring data flow comprises:

an application level agent process that collects configuration data from a software application, the configuration data represents configuration dependencies between the software application and a storage system; and

a storage system level agent process that collects configuration data from the storage system, the configuration data represents the configuration dependencies between the software application and the storage system.

2. The system monitoring process ofclaim 1 further comprises:

an operating system level agent process for collecting configuration data from an operating system, the configuration data represents configuration dependencies between the operating system and the storage system.

3. The system monitoring process ofclaim 1 wherein the storage system level agent process collects configuration data that represents configuration dependencies between components of the storage system.

4. The system monitoring process ofclaim 1 wherein the configuration dependencies between the software application and the storage system includes a software data path.

5. The system monitoring process ofclaim 3 wherein the configuration dependencies between components of the storage system includes a physical data path.

6. The system monitoring process ofclaim 3 wherein the configuration dependencies between components of the storage system includes a logical data path.

7. The system monitoring process ofclaim 1 further comprises:

a mapping process for mapping the configuration dependencies between the software application and the storage system; and

a visualization process for disseminating the mapped configuration dependencies to a user.

8. The system monitoring process ofclaim 1 wherein the storage system includes a host server.

9. The system monitoring process ofclaim 1 wherein the storage system includes a switch.

10. The system monitoring process ofclaim 1 wherein the storage system includes a storage device.

11. The system monitoring process ofclaim 1 wherein the storage system level agent process collects configuration data from a table stored in the storage system.

12. A method for providing monitoring of data flow in resources of an enterprise comprises:

collecting configuration data from a software application that is a first resource of the enterprise, the configuration data represents configuration dependencies between the software application and a storage system that is a second resource of the enterprise;

collecting configuration data from the storage system, the configuration data represents the configuration dependencies between the software application and the storage system;

mapping the configuration dependencies from the software application and the storage system; and

disseminating the mapped configuration dependencies to a user.

Images