System and method for self-healing of application centric infrastructure fabric memory

Abstract

Disclosed is a method that includes obtaining a list of processes in an application centric infrastructure fabric, sorting the list of processes according to an amount of memory increase associated with each respective process in the list of processes to yield a sorted list, selecting a group of processes from the sorted list and collecting a respective live process core for each process in the group of processes without pausing or killing any process in the group of processes. The method includes applying an offline leak detection tool to each process in the group of processes to yield a list of leaked memory addresses for a given process of the group of processes and transmitting a message to the given process with the list of leaked memory addresses, whereby the given process calls a function to release leaked memory associated with the given process as identified in the message.

Description

TECHNICAL FIELD

The present disclosure relates to garbage collection and more specifically to a method of collecting live process cores for a group of processes that have memory leak issues, transmitting the list to a cloud-based container for running a garbage detection on the live process cores, and transmitting back a list of leak address such that a respective process can to free the memory associated with the leak address.

BACKGROUND

A data-center fabric like Application Centric Infrastructure (ACI) could contain up to 400 leaf switches, each potentially running about 50 applications (e.g., linux processes, bgp processes, ospf processes), taking the total to about 20K linux processes. Like most embedded technology, ACI is built using C and C++, both of which are prone to memory leaks due to programming errors. Memory leaks can occur since there is no native garbage collection facility.

Applications leaking even a small amount of memory, such as 1 Kb per day each, can result in 20 MB of wasted memory per day. Such memory degradation is harmful in the long run and could cause serious side effects that include increased response time and non-recoverable process crashes. Moreover, running analytics on the entire fabric memory is close to impossible due to the sheer scale and volume of memory that needs to be analyzed.

Native in-memory garbage collection is getting popular, but is not even close to implementation for C and C++ especially in real-time fabric environments which are sensitive to time delays. Though C++ does have explicit garbage collection facilities like smart pointers, there has been no progress on providing an implicit garbage collector. One of the main reasons for this could be the fear of performance impact. For example, “mark and sweep” garbage collection freezes everything in the process while it is at work. Even milliseconds of paused processing can impact real-time embedded systems.

In essence, for the millions of embedded systems that use C/C++, proper garbage collection is not yet a reality and the only solution at the moment is to kill and restart the offending process.

BRIEF DESCRIPTION OF THE DRAWINGS

In order to describe the manner in which the above-recited and other advantages and features of the disclosure can be obtained, a more particular description of the principles briefly described above will be rendered by reference to specific embodiments thereof which are illustrated in the appended drawings. Understanding that these drawings depict only exemplary embodiments of the disclosure and are not therefore to be considered to be limiting of its scope, the principles herein are described and explained with additional specificity and detail through the use of the accompanying drawings in which:

FIG. 1 illustrates an example system configuration;

FIG. 2 illustrates a Application Centric Infrastructure;

FIG. 3 illustrates a method embodiment;

FIG. 4 illustrates a switch and a messaging being sent to an operating system process; and

FIG. 5 illustrates and effect over time of memory leaks when the concepts disclosed herein are applied.

DESCRIPTION OF EXAMPLE EMBODIMENTS

Various embodiments of the disclosure are discussed in detail below. While specific implementations are discussed, it should be understood that this is done for illustration purposes only. A person skilled in the relevant art will recognize that other components and configurations may be used without parting from the spirit and scope of the disclosure.

Overview

Additional features and advantages of the disclosure will be set forth in the description which follows, and in part will be obvious from the description, or can be learned by practice of the herein disclosed principles. The features and advantages of the disclosure can be realized and obtained by means of the instruments and combinations particularly pointed out in the appended claims. These and other features of the disclosure will become more fully apparent from the following description and appended claims, or can be learned by the practice of the principles set forth herein.

The method includes obtaining a list of processes in an application centric infrastructure fabric, sorting the list of processes according to an amount of memory increase associated with each respective process in the list of processes to yield a sorted list, selecting a group of processes from the sorted list and collecting a respective live process core for each process in the group of processes without pausing or killing any process in the group of processes. The method includes applying an offline leak detection tool to each process in the group of processes to yield a list of leaked memory addresses for a given process of the group of processes and transmitting a message to the given process with the list of leaked memory addresses. The given process calls a function to release leaked memory associated with the given process as identified in the message.

Detailed Description

The present disclosure addresses the issues raised above. The disclosure provides a system, method and computer-readable storage device embodiments. First a general example system shall be disclosed in FIG. 1 which can provide some basic hardware components making up a server, node or other computer system.

First a general example system shall be disclosed in FIG. 1, which can provide some basic hardware components making up a server, node or other computer system. FIG. 1 illustrates a computing system architecture 100 wherein the components of the system are in electrical communication with each other using a connector 105. Exemplary system 100 includes a processing unit (CPU or processor) 110 and a system connector 105 that couples various system components including the system memory 115, such as read only memory (ROM) 120 and random access memory (RAM) 125, to the processor 110. The system 100 can include a cache of high-speed memory connected directly with, in close proximity to, or integrated as part of the processor 110. The system 100 can copy data from the memory 115 and/or the storage device 130 to the cache 112 for quick access by the processor 110. In this way, the cache can provide a performance boost that avoids processor 110 delays while waiting for data. These and other modules/services can control or be configured to control the processor 110 to perform various actions. Other system memory 115 may be available for use as well. The memory 115 can include multiple different types of memory with different performance characteristics. The processor 110 can include any general purpose processor and a hardware module or software module/service, such as service 1132, service 2134, and service 3136 stored in storage device 130, configured to control the processor 110 as well as a special-purpose processor where software instructions are incorporated into the actual processor design. The processor 110 may essentially be a completely self-contained computing system, containing multiple cores or processors, a bus (connector), memory controller, cache, etc. A multi-core processor may be symmetric or asymmetric.

To enable user interaction with the computing device 100, an input device 145 can represent any number of input mechanisms, such as a microphone for speech, a touch-sensitive screen for gesture or graphical input, keyboard, mouse, motion input, speech and so forth. An output device 135 can also be one or more of a number of output mechanisms known to those of skill in the art. In some instances, multimodal systems can enable a user to provide multiple types of input to communicate with the computing device 100. The communications interface 140 can generally govern and manage the user input and system output. There is no restriction on operating on any particular hardware arrangement and therefore the basic features here may easily be substituted for improved hardware or firmware arrangements as they are developed.

Storage device 130 is a non-volatile memory and can be a hard disk or other types of computer readable media which can store data that are accessible by a computer, such as magnetic cassettes, flash memory cards, solid state memory devices, digital versatile disks, cartridges, random access memories (RAMs) 125, read only memory (ROM) 120, and hybrids thereof.

The storage device 130 can include software services 132, 134, 136 for controlling the processor 110. Other hardware or software modules/services are contemplated. The storage device 130 can be connected to the system connector 105. In one aspect, a hardware module that performs a particular function can include the software component stored in a computer-readable medium in connection with the necessary hardware components, such as the processor 110, connector 105, display 135, and so forth, to carry out the function.

Having introduced the basic computing components which can be applicable to embodiments associated with this disclosure, the disclosure now turn to FIG. 2 which illustrates an example network environment.

FIG. 2 illustrates a diagram of example network environment 200. This figure is discussed with the concept of capturing agents on various network components. With reference to FIG. 2, fabric 212 can represent the underlay (i.e., physical network) of network environment 200. Fabric 212 can include spine routers 1-N (202A-N) (collectively “202”) and leaf routers 1-N (204A-N) (collectively “204”). Leaf routers 204 can reside at the edge of fabric 212, and can thus represent the physical network edges. Leaf routers 204 can be, for example, top-of-rack (“ToR”) switches, aggregation switches, gateways, ingress and/or egress switches, provider edge devices, and/or any other type of routing or switching device.

Leaf routers 204 can be responsible for routing and/or bridging tenant or endpoint packets and applying network policies. Spine routers 202 can perform switching and routing within fabric 212. Thus, network connectivity in fabric 212 can flow from spine routers 202 to leaf routers 204, and vice versa.

Leaf routers 204 can provide servers 1-5 (206A-E) (collectively “206”), hypervisors 1-4 (208A-208D) (collectively “208”), and virtual machines (VMs) 1-5 (210A-210E) (collectively “210”) access to fabric 212. For example, leaf routers 204 can encapsulate and decapsulate packets to and from servers 206 in order to enable communications throughout environment 200. Leaf routers 204 can also connect other devices, such as device 214, with fabric 212. Device 214 can be any network-capable device(s) or network(s), such as a firewall, a database, a server, an engine 220 (further described below), etc. Leaf routers 204 can also provide any other servers, resources, endpoints, external networks, VMs, services, tenants, or workloads with access to fabric 212.

VMs 210 can be virtual machines hosted by hypervisors 208 running on servers 206. VMs 210 can include workloads running on a guest operating system on a respective server. Hypervisors 208 can provide a layer of software, firmware, and/or hardware that creates and runs the VMs 210. Hypervisors 208 can allow VMs 210 to share hardware resources on servers 206, and the hardware resources on servers 206 to appear as multiple, separate hardware platforms. Moreover, hypervisors 208 and servers 206 can host one or more VMs 210. For example, server 206A and hypervisor 208A can host VMs 210A-B.

In some cases, VMs 210 and/or hypervisors 208 can be migrated to other servers 206. For example, VM 210A can be migrated to server 206C and hypervisor 208B. Servers 206 can similarly be migrated to other locations in network environment 200. A server connected to a specific leaf router can be changed to connect to a different or additional leaf router. In some cases, some or all of servers 206, hypervisors 208, and/or VMs 210 can represent tenant space. Tenant space can include workloads, services, applications, devices, and/or resources that are associated with one or more clients or subscribers. Accordingly, traffic in network environment 200 can be routed based on specific tenant policies, spaces, agreements, configurations, etc. Moreover, addressing can vary between one or more tenants. In some configurations, tenant spaces can be divided into logical segments and/or networks and separated from logical segments and/or networks associated with other tenants.

An administrator from a device 226 can communicate with one or more application policy infrastructure controllers 218_A, 218_B, 218_C (collectively APICs 218). The administrator can push a policy from their device 226 to one or more of the APICs 218, which can transmit the garbage collection policy to a respective Leaf node or nodes 204 according to the principles disclosed herein to perform garbage collection. A Leaf node 204 can transmit a live process core that is collected and pushed over the cloud 222 (or any network) to a container running on a server 224. The server 224 can run a garbage detection tool on the live process core and send back a list of leaked addresses. The list of leaked addresses can be received at a leaf 204 and the process can “walk” over each address in the list and call a function such as free ( ), or some other function, to perform garbage collection. The details of this process will be set forth in more detail below.

Having introduced the basic ACI environment, the disclosure turns to FIG. 3, which provides an example method embodiment according to the solution and enables a scalable, non-intrusive guard garbage collection method in traditional embedded networking devices. Absent the solution as disclosed herein, it is not possible to provide such a scalable garbage collection approach. One benefit of the approach is that it will work on already deployed network fabrics.

The present approach can run on all processes and according to the solution disclosed herein, an administrator, or a system can pick and choose which processes to apply the method to. In regular garbage collection, a process must be paused during the process of cleaning up memory leaks. According to the present approach, a respective process is not paused during the course snapshot.

In traditional approaches, garbage detection and collection are both done within a process context. According to the disclosed approach, the memory analysis, or garbage detection, is done in the cloud and collection is done within the process context using a special message as disclosed herein. Finally, in traditional garbage collection, the process is computationally expensive and can consume many CPU cycles. The approach disclosed herein is virtually nonintrusive to the running process in that much of the solution is executed in the cloud.

FIG. 3 illustrates an example method embodiment. The example method includes obtaining a list of processes in an application centric infrastructure fabric (302), sorting the list of processes according to an amount of memory increase associated with each respective process in the list of processes to yield a sorted list (304), selecting a group of processes from the sorted list (306) and collecting a respective live process core for each process in the group of processes without pausing or killing any process in the group of processes (308). The method can also include applying a leak detection tool to each process in the group of processes to yield a list of leaked memory addresses for a given process of the group of processes (310) and transmitting a message to the given process with the list of leaked memory addresses, whereby the given process calls a function to release leaked memory associated with the given process as identified in the message (312). The leak detection tool can be an offline tool. An example tool that could be used to perform the step of selecting the top “n” processes to collect the live process core could include proprietary tools such as the one disclosed in U.S. Pat. No. 9,558,055, incorporated herein by reference. Other tools could be utilized as well.

The list of processes can include all processes in the application centric infrastructure fabric or can consist of a group of selected processes from all processes within the application centric infrastructure fabric, according to a criterion. The criteria can be based on a rate of memory leak, a percentage of memory leak relative to some parameter which can be process based or ACI fabric based, cost based, workload based, predictive data based, or according to any other parameter.

In one aspect, the method includes applying the leak detection tool in a container located in a cloud computing environment. In such a scenario, the method can include transmitting the respective live process core to a container running in a cloud computing environment for applying the leak detection tool. The benefit of doing an offline memory leak detection process is that a process core's file should have all the data necessary to detect leaks. The core has the complete memory foot-print of all the processes along with all the memory that were ever allocated by it on the heap.

In another aspect, the method can include transmitting the respective live process core off-line to a container on a same (or different) switch for applying the leak detection tool. Any tool can be used to perform the individual functions disclosed herein. Any memory leak detection tool could be utilized to detect memory leaks. Any tool or process that performs the function of detecting memory leaks could be applied.

Existing technologies involve optimizing an algorithm for native in-memory garbage collection techniques. Garbage collection itself is an expensive operation and more often than not involves pausing the process in order to clean up memory. An example tool for collecting a live process without pausing the process can include Google's Coredumper library (see https://github.com/anatol/google/google-coredumper). Not only does this disclosure make garbage collection possible in C/C++ based legacy applications, but the concepts let the system do the expensive garbage detection outside the ACI fabric (inside a container in the cloud), thereby making the entire operation non intrusive.

Moreover, the concepts disclosed herein can potentially work on solutions that have currently been deployed. The approach is platform and operating system agnostic and can work across any embedded system using standard glibe library for memory allocation.

The concepts disclosed herein allow in-memory garbage collection for native C/C++ based applications with the least possible intrusion to the running system and without pausing or killing the running application. The concept enables a system to execute the expensive garbage detection (data analytics) in a container located in the cloud, in one aspect, and then send in a message in another mode the list of memory addresses that have been leaked to the a respective application process in order to be freed. FIG. 4 illustrates several components 400 including a switch 402 and a message 408 for releasing memory. The message 408 can have a list of leaks and data that could include, for example, a TCP Port identification (such as 4500), a message type (such as msg_memleak_fre) and a payload size (such as 1 Kb). The message could include a leaked memory address such as 0x43416780, for example.

Aspect of this disclosure also include predictive measures that can be developed based on machine learning or artificial intelligence such that memory leak decisions, messages or assignments can be predicted in advance and even transmitted in some cases before the memory issue arises. As shown in FIG. 4. the message 408 could be sent from an operating system switch 402 (Such as Cisco NX-OS switch) having an operating system shell 404 which can be used to transmit a special message 408 to an operating system process 406 such as (a border gateway protocol process).

Applying the approach disclosed herein can result in a sawtooth configuration of the process memory over time, as is shown in the graph 500 of FIG. 5. For example, the process memory may increase over time until an instant when a message is sent to the process to free the memory associated with the memory leak. In response to the message and implementation of the memory release, the process memory would be reduced only to start rising again until the next message is acted on to again release memory. Some of the increase in process memory could be caused by regular operation of the process as well and not due to leakage.

In some embodiments the computer-readable storage devices, mediums, and/or memories can include a cable or wireless signal containing a bit stream and the like. However, when mentioned, non-transitory computer-readable storage media expressly exclude media such as energy, carrier signals, electromagnetic waves, and signals per se.

Methods according to the above-described examples can be implemented using computer-executable instructions that are stored or otherwise available from computer readable media. Such instructions can include, for example, instructions and data which cause or otherwise configure a general purpose computer, special purpose computer, or special purpose processing device to perform a certain function or group of functions. Portions of computer resources used can be accessible over a network. The computer executable instructions may be, for example, binaries, intermediate format instructions such as assembly language, firmware, or source code. Examples of computer-readable media that may be used to store instructions, information used, and/or information created during methods according to described examples include magnetic or optical disks, flash memory, USB devices provided with non-volatile memory, networked storage devices, and so on.

Devices implementing methods according to these disclosures can include hardware, firmware and/or software, and can take any of a variety of form factors. Typical examples of such form factors include laptops, smart phones, small form factor personal computers, personal digital assistants, rackmount devices, standalone devices, and so on. Functionality described herein also can be embodied in peripherals or add-in cards. Such functionality can also be implemented on a circuit board among different chips or different processes executing in a single device, by way of further example.

The instructions, media for conveying such instructions, computing resources for executing them, and other structures for supporting such computing resources are means for providing the functions described in these disclosures.

Although a variety of examples and other information was used to explain aspects within the scope of the appended claims, no limitation of the claims should be implied based on particular features or arrangements in such examples, as one of ordinary skill would be able to use these examples to derive a wide variety of implementations. Further and although some subject matter may have been described in language specific to examples of structural features and/or method steps, it is to be understood that the subject matter defined in the appended claims is not necessarily limited to these described features or acts. For example, such functionality can be distributed differently or performed in components other than those identified herein. Rather, the described features and steps are disclosed as examples of components of systems and methods within the scope of the appended claims. Moreover, claim language reciting “at least one of” a set indicates that one member of the set or multiple members of the set satisfy the claim.

It should be understood that features or configurations herein with reference to one embodiment or example can be implemented in, or combined with, other embodiments or examples herein. That is, terms such as “embodiment”, “variation”, “aspect”, “example”, “configuration”, “implementation”, “case”, and any other terms which may connote an embodiment, as used herein to describe specific features or configurations, are not intended to limit any of the associated features or configurations to a specific or separate embodiment or embodiments, and should not be interpreted to suggest that such features or configurations cannot be combined with features or configurations described with reference to other embodiments, variations, aspects, examples, configurations, implementations, cases, and so forth. In other words, features described herein with reference to a specific example (e.g., embodiment, variation, aspect, configuration, implementation, case, etc.) can be combined with features described with reference to another example. Precisely, one of ordinary skill in the art will readily recognize that the various embodiments or examples described herein, and their associated features, can be combined with each other. For example, while some specific protocols such as 802.11 and 802.3 are mentioned in the examples above, the principles could apply to any communication protocol and does not have to be limited to these particular protocols. Any configuration in which received data is acknowledged through an ACK signal could implement the concepts disclosed herein.

A phrase such as an “aspect” does not imply that such aspect is essential to the subject technology or that such aspect applies to all configurations of the subject technology. A disclosure relating to an aspect may apply to all configurations, or one or more configurations. A phrase such as an aspect may refer to one or more aspects and vice versa. A phrase such as a “configuration” does not imply that such configuration is essential to the subject technology or that such configuration applies to all configurations of the subject technology. A disclosure relating to a configuration may apply to all configurations, or one or more configurations. A phrase such as a configuration may refer to one or more configurations and vice versa. The word “exemplary” is used herein to mean “serving as an example or illustration.” Any aspect or design described herein as “exemplary” is not necessarily to be construed as preferred or advantageous over other aspects or designs.

Moreover, claim language reciting “at least one of” a set indicates that one member of the set or multiple members of the set satisfy the claim. For example, claim language reciting “at least one of A, B, and C” or “at least one of A, B, or C” means A alone, B alone, C alone, A and B together, A and C together, B and C together, or A, B and C together.

Claims

1. A method comprising: collecting a respective live process core for each process in a group of processes without pausing or killing any process in the group of processes;applying a leak detection tool to each process in the group of processes to yield a list of leaked memory addresses for a given process of the group of processes; andtransmitting a message to the given process with the list of leaked memory addresses, whereby the given process calls a function to release leaked memory associated with the given process as identified in the message.

2. The method of claim 1, wherein the group of processes comprises all processes in an application centric infrastructure fabric.

3. The method of claim 1, wherein the group of processes comprises a group of selected processes from all processes within an application centric infrastructure fabric, according to a criteria.

4. The method of claim 1, wherein applying the leak detection tool occurs in a container located in a cloud computing environment.

5. The method of claim 1, further comprising: transmitting the respective live process core off-line to a container on a same switch for applying the leak detection tool.

6. The method of claim 1, further comprising: transmitting the respective live process core to a container running in a cloud computing environment for applying the leak detection tool.

7. A system comprising: at least one processor; anda computer-readable media storing instructions which, when executed by the at least one processor, cause the at least one processor to perform operations comprising: collecting a respective live process core for each process in a group of processes without pausing or killing any process in the group of processes;applying a leak detection tool to each process in the group of processes to yield a list of leaked memory addresses for a given process of the group of processes; andtransmitting a message to the given process with the list of leaked memory addresses, whereby the given process calls a function to release leaked memory associated with the given process as identified in the message.

8. The system of claim 7, wherein the group of processes comprises all processes in an application centric infrastructure fabric.

9. The system of claim 7, wherein the group of processes comprises a group of selected processes from all processes within an application centric infrastructure fabric, according to a criteria.

10. The system of claim 7, wherein applying the leak detection tool occurs in a container located in a cloud computing environment.

11. The system of claim 7, wherein the computer-readable media stores additional instructions which, when executed by the at least one processor, cause the at least one processor to perform operations further comprising: transmitting the respective live process core off-line to a container on a same switch for applying the leak detection tool.

12. The system of claim 7, wherein the computer-readable media stores additional instructions which, when executed by the at least one processor, cause the at least one processor to perform operations further comprising: transmitting the respective live process core to a container running in a cloud computing environment for applying the leak detection tool.

13. A non-transitory computer-readable media storing instructions which, when executed by a processor, cause the processor to perform operations comprising: collecting a respective live process core for each process in a group of processes without pausing or killing any process in the group of processes;applying a leak detection tool to each process in the group of processes to yield a list of leaked memory addresses for a given process of the group of processes; andtransmitting a message to the given process with the list of leaked memory addresses, whereby the given process calls a function to release leaked memory associated with the given process as identified in the message.

14. The non-transitory computer-readable media of claim 13, wherein the group of processes comprises all processes in an application centric infrastructure fabric.

15. The non-transitory computer-readable media of claim 13, wherein the group of processes comprises a group of selected processes from all processes within an application centric infrastructure fabric, according to a criteria.

16. The non-transitory computer-readable media of claim 13, wherein applying the leak detection tool occurs in a container located in a cloud computing environment.

17. The non-transitory computer-readable media of claim 13, wherein the non-transitory computer-readable media stores additional instructions which, when executed by the processor, cause the processor to perform operations further comprising: transmitting the respective live process core off-line to a container on a same switch for applying the leak detection tool.

18. The non-transitory computer-readable media of claim 13, wherein the non-transitory computer-readable media stores additional instructions which, when executed by the processor, cause the processor to perform operations further comprising: transmitting the respective live process core to a container running in a cloud computing environment for applying the leak detection tool.