DE102020102936B4 - LOCATE POTENTIAL PROBLEMS ON OBJECTS - Google Patents

Abstract

Non-transitory machine-readable storage medium (300; 404) containing instructions that cause a system to be executed: based on a comparison of measurement data collected at different hierarchical levels (104; 108; 112; 116) of a computing environment (100), a potential problem and to identify (302; 406) a specific hierarchy level of the different hierarchy levels (104; 108; 112; 116) as a potential cause of the potential problem, wherein the different hierarchy levels (104; 108; 112; 116) have a physical machine level (112 ) and a program layer (104) comprising programs (106) for execution on one or more physical machine layer (112) hosts (114); andbased on measurement data collected for objects within a particular hierarchy level of the different hierarchy levels (104; 108; 112; 116):compute the respective localization values for a plurality of subsets of the objects within the particular hierarchy level; andorder the plurality of subsets according to the respective location values;determine that the location values for the plurality of subsets of objects each violate a criterion;based on the determination, determine that the potential problem in the plurality of subsets of objects in the particular hierarchical level is localized; andtrigger a corrective action to fix the potential problem responsive to the ranking of the multiple subsets of the objects.

Description

State of the art

A computing environment includes sensors or monitoring agents to collect information about objects in the computing environment. A "computing environment" may refer to any arrangement that contains objects capable of performing various tasks, including processing tasks, storage tasks, and/or communication tasks.

SOLE, Marc et al.: Survey on Models and Techniques for Root-Cause Analysis, 3 July 2017, https://arxiv.org/abs/1701.08546v2, relates to an examination of models and techniques for root-cause analysis.

US 2014/0 172 371 A1 describes an adaptive fault diagnosis system including a memory storing machine-readable instructions to receive metrics and events from an enterprise system and using a substitution graph to determine whether a received metric or event belongs to a cluster containing one or more correlated metrics and/or events grouped based on similarity. Further, if the received metric or event belongs to the cluster, the memory may store machine-readable instructions to use a detection graph to determine whether the received metric or event is identifiable to form an error pattern by a error path of the detection graph is traversed. Further, the memory may further store machine-readable instructions for diagnosing a fault based on traversing the detection graph's fault path.

US 2017/0 017 537 A1 relates to a host device having a memory and a processor, the host device being configured to determine an anomaly associated with an attribute of a computing environment resource of the computing infrastructure. The host device is configured to correlate an object associated with the attribute of the detected anomaly with an associated computer infrastructure object. The host device is configured to determine a root cause probability for each of the correlated objects, the root cause probability identifying a probability that the correlated object functions as a cause of the detected anomaly. The host device is configured to output an identification of a root object associated with the anomaly based on the identified root cause probability.

US 2013/0 097 463 A1 describes an approach to determine root cause in a complex system based on monitoring and event data. It includes historical analysis of events with their probabilistic correlations. By applying measures of information between the random variables that embody these events, one can identify causes of problems and generate real-time recommendations for their locations in a hierarchical system. The estimation of system bottlenecks as well as the risk of black swan type events are also calculated. The processes are based on statistical processing of a virtual directed graph generated from historical events.

Summary

Aspects of the present disclosure relate to a non-transitory machine-readable storage medium, system and method according to the appended claims.

character list

• 1 ist ein Blockdiagramm einer Computerumgebung und einer Problemlokalisierungsmaschine gemäß einigen Beispielen.
• 2 ist ein Flussdiagramm eines Prozesses gemäß einigen Beispielen.
• 3 ist ein Blockdiagramm eines Speichermediums, das maschinenlesbare Anweisungen gemäß einigen Beispielen speichert.
• 4 ist ein Blockdiagramm eines Systems gemäß einigen Beispielen.
• 5 ist ein Flussdiagramm eines Prozesses gemäß weiteren Beispielen.

Some implementations of the present disclosure are described with reference to the following figures.

• 1 12 is a block diagram of a computing environment and a problem location engine, according to some examples.
• 2 12 is a flowchart of a process according to some examples.
• 3 12 is a block diagram of a storage medium storing machine-readable instructions, according to some examples.
• 4 12 is a block diagram of a system, according to some examples.
• 5 12 is a flowchart of a process according to further examples.

Throughout the drawings, identical reference numbers indicate similar, but not necessarily identical, elements. The figures are not necessarily to scale and some parts may be exaggerated in size to more clearly illustrate the example shown. In addition, the drawings include examples and/or implementations consistent with the description; However, the description is not limited to the examples and/or implementations given in the drawings.

Detailed description

In the present disclosure, use of the term "a", "an" or "the" is intended to include the plural forms as well, unless the context clearly dictates otherwise. Also, the term "comprises," "including," "includes," "comprising," "have," or "have," when used in this disclosure, specifies the presence of the specified items, but does not preclude the presence or addition of from other elements.

Examples of computing environments include one or a combination of the following: a cloud environment, a data center, a storage network, a communications fabric, etc.

A computing environment can include a virtualized environment that contains various objects, including hosts (physical machines), virtual machines (VMs) that can run on hosts, storage systems (a storage system can include a single storage device or a storage array that includes a) multiple storage devices) and/or other objects. An "object" can refer to any physical electronic device (e.g. computer, switch, router, storage system, etc.) of the computing environment, a virtual device (e.g. virtual machine, container, virtual storage volume, etc.) Obtain .), a program (including machine-readable instructions), and so on.

Information gathered by sensors or monitoring agents is analyzed by a system to identify problems within the computing environment. The identified issues can include bugs, failures, bugs, etc. in the computing environment. Once a problem has been identified, the system can generate a recommendation on how to fix the problem. A corrective action to resolve the issue can be specified in the generated recommendation.

Corrective Actions may include corrective actions to be performed by existing components in the Computing Environment and/or corrective actions to be performed by components that may not yet have been deployed to the Computing Environment. For example, the recommended actions may include actions related to configurations of a virtualized environment, e.g. B. Configurations of VMs. As further examples, the recommended actions may include actions related to host and/or storage system configurations. An action related to a configuration of an object may relate to setting or adjusting a configuration of the object. For example, if the object is a VM, the VM's configuration may include a physical resource set (e.g., a processing resource, a storage resource, or a communication resource) that is allocated to the VM. Other sample configurations of other object types can be specified or customized.

In addition, corrective actions may include actions to disable or stop objects in the computing environment, actions to start objects in the computing environment, actions to adjust the operating speed of objects, and so on.

A computing environment can contain a large number of objects. As a result, a large amount of measurement data can be collected for the large number of objects, making analysis of the measurement data difficult. "Measurement Data" may refer to any information collected by a sensor or monitoring agent about an object or group of objects or other part of the computing environment. Challenges associated with analyzing metrics can include spotting patterns (to identify problems) and pinpointing potential problems for specific objects in the computing environment. As used herein, locating a problem on an object may refer to assessing that the problem is caused by the object (either individually or as part of a collection of objects).

Examples of metrics that may be collected in a computing environment may include latency metrics that may be reported by sensors or monitoring agents for VMs, hosts (ie, physical machines), storage systems, programs, and other objects. A latency metric provides an indication of the latency associated with an object, e.g. B. A period of time for the execution of a specific task (e.g. a processing task, a storage task and/or a communication task). In other examples, measurement data may include performance metrics representing performance levels of the respective objects, e.g. B. a count of operations performed by an object in a unit of time, a count of input/output (I/O) operations performed by an object in a unit of time, a depth of a queue for buffering data during the operation of an object and so on. Other types of metrics may be included as part of the metrics collected in other example computing environments.

1 10 is a block diagram of a computing environment 100 that includes various objects at various hierarchical levels of a hierarchy 102, according to some examples. The hierarchy 102 includes the following hierarchical levels: a program level 104 containing programs 106, a virtual level 108 containing VMs 110 (and/or other virtual resources such as containers, virtual storage volumes, etc.), a physical level 112 containing hosts 114 (and/or other physical resources such as processors, network interface controllers, graphics controllers, memory controllers, etc.) and a memory tier 116 containing memory systems 118 (and/or other memory resources such as memory caches, etc.).

Although in 1 While specific example hierarchy levels are shown, different hierarchy levels may be provided in other hierarchies.

Note that while FIG. 1 shows the various hierarchy levels 104, 108, 112, and 116 as being at different levels, it is noted that one hierarchy level may be included or otherwise overlap another hierarchy level. For example, virtual tier 108 VMs may run within physical tier 112 hosts. As another example, virtual tier VMs 108 may include storage tier 116 storage system disks.

A program 106 may refer to one or a combination of the following: a boot program, an operating system, an application program, a utility, etc. A storage system 118 may include one storage device or multiple storage devices. The storage device may be implemented using a hard disk-based storage device, a solid-state storage device, a storage device, and so on. A host 124 can be in the form of a physical machine, such as a computer, a collection of computers, a portion of a computer, and so on.

The computing environment 100 further includes sensors 120 and/or monitoring agents 122 to measure data related to the operation of various objects at the various levels of the hierarchy 102 . Computing environment 100 may also include monitoring agents 122 . A sensor 120 may refer to a hardware sensor that can physically measure a metric. A monitoring agent 122 can reference machine-readable instructions that can monitor the operation of a given object and derive a metric based on the monitored operation.

The sensors 120 and/or monitoring agents 122 can provide measurement data 124 to a problem location engine 126 . Measurement data 124 may include various metrics, including a latency metric, a performance metric, and so on.

As used herein, an "engine" may refer to hardware processing circuitry that includes any one or combination of a microprocessor, a core of a multi-core microprocessor, a microcontroller, a programmable integrated circuit, a programmable gate array, a digital Signal processor or other hardware processing circuit. Alternatively, an "engine" may refer to a combination of hardware processing circuitry and machine-readable instructions (software and/or firmware) executable on the hardware processing circuitry.

The various objects of the computing environment, including programs 106, VMs 110, hosts 124, and storage systems 118, may have relationships with one another. For example, a VM 110 (or multiple VMs 110) may be running on a host 114. A program 106 (or multiple programs 106) can be executed in a VM 110 or in a host 114. A storage system 118 may be part of a host 114 or may be reached by a host 114 (or multiple hosts 114) or a VM 110 (or multiple VMs). Other relationships between objects in the various hierarchical levels of hierarchy 102 are also possible. More generally, a relationship between objects may include a containment relationship (in which a first object is contained within a second object), a link relationship (in which a first object is physically or logically linked to a second object), or any other type of relationship.

The problem localization engine 126 includes vertical evaluation logic 128 and horizontal evaluation logic 130. The vertical evaluation logic 128 and the horizontal evaluation logic 130 can be implemented as parts of the hardware processing circuitry of the problem localization engine 126 or can be implemented as machine-readable instructions that are executable by the problem localization engine 126.

Vertical scoring logic 128 is used to score measurement data collected for objects in the various hierarchical levels of hierarchy 102 . The assessment performed by the vertical assessment logic 128 may include a determination of a potential problem in the computing environment 100 based on the measurement data collected for the objects at the various hierarchical levels. The vertical scoring logic 128 may compare the respective metrics collected for the different hierarchical levels of the hierarchy 102 (e.g., by comparing the metrics collected for the program level 104, the metrics collected for the virtual level 108, and the metrics collected for the physical layer Measurement data tier 112 and measurement data collected for storage tier 116).

As a more specific example, the measurement data may include values of latency metrics associated with the different levels of the hierarchy. For example, there may be a latency metric for the storage tier 116 (e.g., referred to as "memory latency"), a latency metric for the physical tier 112 (e.g., referred to as "host latency", a latency metric) for the virtual tier 108 (e.g., referred to as "VM latency") ) a latency metric for the program layer 104 (e.g., referred to as "program latency") or another type of latency metric (e.g., a network latency metric, an overall latency metric composed of multiple latency metrics, etc.).

The latency metric for a corresponding hierarchy level may express the latency experienced by objects of the hierarchy level. Thus, the vertical evaluation logic 128 can determine that the physical layer and the storage layer experience latencies that are comparable to each other (i.e., the latency experienced by the physical layer is within a certain range of the latency experienced by the storage layer). . whereas the virtual layer 108 and program layer 104 may have low latency. Based on this determination, vertical assessment logic 128 may conclude that a latency issue may be associated with a communication network between hosts 124 of physical tier 112 and storage systems 118 of storage tier 116 . The potential problem in this example is a network delay issue.

In some cases, the values of the metric values reported for a first hierarchy level might not match the metric values reported for a second hierarchy level. This can be caused by inconsistencies in the metric measurements reported by sensors 120 and/or monitoring agents 122 . For example, sensors 120 and/or monitoring agents 122 may encounter errors or errors that may prevent sensors 120 and/or monitoring agents 122 from properly reporting measurement data. Alternatively, noise in the computing environment 100 may cause sensors 120 and/or monitoring agents 122 to report unreliable measurement data. The noise may be due to faulty objects in the computing environment 100, for example.

Inaccurate measurement data may falsely indicate that the latency experienced by the virtual layer 108 and the latency experienced by the physical layer 112 are relatively close, when in fact the virtual layer 108 does not exhibit any significant latency, indicating a potential problem with the virtual layer .

Inaccurate measurement data from the sensors 120 and/or monitoring means 122 can make diagnosing potential problems difficult or inaccurate.

In accordance with some implementations of the present disclosure, the horizontal measurement logic 130 can increase confidence in the association of metrics with different objects of the different hierarchical levels of the hierarchy 102 . For example, if vertical assessment logic 128 determines that objects at a particular hierarchy level are the cause of a potential problem, horizontal assessment logic 130 may perform an assessment as to whether the potential problem can be localized to a subset of the objects within that hierarchy level. A subset of objects can contain a single object or multiple objects (e.g. fewer than all objects in the respective hierarchy level). Locating a potential problem in a subset of objects may relate to assessing that the potential problem is caused by the subset of objects.

In some examples, the assessment of whether a potential problem is located in a particular subset of objects includes an assessment based on metrics collected for the same or a similar object type within the hierarchy level. For example, if a given hierarchical level contains multiple different types of objects, the evaluation performed by horizontal scoring logic 130 is based on metrics for objects of the same or similar type within the given hierarchical level. For example, the virtual tier 108 may include the following different types of virtual resources: VMs (a first type of virtual resource), virtual storage volumes (a second type of virtual resource), etc. Locating a potential problem in a subset of objects in the virtual Level 108 may be based on metrics collected for the first object type (e.g. VMs) and/or for the second object type (e.g. virtual storage volumes).

The assessment performed by the horizontal assessment logic 130 can also be referred to as a horizontal comparison that attempts to pinpoint a potential problem to a specific subset of objects such as hosts, storage systems, data stores, VMs, programs, and so on. The horizontal comparison by the horizontal scoring logic 130 is used to verify the "signal" provided by the vertical scoring logic 128, where the "signal" may refer to a scoring result that the vertical scoring logic 128 outputs. The score from the vertical score logic 128 may include an indication that a particular hierarchy level (or multiple hierarchy levels) is (are) the potential cause of a potential problem.

For example, if the horizontal assessment logic 130 determines that a potential latency problem is localized to a particular physical layer 112 host(s) 124, this would provide greater confidence in the assessment that the physical layer hosts 124 112 are causing the latency problem in comparison to another scenario where the horizontal assessment logic 130 cannot locate the latency problem in a subset of the hosts 124.

The horizontal comparison is performed based on the use of a technique that calculates location values for respective objects within a hierarchy level. In some examples, Eq. 1 can be used to calculate a localization score Lscore for a given object within a hierarchy level. $$LscO\mathrm{right}e=p{\left(ScOpeMet\mathrm{right}ic\right)}^n\cdot \left(p\left(P\mathrm{right}OblemMet\mathrm{right}ic|ScOpeMet\mathrm{right}ic\right)-p\left(P\mathrm{right}OblemMet\mathrm{right}ic\right)\right).$$

Eq. 1 returns the Lscore localization values, which are based on Lscore metric values collected for objects of a hierarchy level. The localization value gives an indication of the importance of an identified potential problem for a specific subset of objects at the hierarchy level. If the localization score Lscore meets a criterion (eg, the localization score Lscore exceeds a certain threshold), this gives an indication that the potential problem is sufficiently localized for a certain subset of objects in the hierarchy level.

In Eq. In 1 the value of n is greater than 0 and less than 1. Also a function p() represents a probability. The ScopeMetric parameter represents a condition that must be satisfied by a ScopeMetric metric value that indicates a potential problem (for example, ScopeMetric is a threshold). The ProblemMetric parameter represents a metric value that is problematic. The probability p(ScopeMetric) represents a probability that p(ScopeMetric)metric values of Objects in the entire computing environment 100 (or part of the computing environment 100) meet the condition. The probability p(ProblemMetric) represents a probability of problematic p(ProblemMetric) metric values. The probability p(ProblemMetricl ScopeMetric) represents a probability of problematic p(ProblemMetric I ScopeMetric) metric values when metric values of objects in all (or a portion) of the computing environment 100 satisfy the condition. For example, if p(ProblemMetric) for VMs in a computing environment (or part) overall p(ProblemMetric) is low, but p(ProblemMetric | ScopeMetric) ) (which in this example may represent a probability that a VM is in a problem state , if the VM is on a specific host) is high, the following would be indicated: Being on the specific host predicts that a VM has the potential problem and therefore the host is worth further investigation (to determine the cause or to determine correlation).

More generally, the localization determination is based on the relative values of p(ProblemMetric) versus p(ProblemMetric | ScopeMetric). The larger the difference between these values, the more important the condition (represented by ScopeMetric) is in determining the level of the potential problem.

Eq. 1 produces a localization score based on a trade-off between the generality of the rules (or the relative size of a subset p(ScopeMetric)) and the unusualness of the distribution or relative accuracy (the difference between the
p(ProblemMetric | ScopeMetric)rules accuracy and the default accuracy p(ScopeMetric)) p(ScopeMetric)based .

Localization score is based on metric data collected from an environment and provides an indication of the importance of an identified potential issue for a specific subset of objects.

Note that multiple objects (or multiple subsets of objects) within a hierarchical level can have corresponding localization scores Lscore that each satisfies the criterion (e.g., the localization scores Lscore of multiple subsets of objects that each exceed the specified threshold). In such cases, the location values Lscore for the multiple subsets of objects can be used to Lscore the multiple subsets of objects. For example, a subset of objects with a higher localization score Lscore may be ranked higher than another subset of objects with a lower localization score Lscore.

The problem location engine 126 generates an evaluation result 132 which is output to a resolution engine 134 . In examples where the problem location engine 126 determines that the potential problem is located in a subset of objects, the score 132 may include information identifying the subset of objects. In examples where the problem location engine 126 determines that the potential problem is located in multiple subsets of objects, the score 132 may include ranking information that ranks the multiple subsets of objects, where the ranking is based on the location values, for example.

The resolution engine 134 can respond to the assessment result 132 by triggering an automated corrective action to address the potential problem. See above for examples of remedial measures.

In other examples, the resolution engine 132 may also provide information to a user device 136 (such as a user device associated with an administrator). User device 136 may include a desktop computer, notebook computer, tablet computer, smartphone, and so on. The information provided to user device 136 may identify the potential problem and the objects within a hierarchy level (or multiple hierarchy levels) at which problem location engine 126 located the potential problem. The information may be presented in a user interface of user device 136 to allow the user to take an action in response to the potential problem.

2 is a flowchart of a process used, for example, by various in 1 units shown can be executed. The problem location engine 126 receives (at 202) Metric values for different hierarchical levels of the computer environment 100. The measurement data are provided by the sensors 120 and/or monitoring means 122.

The vertical scoring logic 128 compares (at 204) the metric values for the various hierarchical levels. For example, vertical scoring logic 128 compares program tier 104 metrics to virtual tier metrics 108, compares virtual tier 108 metrics to physical tier metrics 112, and compares physical tier 112 metrics to storage tier metrics 116, etc. Based on the comparisons, the vertical assessment logic 128 may determine whether a potential problem is associated with a given hierarchy level (or multiple given hierarchy levels).

Within each of the given hierarchical levels identified by the vertical scoring logic 128 as being associated with the potential problem, the horizontal scoring logic 130 performs (at 206) a horizontal comparison to assess whether the potential problem can be located in a subset of the Objects within the specified hierarchy level. As discussed above, horizontal scoring logic 130 calculates location values (e.g., according to Eq. 1) for respective subsets of objects within the given hierarchy level. The horizontal comparison performed by horizontal scoring logic 130 may include comparing a location score of a subset of objects to a specified threshold. For example, if the localization value exceeds the specified threshold, the potential problem is localized to the appropriate subset of objects.

It is possible that the localization values for several subsets of objects exceed the threshold. In this case, the potential problem is localized to multiple objects.

The importance of a potential problem is based on the localization between types of metric values and respected types of objects of a hierarchy level. For example, if a host-related metric is in a host-related object, it means that the related issue is more important. This can be based on the application of a specific rule (or rules). More generally, determining that the potential problem is located in a subset of the objects includes determining that a type of metric for the subset of objects matches an object type of an object in the subset of objects. It is assumed that the type of measurement data corresponds to an object type based, for example, on the predefinition of an administrator, a program or a machine.

After the horizontal comparison performed by the horizontal scoring logic 130, the problem location engine 126 provides (at 208) the score 132 to the resolution engine 134. The score 132 may identify a subset of objects as being located for the potential problem, or it may identify multiple Identify subsets of objects as localized to the potential problem. In the latter case, the score 132 may include a ranking of the multiple subsets of objects, for example based on the location values of the multiple subsets of objects.

The resolution engine 134 initiates (at 210) corrective action to address the potential problem. If the scoring result 132 contains information that locates the potential problem in a subset (or subsets) of objects, then this provides the resolution engine 134 with an indication that the score made by the vertical scoring logic 128 associated the potential problem with a is given hierarchy level is accurate with a high confidence. In such a scenario, the resolution engine 134 may trigger corrective action to be taken.

In response to determining that the potential problem is located in the subset of objects, the correction engine 134 may identify a subset of the objects as the cause of the potential problem. In cases where there are multiple subsets of objects for which the potential problem is located, the ranking of the multiple subsets of objects can be used by the correction engine 134 to identify the subset of objects that is ranked as the highest cause the potential problem.

If the evaluation result 132 does not indicate that the potential problem is located in a subset of objects, the correction engine 134 indicates that a cause of the potential problem lems is not specific to any subset of the objects. This may also be an indication that the vertical scoring logic 128's assessment of the potential problem as being associated with a particular hierarchical level has low confidence.

3 Figure 3 is a block diagram of a non-transitory machine-readable or computer-readable storage medium 300 storing machine-readable instructions that, when executed, cause a system to perform various tasks. The machine-readable instructions include potential problem identification instructions 302 for identifying a potential problem based on comparing metrics collected at different hierarchical levels of a computing environment. For example, the instructions for identifying potential problems 302 may be part of the vertical scoring logic 128 of FIG 1 be.

The machine-readable instructions further include problem location instructions 304 to determine, within one of the various hierarchy levels, whether the potential problem is located in a subset of the objects based on metrics collected for objects in the hierarchy level. For example, the problem location instructions 304 may be part of the horizontal scoring logic 130 of FIG 1 be.

4 4 is a block diagram of a system 400, according to some examples. The system 400 can be implemented using one or more computers.

The system 400 includes a hardware processor 402 (or multiple hardware processors). A hardware processor may include a microprocessor, a core of a multi-core microprocessor, a microcontroller, a programmable integrated circuit, a programmable gate array, a digital signal processor, or other hardware processing circuitry.

The system 400 includes a storage medium 404 storing machine-readable instructions executable on the hardware processor 402 to perform various tasks. Machine-readable instructions executable on a hardware processor may refer to the instructions executable on a single hardware processor or to the instructions executable on multiple hardware processors.

The machine-readable instructions include potential problem identification instructions 406 for identifying a potential problem based on comparing metrics collected at different hierarchical levels of a computing environment. For example, instructions 406 for identifying potential problems may be part of vertical scoring logic 128 of FIG 1 be.

The machine-readable instructions further include location score calculation instructions 408 for calculating location scores for objects within a hierarchy level of the various hierarchy levels, the location scores calculated based on measurement data collected for objects in the hierarchy level, and the location scores that indicate a degree of association between the objects within the hierarchy level and the potential problem.

The machine-readable instructions further include problem location instructions 410 to determine whether the potential problem is located in a subset of the objects based on the location scores. For example, location score computation instructions 408 and problem location location instructions 410 may be part of horizontal scoring logic 130 of FIG 1 be.

5 12 is a flowchart of a process according to some examples. The process of 5 includes receiving (at 502) metrics collected for objects collected at various hierarchical levels of a computing environment. The metrics may be obtained from sensors 120 and/or monitoring agents 122 1 to be collected.

The process further includes identifying (at 504) a potential problem based on comparing values of the metrics collected at the various hierarchical levels of the computing environment. The identification can be performed, for example, by the vertical evaluation logic 128 of 1 be performed.

The process further includes, for a hierarchy level of the different hierarchy levels, determining (at 506) whether the potential problem is located in a subset of the objects based on the values of the metrics collected for the objects in the hierarchy level.

A storage medium (e.g. 300 in 3 or 404 in 4 ) may include any one or combination of the following: a semiconductor memory device such as dynamic or static random access memory (a DRAM or SRAM), erasable and programmable read only memory (EPROM), electrically erasable and programmable read only memory (EEPROM ) and flash memory; a magnetic disk such as a fixed floppy disk and a removable floppy disk; another magnetic medium including adhesive tape; an optical medium such as a compact disc (CD) or a digital video disc (DVD); or some other type of storage device. It should be noted that the instructions discussed above may be provided on one computer-readable or machine-readable storage medium, or alternatively may be provided on multiple computer-readable or machine-readable storage media distributed in a large system with possibly multiple nodes. Such computer-readable or machine-readable storage media or media shall be considered part of an article (or article of manufacture). An article or article of manufacture may refer to each individual component or multiple components manufactured. The storage medium or media may reside either on the computer on which the machine-readable instructions are executed or at a remote location from which machine-readable instructions may be downloaded over a network for execution.

In the foregoing description, numerous details are set forth in order to provide an understanding of the examples disclosed herein. However, implementations can be practiced without some of these details. Other implementations may include modifications and variations from the details discussed above. It is intended that the appended claims cover such modifications and variations as come within the scope of the examples.

Claims (20)

A non-transitory machine-readable storage medium (300; 404) containing instructions that, when executed, cause a system to: based on a comparison of measurement data recorded at different hierarchical levels (104; 108; 112; 116) of a computer environment (100), a potential problem and a specific hierarchical level of the various hierarchical levels (104; 108; 112; 116) as a potential cause identifying (302; 406) the potential problem, the various hierarchical levels (104; 108; 112; 116) a physical machine level (112) and a program level (104), the programs (106) for execution on one or more physical machine level (112) hosts (114); and based on measurement data collected for objects within a specific hierarchy level of the different hierarchy levels (104; 108; 112; 116): calculate the respective location values for a plurality of subsets of the objects within the determined hierarchy level; and order the plurality of subsets according to the respective location values; determine that the location values for the plurality of subsets of objects each violate a criterion; determine based on the determination that the potential problem is located in the plurality of subsets of the objects in the determined hierarchy level; and trigger a corrective action to address the potential issue responsive to the ranking of the multiple subsets of the objects. Non-transitory machine-readable storage medium (300; 404). claim 1 , wherein the different hierarchical levels (104; 108; 112; 116) further comprise a memory level (116). Non-transitory machine-readable storage medium (300; 404). claim 1 wherein the instructions, when executed, cause the system to: in response to determining that the potential problem is located in the subset of objects, identifying the subset of objects that is highest ranked as the cause of the potential problem. Non-transitory machine-readable storage medium (300; 404). claim 1 , wherein the various hierarchical levels (104; 108; 112; 116) further comprise a virtual level (108). Non-transitory machine-readable storage medium (300; 404). claim 4 , wherein the virtual layer (108) comprises a virtual machine (110) for executing on the one or more hosts (114) of the physical layer (112). Non-transitory machine-readable storage medium (300; 404). claim 5 , wherein the program to be executed on the one or more hosts (114) is to be executed on the virtual machine (110) for execution on the physical layer (112). Non-transitory machine-readable storage medium (300; 404). claim 1 , wherein the different hierarchical levels (104; 108; 112; 116) further comprise a virtual level (108) and a physical level (112). Non-transitory machine-readable storage medium (300; 404). claim 7 , wherein a virtual machine (110) of the virtual tier (108) comprises at least one hard disk of at least one storage system (118) of the storage tier (116). Non-transitory machine-readable storage medium (300; 404). claim 1 , wherein the objects of the specific hierarchical level include programs (106), virtual machines (110), hosts (114) or storage systems (118). System (400) which includes: a processor (402); and a non-transitory machine-readable storage medium (300; 404) storing machine-readable instructions executable on the processor (402) to: based on the comparison of measurement data recorded at different hierarchical levels (104; 108; 112; 116) of a computer environment (100), a potential problem and a specific hierarchical level of the various hierarchical levels (104; 108; 112; 116) as a potential cause identifying (406) the potential problem, the various hierarchical levels (104; 108; 112; 116) a physical machine level (112) and a program level (104) containing programs (106) for execution on one or more hosts ( 114) of the physical machine layer (112); calculate (408) respective location values for a plurality of subsets of objects within the given hierarchy level of the different hierarchy levels (104; 108; 112; 116) based on measurement data collected for objects within the given hierarchy level and the location values give a degree of association between the objects within the given hierarchical level and the potential problem; order the plurality of subsets according to the respective location values; determine that the location values for the plurality of subsets of objects each violate a criterion; based on the determination, determining (410) that the potential problem is located in the plurality of subsets of the objects in the determined hierarchy level; and trigger a corrective action to address the potential issue responsive to the ranking of the multiple subsets of the objects. system (400) after claim 10 , wherein the different hierarchical levels (104; 108; 112; 116) further comprise a memory level (116). system (400) after claim 10 , wherein the various hierarchical levels (104; 108; 112; 116) further comprise a virtual level (108), and wherein the virtual level (108) comprises a virtual machine (110) for execution on the one or more hosts (114) of physical layer (112). system (400) after claim 12 wherein the program to be executed on the one or more hosts (114) is to be executed on the virtual machine for execution on the physical layer (112). system (400) after claim 10 , wherein the different hierarchical levels (104; 108; 112; 116) further comprise a virtual level (108) and a storage level (116). A method performed by a system (400) of a hardware processor (402), comprising: receiving (502) metrics collected for objects collected at different hierarchical levels (104; 108; 112; 116) of a computing environment (100), wherein the different hierarchical levels (104; 108; 112; 116) a physical machine level (112) and a program level (104), the programs (106) for execution on one or more hosts (114) of the physical machine level (112 ), comprises, comprise; Identifying (504) a potential problem and a particular hierarchy level of the different hierarchy levels (104; 108; 112; 116) as a potential cause of the potential problem based on comparing values of the metrics determined on the different hierarchy levels (104; 108; 112; 116) of the computing environment (100); and based on the values of the metrics recorded for the objects in the specific hierarchy level: calculating respective localization values for a plurality of subsets of the objects within the specific hierarchy level; ordering the plurality of subsets according to the respective location values; determining that the location values for the plurality of subsets of objects each violate a criterion; Based on the determination, determining (506) that the potential problem is located in a plurality of subsets of the objects in the determined hierarchy level; and triggering a corrective action to fix the potential problem responsive to the ranking of the multiple subsets of the objects. procedure after claim 15 , wherein the different hierarchical levels (104; 108; 112; 116) include programs (106), virtual machines (110), hosts (114) or storage systems (118). procedure after claim 15 , wherein the different hierarchical levels (104; 108; 112; 116) further comprise a memory level (116). procedure after claim 15 , wherein the various hierarchical levels (104; 108; 112; 116) further comprise a virtual level (108), and wherein the virtual level (108) comprises a virtual machine (110) for execution on the one or more hosts (114) of physical layer (112). procedure after Claim 18 wherein the program to be executed on the one or more hosts (114) is to be executed on the virtual machine for execution on the physical layer (112). procedure after claim 15 , wherein the different hierarchical levels (104; 108; 112; 116) further comprise a virtual level (108) and a storage level (116).

Images