Determining Quiesce Timeouts for Containerized Workloads

Abstract

A technique for determining a quiesce timeout for a containerized workload is described, which is used to identify a storage unit for the containerized workload. The technique includes determining an input / output (I / O) rate associated with a containerized workload executing in a container environment, where the containerized workload interfaces with a storage system to store data. The technique further includes determining a quiesce timeout for the containerized workload based at least in part on the I / O rate of data associated with the containerized workload and an amount of memory allocated to buffer the data during execution of a backup operation without causing I / O overflow. The technique further includes determining a storage unit specification that enables execution of the backup operation within the quiesce timeout; and evaluating storage units available to the container environment to identify a storage unit for the containerized workload based on the storage unit specification.

Description

[Background technology]

[0001] The present disclosure relates to container orchestration, and more particularly to mapping containerized workloads to storage units.

[0002] Containers are a way to build, package, and deploy software. Simply put, a container contains both the application code and the dependencies the application code needs to run properly. Multiple containers (also called containerized workloads or containerized applications) can run on the same machine and share an operating system (OS) kernel with other containers, each running as an isolated process in user space. Container orchestration is the automation of the operational tasks required to run containerized workloads or applications and services. These operational tasks include managing the container lifecycle, such as container provisioning, deployment, scaling (up and down), networking, and load balancing. Summary of the Invention

[0003] Aspects of the present disclosure are directed to a computer-implemented method comprising determining an input / output (I / O) rate associated with a containerized workload executing in a container environment, where the containerized workload interfaces with a storage system to store data. Aspects of the present disclosure further comprise determining a quiesce timeout for the containerized workload based at least in part on the I / O rate of data associated with the containerized workload and an amount of memory allocated to buffer the data during execution of a backup operation without causing an I / O overflow. Aspects of the present disclosure further comprise determining a storage unit specification that enables execution of the backup operation within the quiesce timeout; and evaluating storage units available to the container environment to identify a storage unit for the containerized workload based on the storage unit specification.

[0004] Additional aspects of the present disclosure are directed to systems and computer program products configured to perform the methods described above. This Summary is not intended to describe each aspect, every implementation, and / or every embodiment of the present disclosure. [Brief explanation of the drawings]

[0005] The drawings included in this application are incorporated in and constitute a part of this specification. They illustrate embodiments of the present disclosure and, together with the description, serve to explain the principles of the disclosure. The drawings are merely illustrative of particular embodiments and are not intended to limit the disclosure.

[0006] [Figure 1] FIG. 1 is a block diagram illustrating an example container orchestration system, according to some embodiments of the present disclosure.

[0007] [Figure 2]FIG. 1 illustrates a hypothetical scenario for mapping a containerized workload to a storage unit based in part on an estimated quiesce timeout for the containerized workload, in accordance with some embodiments of the present disclosure.

[0008] [Figure 3] FIG. 1 is a flow diagram illustrating an example method for managing storage unit allocation based on quiesce timeouts for containerized workloads, in accordance with some embodiments of the present disclosure.

[0009] [Figure 4] FIG. 1 is a flow diagram illustrating an example method for determining a quiesce timeout for a containerized workload used to identify a storage unit for the containerized workload, in accordance with some embodiments of the present disclosure.

[0010] [Figure 5] FIG. 1 is a block diagram illustrating an exemplary computing environment in which aspects of the present disclosure can be implemented, according to some embodiments of the present disclosure.

[0011] While the present disclosure is susceptible to various modifications and alternative forms, specifics thereof have been shown by way of example in the drawings and will be described in detail below. It should be understood, however, that the intention is not to limit the disclosure to the particular embodiments described. Rather, the intention is to cover all modifications, equivalents, and alternatives falling within the scope of the present disclosure. DETAILED DESCRIPTION OF THE INVENTION

[0012] Aspects of the present disclosure are directed to determining a quiesce timeout for a containerized workload that allows a backup operation to complete successfully, and mapping the containerized workload to a storage unit based at least in part on the quiesce timeout. While not limited to such applications, embodiments of the present disclosure may be better understood in light of the above context.

[0013] Container orchestration automates the deployment, management, scaling, and networking of containerized workloads. Containerization is the packaging of software code with the operating system (OS) libraries and dependencies required to run the software code, thereby producing a single, lightweight executable file, referred to herein as a containerized workload, that runs consistently on common computing platforms.

[0014] Containerized workloads have complex management specifications that become even more complex as they scale. To manage this complexity, container orchestration systems include data stores and file systems that provide a consistent source of truth about the status of containerized workloads at any given time. For example, in the context of the KUBERNETES® container orchestration system, etcd is an open-source distributed key-value store used to hold and manage critical information needed for proper execution, such as configuration data, state data, and metadata, about containerized workloads being managed by the container orchestration system.

[0015] Performing regular backups of file systems used by containerized workloads is important for recovering containerized workloads in the event of a disaster scenario. File system backups can be generated to maintain the contents of a file system at a single point in time. The backups include the state of the containerized workload and other important information. Backup operations (e.g., snapshot creation and / or deletion) provide online backup capabilities that enable recovery from common issues such as accidental deletion of files and comparison with previous versions of files.

[0016] Performing a backup operation requires a brief interruption of access to the file system. The interruption allows the backup process to obtain a consistent point-in-time image of the file system and update various internal databases with the backup information. The mechanism for suspending access to the file system is called quiesce. While a backup operation can take only a few seconds to perform, the process of reaching a quiesced state can take longer due to the operations required to reach the quiesced state. Illustratively, snapshot quiesce works by sending a remote procedure call (RPC) from the backup process to all nodes (virtual or physical computers) that host the file system. On each node, an RPC handler blocks the initiation of new file system operations and waits for currently running file system operations to complete. The backup process also blocks the execution of daemon threads (e.g., housekeeping threads) that access the file system. The backup process then performs the snapshot operation. Afterwards, the backup process sends a resume RPC to the nodes, instructing them to allow file system operations and daemon threads to return to normal operation. The time to reach a quiesced state, perform one or more snapshot operations, and return to normal operation is the snapshot quiesce period.

[0017] The snapshot quiescing process described above becomes problematic in high-frequency containerized workload environments where snapshots are frequently created and deleted. This is because the container orchestration system does not know whether the storage units (units of processing and memory) allocated to the containerized workload are sufficient to buffer the input / output (I / O) of the containerized workload during the snapshot quiescing period. Failure to buffer I / O during the snapshot quiescing period can cause the snapshot operation to fail, resulting in I / O data loss for the containerized workload. This lack of intelligence by the container orchestration system can result in suboptimal scheduling of the containerized workload, causing problems in terms of snapshot failures, backup failures, application downtime, and storage resource inefficiencies.

[0018] Advantageously, aspects of the present disclosure overcome these challenges by determining a quiesce timeout for each containerized workload that enables storage backup operations to be performed without causing I / O overflow, and mapping each containerized workload to a storage unit based at least in part on the quiesce timeout. More specifically, aspects of the present disclosure determine an I / O rate associated with a containerized workload executing in a container environment, where the containerized workload interfaces with a storage system to perform storage operations. Aspects of the present disclosure then determine a quiesce timeout for the containerized workload, where the quiesce timeout is based at least in part on the I / O rate and an amount of memory required to buffer the I / O associated with the containerized workload during backup operations without causing I / O overflow. Aspects of the present disclosure then determine a storage unit specification that enables backup operations to be performed within the quiesce timeout. Aspects of the present disclosure then evaluate storage units available to the container environment and identify a storage unit for the containerized workload based on the storage unit specification. Aspects of the present disclosure can then allocate storage units to the containerized workload, thereby providing the containerized workload with a sufficient amount of computing resources (processing and memory) to allow backup operations to be performed without experiencing I / O overflow. Accordingly, aspects of the present disclosure provide an improvement to computer-related technologies by optimizing the allocation of storage units to the containerized workload using a quiesce timeout calculated for the containerized workload, thereby increasing backup reliability of the associated file system by reducing the number of backup failures.

[0019] Referring now to the figures, FIG. 1 illustrates a block diagram of an exemplary container environment 100 including a container orchestration system 120 configured to determine a quiesce timeout for containerized workloads 112A, 112N (collectively 112, where N may refer to any positive integer representing any number of containerized workloads) and match the containerized workloads 112 to storage units 114A, 114N (collectively 114, where N may refer to any positive integer representing any number of storage units) based on the quiesce timeout, in accordance with some embodiments of the present disclosure. As shown, the container orchestration system 120 is in network communication with multiple nodes 110A, 110N (collectively 110, where N may refer to any positive integer representing any number of nodes). The nodes 110 include virtual or physical machines that host one or more containerized workloads 112. The container orchestration system 120 manages various aspects of the container environment 100, including automated container deployment, scaling, and management.

[0020] The control plane 102 of the container orchestration system 120 includes a collection of modules (processes) used to manage various aspects of the containerized workload lifecycle. The modules may include a storage mapping module 104, a container storage interface module 106, a quiesce timeout module 108, and other modules, as may be appreciated.

[0021] The storage mapping module 104 of the control plane 102 maps (e.g., assigns) containerized workloads 112 to storage units 114. Generally, the storage mapping module 104 selects storage units 114 for newly created or not-yet-scheduled containerized workloads 112. Because containerized workloads 112 may have different requirements, the storage mapping module 104 identifies storage units 114 specific to the requirements of the containerized workload. When scheduling a containerized workload 112, the storage mapping module 104 may assign a storage unit 114 to the containerized workload 112 that corresponds to a quiesce timeout calculated for the containerized workload 112. In some embodiments, as described below in connection with FIG. 3 , before performing a backup operation (e.g., a snapshot operation), the storage mapping module 104 may determine whether the storage unit 114 currently assigned to the containerized workload 112 corresponds to the quiesce timeout for the containerized workload 112, and may perform one or more corrective actions if the currently assigned storage unit 114 does not comply with the quiesce timeout.

[0022] The storage units 114 include heterogeneous data storage devices and systems available for allocation to the containerized workloads 112 managed by the container orchestration system 120. The storage units 114 host file systems that store data associated with processes executed by the containerized workloads 112. The storage units 114 include persistent storage that can be installed in containers (e.g., pods) that include the containerized workloads 112. The storage units 114 can be heterogeneous and include a variety of storage systems and devices, including cloud-based storage (e.g., storage provided by a storage service provider). The computing resources (e.g., processing, memory, storage, and software) used to implement the storage units 114 vary, with some storage units 114 being able to perform storage and storage-related operations more quickly than other storage units 114. For example, a first storage unit 114A may be implemented using computing resources that allow a snapshot operation to be performed within a 10-second period, while a second storage unit 114N may be implemented using computing resources that allow the same snapshot operation to be performed within a 5-second period. The ability of the storage unit 114 to perform a backup operation within the quiesce timeout of the containerized workload 112 depends at least in part on the computing resource specifications of the storage unit 114.

[0023] Accordingly, as part of mapping a storage unit 114 to a containerized workload 112, the storage mapping module 104 identifies a storage unit 114 that corresponds to the quiesce timeout of the containerized workload 112 and schedules the containerized workload 112 to be assigned to the storage unit 114. As an example, the storage mapping module 104 obtains (or requests) a quiesce timeout from a quiesce timeout module 108, which is configured to determine a quiesce timeout for an individual containerized workload 112, as described below. Based on the quiesce timeout for the containerized workload 112, the storage mapping module 104 identifies a storage unit specification (e.g., read / write speed, buffer flush rate, etc.) that enables a backup operation to be performed by the storage unit 114 within the quiesce timeout of the containerized workload 112. The storage mapping module 104 uses the storage unit specification to identify storage units 114 available for allocation to the containerized workload 112. For example, the storage mapping module 104 may query one or more container storage interface modules 106 for storage units 114 that meet the storage unit specification. After identifying the storage units 114 that correspond to the storage unit specification, the storage mapping module 104 schedules the containerized workloads 112 to be assigned to the storage units 114.

[0024] The quiesce timeout module 108 calculates an estimated quiesce timeout for a containerized workload 112 managed by the container orchestration system 120. The quiesce timeout calculated for the containerized workload 112 by the quiesce timeout module 108 is provided to the storage mapping module 104 to enable identification of a storage unit 114 on which a backup operation can be performed within the quiesce timeout of the containerized workload 112. The quiesce timeout for the containerized workload 112 is the amount of time a backup operation (e.g., a snapshot operation) associated with a file system utilized by the containerized workload 112 can execute without causing an I / O overflow (e.g., an overflow of buffered I / O data associated with the containerized workload 112) while the backup operation is executing. The quiesce timeout may include the time to reach a quiesce state and the time to execute the backup operation. An I / O overflow occurs when the containerized workload 112 attempts to write more I / O data to its allocated memory than the memory can hold. When in a quiescent state, I / O data held in memory cannot be written to the file system, which can result in I / O data loss, failed backup operations, application downtime, etc., as a result of attempting to write more I / O data to memory than the memory buffers can hold.

[0025] The quiesce timeout module 108 calculates a quiesce timeout for the containerized workload 112 by determining the amount of time to quiesce a file system utilized by the containerized workload 112, determining an I / O buffer size for storing I / O data related to the containerized workload 112 while performing a backup operation, and determining an I / O rate associated with the containerized workload 112. Based on the amount of time to quiesce the file system, the I / O buffer size, and the I / O rate, the quiesce timeout module 108 calculates the amount of time that I / O data associated with the containerized workload 112 can be buffered without causing an I / O overflow. This amount of time comprises the quiesce timeout for the particular containerized workload 112.

[0026] More specifically, in some embodiments, the quiesce timeout module 108 determines the I / O buffer size for the containerized workload 112 by identifying the amount of computing resources (processing, memory, local storage) available to buffer the containerized workload's I / O data. A specification of the computing resources allocated to buffer the containerized workload's I / O data may be obtained via the control plane 102 of the container orchestration system 120, where the computing resource specification indicates the amount of node 110 resources (processing, memory, local storage) allocated to buffer the containerized workload's I / O data. Illustratively, the container orchestration system 120 allocates the node 110 computing resources to the containerized workload 112 when scheduling (assigning) the containerized workload 112 to the node 110. For example, in the context of a KUBERNETES® container orchestration system, the quiesce timeout module 108 may obtain the computing resource specification from a node object (node ​​representation) associated with the node 110 hosting the containerized workload 112.

[0027] Along with determining the I / O buffer size available to the containerized workload 112, the quiesce timeout module 108 obtains an I / O rate for the containerized workload 112. In some embodiments, the quiesce timeout module 108 analyzes historical workload data 118 associated with the containerized workload 112 to estimate the I / O rate. The I / O rate may be based on historical I / O execution metrics (e.g., I / O operations per second, bandwidth, response time, etc.). Alternatively, in some embodiments, if historical workload data 118 is not available for the containerized workload 112, the quiesce timeout module 108 attempts to identify historical workload data 118 associated with one or more similar containerized workloads (containerized workloads having similar specifications as the particular containerized workload 112). If historical workload data 118 for one or more similar containerized workloads is identified, the quiesce timeout module 108 analyzes the historical workload data 118 to estimate an I / O rate that can be used as part of determining a quiesce timeout for the individual containerized workload 112.

[0028] When historical workload data 118 may not be available for a containerized workload 112 (e.g., a particular application), the quiesce timeout module 108, in some embodiments, determines the I / O rate for the containerized workload 112 by monitoring I / O operations associated with the containerized workload 112. For example, in the context of the KUBERNETES® container orchestration system, the Kubelet (node ​​agent) reports a pod's resource usage as part of the pod status. The quiesce timeout module 108 may analyze the resource usage reported by the Kubelet to determine the I / O rate for the pod.

[0029] In some embodiments, both the historical I / O rate and the current I / O rate may be used to calculate the quiesce timeout for the containerized workload 112. Illustratively, the quiesce timeout module 108 obtains the historical I / O rate from the historical workload data 118 and obtains the current I / O rate through monitoring I / O operations associated with the containerized workload 112. The quiesce timeout module 108 then evaluates (compares) the historical I / O rate and the current I / O rate to estimate an I / O rate for calculating the quiesce timeout for the containerized workload 112.

[0030] As described above, in some embodiments, the quiesce timeout module 108 determines an estimated time to quiesce a file system utilized by the containerized workload 112. The estimated quiesce time is the time that may be required for the file system to reach a quiesced state, in which storage operations associated with the file system are blocked, paused, or suspended. Illustratively, quiescing a file system includes the process of bringing on-disk data of a physical or virtual storage system into a state suitable for backup. Such a process may include flushing dirty memory buffers from the operating system's in-memory cache to storage disks or other higher-level, special-purpose tasks, as well as operations such as blocking the initiation of new storage operations, waiting for currently executing storage operations to complete, and / or pausing some storage-related processes (including housekeeping processes).

[0031] In some embodiments, determining the estimated time to reach a quiescent state may include estimating an amount of time to flush (copy) I / O data from the node's I / O buffer to persistent storage provided by the storage unit 114. The rate at which I / O data can be flushed from the I / O buffer to the storage unit 114 may be determined based on data throughput rates for the components involved (e.g., processor, memory, network hardware, etc.). Also, in some embodiments, determining the estimated time to reach a quiescent state may include estimating an amount of time to complete pending storage operations associated with the containerized workload 112. The estimated time may be based on typical processing times associated with storage operations performed on behalf of the containerized workload 112. For example, a containerized workload 112 that handles large amounts of data may be associated with longer processing times compared to a containerized workload 112 that handles small amounts of data.

[0032] After collecting the backup-related information described above, the quiesce timeout module 108 calculates a quiesce timeout for the containerized workload 112 based at least in part on the backup-related information (I / O buffer size, I / O rate, and estimated file system quiesce time). Illustratively, the calculation may include summing the amount of time it takes for the file system to reach a quiesce state and the amount of time the containerized workload's I / O data can buffer at the node 110 before reaching capacity, as determined by the I / O rate. The resulting quiesce timeout indicates the amount of time a file system backup operation can be successfully performed without experiencing a buffer overflow of the containerized workload I / O data. Once the quiesce timeout for the containerized workload 112 has been calculated, the storage mapping module 104 can use the quiesce timeout to map the containerized workload 112 to a storage unit 114 having a storage unit specification indicating that the storage unit 114 is capable of performing a backup operation within the quiesce timeout.

[0033] As previously described, the storage mapping module 104 identifies a storage unit 114 that corresponds to the quiesce timeout of the containerized workload 112 and schedules the containerized workload 112 to be assigned to the storage unit 114. In the event that the storage mapping module 104 is unable to identify a storage unit 114 having specifications that correspond to the quiesce timeout of the containerized workload 112, the storage mapping module 104 may allocate additional memory on the node 110 to the containerized workload 112 to enable buffering of the containerized workload's I / O during backup operation processing, thereby potentially preventing I / O overflow while the backup operation is running.

[0034] In some embodiments, the storage mapping module 104 may evaluate the current allocation of storage units 114 to the containerized workload 112 to determine whether the storage unit 114 specifications correspond to a quiesce timeout calculated for the containerized workload 112. In some embodiments, the evaluation may be performed in response to a backup operation error associated with the containerized workload 112. In the event that the storage unit specifications do not allow for the backup operation to be performed within the quiesce timeout, the storage mapping module 104 may identify an alternative storage unit 114 that is capable of performing the backup operation within the quiesce timeout and migrate the containerized workload 112 to the alternative storage unit 114. For example, the storage mapping module 104 may determine that the currently allocated storage unit 114 has insufficient computing resources to allow for the backup operation to be successfully completed within the quiesce timeout and migrate the containerized workload 112 to another storage unit 114 identified as capable of performing the backup operation within the quiesce timeout.

[0035] 1 illustrates that a network 116 is provided to enable communication between components of the container environment 100. The network 116 may include any useful computing network, including an intranet, the Internet, a local area network, a wide area network, a wireless data network, or any other such network or combination thereof. The components utilized for the network 116 may depend at least in part on the type of network and / or environment selected. Wired or wireless connections and combinations thereof may enable communication over the network 116.

[0036] In this illustrative example, container environment 100 includes components that may be implemented as hardware, such as the hardware shown in computing environment 100 in FIG. 5. For example, container orchestration system 120 may be implemented as software, hardware, firmware, or a combination thereof. When software is used, the operations performed by container orchestration system 120 may be implemented in program instructions configured to execute on hardware, such as a processor unit. When firmware is used, the operations performed by container orchestration system 120 may be implemented as program instructions and data and stored in persistent memory for execution on a processor unit. When hardware is utilized, the hardware may include circuitry that operates to perform the operations of container orchestration system 120.

[0037] Furthermore, the hardware may take the form of at least one of a circuit system, an integrated circuit, an application specific integrated circuit (ASIC), a programmable logic device, or any other suitable type of hardware configured to perform a plurality of operations. In a programmable logic device, the device may be configured to perform a plurality of operations. The device may be later reconfigured or may be permanently configured to perform a plurality of operations. Programmable logic devices include, for example, programmable logic arrays, programmable array logic, field programmable logic arrays, field programmable gate arrays, and other suitable hardware devices.

[0038] As used herein, a processor is a hardware device that is comprised of hardware circuitry, such as integrated circuits, that responds to and processes instructions and program instructions to operate a computer. A processor may be implemented using processor set 510 in FIG. 5. A processor may be one or more processor units on the same computer or on different computers. That is, processes may be distributed among processor units on the same or different computers within container environment 100. Furthermore, the multiple processor units may be the same type or different types of processor units. For example, the multiple processor units may be selected from at least one of a single-core processor, a dual-core processor, a multi-processor core, a general-purpose central processing unit (CPU), a graphics processing unit (GPU), a digital signal processor (DSP), or some other type of processor unit.

[0039] In some embodiments, the modules described in connection with FIG. 1 can be implemented as computing services hosted in a computing service environment. For example, a module can be considered a service involving one or more processes running on a server or other computer hardware. Such a service can provide a service application that receives requests and provides output to other services or consumer devices. An API can be provided for each module to enable a first module to send requests to and receive output from a second module. Such an API can also enable third parties to interface with the module to make requests and receive output from the module.

[0040] Although Figure 1 illustrates one example of a container environment in which the above techniques can be implemented, many other similar or different environments are possible. The illustration of a container environment in Figure 1 is not meant to imply physical or architectural limitations to the manner in which an example embodiment may be implemented. Other components in addition to or in place of those shown may be used. Some components may be unnecessary. Also, blocks are presented to illustrate some functional components. When implemented in an example embodiment, one or more of these blocks may be combined, divided, or combined and then divided into different blocks.

[0041] FIG. 2 illustrates a hypothetical scenario for mapping containerized workloads 202A, 202B, 202N (collectively 202, where N may refer to any positive integer representing any number of containerized workloads) to storage units 206A, 206B, 206N (collectively 206, where N may refer to any positive integer representing any number of storage units) based in part on an estimated quiesce timeout calculated for the containerized workload 202, in accordance with some embodiments of the present disclosure.

[0042] In a first hypothetical scenario, containerized workload 202A is currently assigned to storage unit 206A (mount) that has computing resources capable of performing a backup of the file system utilized by containerized workload 202A within 15 seconds. Container orchestration system 204 determines (based on incoming I / O rates and memory buffer sizes) that the quiesce timeout for containerized workload 202A is approximately 10 seconds, which is not enough time for storage unit 206A to reliably perform the backup operation. In response, before performing the file system backup, container orchestration system 204 migrates containerized workload 202A to storage unit 206N, which has computing resources capable of reliably performing the backup operation within the 5 seconds corresponding to the quiesce timeout, because the backup can be performed without causing an I / O overflow. Migrating containerized workload 202A includes copying data stored on storage unit 206A associated with containerized workload 202A to storage unit 206N and pointing (mounting) containerized workload 202A to storage unit 206N.

[0043] In a second hypothetical scenario, containerized workload 202B is currently assigned to storage unit 206B, which can perform a backup operation within 20 seconds. In response to determining that the quiesce timeout for containerized workload 202B is approximately 18 seconds (which is less than the time required to ensure that storage unit 206B performs the backup), container orchestration system 204 migrates containerized workload 202B to storage unit 206A (which can perform a backup operation within 15 seconds) before performing a file system backup to ensure the backup is successful.

[0044] In a third hypothetical scenario, containerized workload 202N is currently assigned to storage unit 206N, which can perform a backup operation within 5 seconds. Container orchestration system 204 determines that the quiesce timeout for containerized workload 202N is approximately 10 seconds. Because the time to perform a backup with storage unit 206N is significantly less than what is required to meet the quiesce time of containerized workload 202N, and to balance the storage load of containerized workload 202 across storage units 206, container orchestration system 204 migrates containerized workload 202N to storage unit 206A, which can perform a backup operation within the quiesce timeout of containerized workload 202N.

[0045] 3 is a flow diagram illustrating an example method 300 for managing storage unit allocation based on quiesce timeouts for containerized workloads, according to some embodiments of the present disclosure. At operation 302, method 300 begins by scheduling a backup operation of a file system utilized by containerized workloads executing in a container environment. As part of scheduling the backup operation, method 300 determines whether the backup operation can be successfully executed. The success of the backup operation is based in part on whether a backup (e.g., a snapshot) can be created and / or deleted without incurring I / O overflow associated with one or more containerized workloads utilizing the file system. To aid in determining whether the backup operation can be successfully executed, method 300 calculates a quiesce timeout for each containerized workload utilizing the file system and evaluates storage units allocated to the containerized workloads to determine whether the storage unit specifications correspond to the containerized workload's quiesce timeout.

[0046] More specifically, at operation 304, method 300 determines the existence of historical data for the containerized workload (historical workload data 118 in FIG. 1 ). The historical data may include historical I / O execution metrics associated with the containerized workload. If historical data for the containerized workload is available, then at operation 306, method 300 uses the historical data to determine a quiesce timeout for the containerized workload. For example, the historical data may be used to estimate a current I / O rate for the containerized workload, and the estimated I / O rate and a specification of a memory buffer available for storing I / O associated with the containerized workload during execution of a backup operation may be used to calculate a quiesce timeout for the containerized workload.

[0047] If historical data is not available for the containerized workload, then in operation 308, method 300 determines a quiesce timeout for the containerized workload using a current (e.g., real-time or near real-time) I / O rate associated with the containerized workload and a specification of a memory buffer available for storing I / O associated with the containerized workload while the backup operation is running. For example, the current I / O rate may be determined by monitoring I / O operations associated with the containerized workload. Resource usage of the containerized workload may be provided to the container orchestration system by a node agent running on the node hosting the containerized workload, and method 300 may use the resource usage information to determine the I / O rate for the containerized workload.

[0048] Alternatively, in some embodiments, if historical data is not available for each containerized workload, method 300 may attempt to identify historical data associated with one or more similar containerized workloads having specifications similar to those of the respective containerized workload. If one or more similar containerized workloads are identified, method 300 analyzes the historical data for the similar containerized workloads to estimate I / O rates to use in calculating the quiesce timeout for each containerized workload.

[0049] After determining the quiesce timeout for the containerized workload, in operation 310, method 300 determines whether a backup operation can be successfully performed based on the quiesce timeout. That is, method 300 determines whether the storage unit assigned to the containerized workload is capable of performing a file system backup within the period indicated by the quiesce timeout. Determining whether the storage unit is capable of performing a backup within the quiesce timeout can be determined via evaluation of the storage unit's specifications (e.g., read / write speeds, buffer flush rate, etc.), as previously described in connection with FIG. 2.

[0050] If it is determined that the backup operation can be successfully performed by the storage unit currently assigned to the containerized workload, then method 300 continues with operation 320 to determine whether there are additional containerized workloads to evaluate. If it is determined that the backup operation cannot be successfully performed by the storage unit, method 300 continues with operation 312, which attempts to identify another storage unit that can perform the backup operation. For example, as previously described in connection with FIG. 2, various storage units in a container environment may be available for allocation to containerized workloads. Method 300 may evaluate storage unit specifications (e.g., read / write speeds, buffer flush rates, etc.) to identify alternative storage units that can perform the backup operation within the quiesce timeout of the containerized workload.

[0051] If an alternative storage unit is found at operation 314, then at operation 318, method 300 migrates the containerized workload from the originally assigned storage unit determined to have insufficient computing resources to enable backup processing within the containerized workload's quiesce timeout to the alternative storage unit. Migrating the containerized workload may include copying data associated with the containerized workload from disk storage of the originally assigned storage unit to disk storage of the alternative storage unit and pointing the containerized workload to the alternative storage unit. Method 300 then continues at operation 320 to determine whether there are additional containerized workloads to evaluate.

[0052] Returning again to operation 314, if an alternative storage unit cannot be found for the containerized workload, then in operation 316, the container orchestration system performs memory scaling to increase the amount of memory of the node allocated to buffering the containerized workload's I / O. The increase in allocated memory may be an amount that prevents I / O overflow while performing the backup operation. Method 300 then continues with operation 320 to determine whether there are additional containerized workloads to evaluate. After performing the above operations for each containerized workload that utilizes the file system, method 300 continues with operation 322 to allow a backup operation to be scheduled.

[0053] 4 is a flow diagram illustrating an example method 400 for determining a quiesce timeout for a containerized workload that is used to identify a storage unit for the containerized workload, in accordance with some embodiments of the present disclosure. In some embodiments, the container orchestration system automates the allocation of the storage unit to enable creation of a snapshot within the quiesce timeout for the containerized workload.

[0054] Beginning at operation 402, method 400 determines an I / O rate associated with a containerized workload executing in a container environment, where the containerized workload interfaces with a storage system to store data. In some embodiments, determining the I / O rate for the containerized workload includes obtaining a historical I / O rate for the containerized workload and estimating a current I / O rate for the containerized workload based on the historical I / O rate. In some embodiments, determining the I / O rate associated with the containerized workload includes monitoring current I / O operations associated with the containerized workload to determine the I / O rate.

[0055] At operation 404, method 400 determines a quiesce timeout for the containerized workload based at least in part on the I / O rate of data associated with the containerized workload and the amount of memory allocated for buffering data while performing backup operations without causing I / O overflow. In some embodiments, determining the quiesce timeout includes estimating the time to complete pending file system operations associated with the containerized workload and estimating the time to copy data from the memory allocated for buffering to persistent storage (e.g., a storage unit allocated to the containerized workload).

[0056] In operation 406, method 400 determines a storage unit specification that enables the backup operation to be performed within the quiesce timeout, and in operation 408, method 400 evaluates storage units available to the container environment to identify a storage unit for the containerized workload based on the storage unit specification.

[0057] In some embodiments, method 400 migrates a containerized workload to a storage unit corresponding to the storage unit specification. Migrating the containerized storage unit includes moving the containerized workload from an originally assigned storage unit determined to have insufficient computing resources to enable snapshot processing within a quiesce timeout to a storage unit identified as having a storage unit specification that enables backup operations to be performed within the quiesce timeout.

[0058] In some embodiments, in response to determining that a storage unit available to the container environment does not comply with a storage unit specification that enables creation of a snapshot within the quiesce timeout, method 400 allocates additional memory to the containerized workload to enable buffering of I / O associated with the containerized workload during backup processing.

[0059] The methods described above may be performed by a computer (e.g., computer 501 in FIG. 5), performed in a cloud environment (e.g., cloud 506 or 505 in FIG. 5), and / or generally implemented in fixed function hardware, configurable logic, logic instructions, etc., or any combination thereof.

[0060] In some alternative implementations of the exemplary embodiments, the function or functions noted in the blocks may occur out of the order noted in the figures. For example, in some cases, two blocks shown in succession may be executed substantially concurrently, or the blocks may even be executed in reverse order, depending on the functionality involved. Also, other blocks may be added in addition to the blocks noted in the flowcharts or block diagrams.

[0061] Various aspects of the present disclosure are described through text, flowcharts, block diagrams of computer systems, and / or block diagrams of machine logic included in computer program product (CPP) embodiments. For any flowchart, depending on the technology involved, operations may be performed in an order different from that shown in a given flowchart. For example, again depending on the technology involved, two operations shown in successive flowchart blocks may be performed in the reverse order, as a single integrated step, simultaneously, or in an at least partially overlapping manner.

[0062] A computer program product embodiment ("CPP embodiment" or "CPP") is a term used in this disclosure to describe any set of one or more storage media (also referred to as "media"), collectively contained in one or more storage devices, that collectively contain machine-readable code corresponding to instructions and / or data for performing the computer operations specified in a given CPP claim. A "storage device" is any tangible device that can hold and store instructions for use by a computer processor. The computer-readable storage medium may be, but is not limited to, an electronic storage medium, a magnetic storage medium, an optical storage medium, an electromagnetic storage medium, a semiconductor storage medium, a mechanical storage medium, or any suitable combination of the foregoing. Some known types of storage devices that include these media include diskettes, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), compact disk read-only memory (CD-ROM), digital versatile disk (DVD), memory stick, floppy disk, mechanically encoded device (such as punch cards or pits / lands formed on the major surface of a disk), or any suitable combination of the foregoing. Computer-readable storage media, as the term is used in this disclosure, is not to be construed as storage in the form of a transitory signal per se, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through a waveguide, light pulses passing through fiber optic cables, electrical signals communicated over wires, and / or other transmission media. As will be appreciated by those skilled in the art, data is typically moved at some infrequent time during the normal operation of a storage device, such as during access, defragmentation, or garbage collection, but this does not qualify a storage device as transitory since the data is not transitory while it is stored.

[0063] 5 , computing environment 500 includes an example of an environment for execution of at least some of the computer code involved in performing the disclosed methods, such as block 550, which includes container orchestration code for determining a quiesce timeout for a containerized workload used to identify a storage unit for the containerized workload. In addition to block 550, computing environment 500 includes, for example, a computer 501, a wide area network (WAN) 502, an end user device (EUD) 503, a remote server 504, a public cloud 505, and a private cloud 506. In this embodiment, computer 501 has a set of processors 510 (including processing circuitry 520 and cache 521), a communications fabric 511, volatile memory 512, persistent storage 513 (including operating system 522 and block 550, as identified above), a set of peripheral devices 514 (including a user interface (UI), a set of devices 523, storage 524, and a set of Internet of Things (IoT) sensors 525), and a network module 515. Remote server 504 includes a remote database 530. The public cloud 505 includes a gateway 540, a cloud orchestration module 541, a set of host physical machines 542, a set of virtual machines 543, and a set of containers 544.

[0064] Computer 501 may take the form of a desktop computer, a laptop computer, a tablet computer, a smartphone, a smartwatch or other wearable computer, a mainframe computer, a quantum computer, or any other form of computer or mobile device now known or later developed that is capable of executing programs, accessing a network, or querying a database, such as remote database 530. As is well understood in the field of computer technology, and depending on the technology, execution of a computer-implemented method may be distributed among multiple computers and / or among multiple locations. However, in this presentation of computing environment 500, to keep the presentation as concise as possible, the detailed discussion focuses on a single computer, specifically computer 501. Computer 501 may be located in a cloud, although it is not depicted within the cloud in FIG. 5 . However, computer 501 is not required to reside in a cloud except to any extent that may be expressly indicated.

[0065] The processor set 510 includes one or more computer processors of any type now known or later developed. The processing circuitry 520 may be distributed across multiple packages, e.g., multiple tailored integrated circuit chips. The processing circuitry 520 may implement multiple processor threads and / or multiple processor cores. The cache 521 is memory located within the processor chip package and is typically used for data or code that should be available for fast access by threads or cores executing on the processor set 510. Cache memory is typically organized into multiple levels depending on relative proximity to the processing circuitry. Alternatively, some or all of the cache for a processor set may be located “off-chip.” In some computing environments, the processor set 510 may be designed to operate with qubits and perform quantum computing.

[0066] Computer-readable program instructions are typically loaded onto computer 501 and cause a series of operational steps to be performed by processor set 510 of computer 501, thereby performing a computer-implemented method, whereby the instructions so executed instantiate the method specified in the computer-implemented method flowcharts and / or narrative descriptions contained herein (collectively referred to as the "disclosed method"). These computer-readable program instructions are stored in various types of computer-readable storage media, such as cache 521 and other storage media discussed below. The computer-readable program instructions and associated data are accessed by processor set 510 to control and direct the execution of the disclosed method. In computing environment 500, at least some of the instructions that perform the disclosed method may be stored in block 550 in persistent storage 513.

[0067] Communications fabric 511 is the signal-conducting pathway that allows various components of computer 501 to communicate with one another. Typically, this fabric is made up of switches and conductive pathways, such as those that make up buses, bridges, physical input / output ports, etc. Other types of signal communication pathways may be used, such as fiber optic and / or wireless communication pathways.

[0068] Volatile memory 512 may be any type of volatile memory, now known or later developed. Examples include dynamic type random access memory (RAM) or static type RAM. Typically, volatile memory is characterized by random access, although this is not required unless expressly indicated. In computer 501, volatile memory 512 is located in a single package and is internal to computer 501; however, alternatively or additionally, volatile memory may be distributed across multiple packages and / or located external to computer 501.

[0069] Persistent storage 513 is any form of non-volatile storage for a computer, now known or developed in the future. The non-volatility of this storage means that stored data remains regardless of whether power is supplied to computer 501 and / or directly to persistent storage 513. Persistent storage 513 can be read-only memory (ROM), but typically at least a portion of persistent storage allows data to be written, data to be deleted, and data to be rewritten. Some well-known forms of persistent storage include magnetic disks and solid-state storage devices. Operating system 522 can take several forms, such as various known proprietary operating systems or open-source Portable Operating System Interface-type operating systems that employ a kernel. The code contained in block 550 typically includes at least some of the computer code involved in performing the disclosed methods.

[0070] Peripheral device set 514 includes a set of peripheral devices of computer 501. Data communication connections between peripheral devices and other components of computer 501 may be implemented in various ways, such as Bluetooth connections, near field communication (NFC) connections, connections formed by cables (such as universal serial bus (USB)-type cables), insertion-type connections (e.g., Secure Digital (SD) cards), connections formed through local area communication networks, and even connections formed through wide area networks such as the Internet. In various embodiments, UI device set 523 may include components such as display screens, speakers, microphones, wearable devices (such as goggles and smartwatches), keyboards, mice, printers, touchpads, game controllers, and haptic devices. Storage 524 may be external storage, such as an external hard drive, or insertable storage, such as an SD card. Storage 524 may be persistent and / or volatile. In some embodiments, storage 524 may take the form of a quantum computing storage device for storing data in the form of qubits. In embodiments where computer 501 is required to have a large amount of storage (e.g., where computer 501 stores and manages a large database locally), in which case this storage may be provided by a peripheral storage device designed to store very large amounts of data, such as a storage area network (SAN) shared by multiple, geographically distributed computers. IoT sensor set 525 consists of sensors that can be used in Internet of Things applications. For example, one sensor may be a thermometer and another sensor may be a motion detector.

[0071] The network module 515 is a collection of computer software, hardware, and firmware that enables the computer 501 to communicate with other computers over the WAN 502. The network module 515 may include hardware such as a modem or Wi-Fi signal transceiver, software for packetizing and / or depacketizing data for communication network transmission, and / or web browser software for communicating data over the Internet. In some embodiments, the network control and network forwarding functions of the network module 515 are performed on the same physical hardware device. In other embodiments (e.g., embodiments utilizing software-defined networking (SDN)), the control and forwarding functions of the network module 515 are performed on physically separate devices, such that the control function manages several different network hardware devices. Computer-readable program instructions for performing the disclosed methods can typically be downloaded to the computer 501 from an external computer or external storage device through a network adapter card or network interface included in the network module 515.

[0072] WAN 502 is any wide area network (e.g., the Internet) capable of communicating computer data over non-local distances by any technology for communicating computer data, now known or later developed. In some embodiments, a WAN may be replaced and / or supplemented by a local area network (LAN) designed to communicate data between devices located in a local area, such as a Wi-Fi network. WANs and / or LANs typically include computer hardware such as copper transmission cables, optical fiber transmissions, wireless transmissions, routers, firewalls, switches, gateway computers, and edge servers.

[0073] End-user device (EUD) 503 is any computer system used and controlled by an end user (e.g., a customer of the enterprise operating computer 501) and may take any of the forms discussed above in connection with computer 501. EUD 503 typically receives useful and useful data from the operation of computer 501. For example, in the hypothetical case where computer 501 is designed to provide recommendations to the end user, the recommendations would typically be communicated from computer 501's network module 515 over WAN 502 to EUD 503. In this manner, EUD 503 can display or otherwise present the recommendations to the end user. In some embodiments, EUD 503 may be a client device such as a thin client, a heavy client, a mainframe computer, a desktop computer, and the like.

[0074] Remote server 504 is any computer system that provides at least some data and / or functionality to computer 501. Remote server 504 may be controlled and used by the same entity that operates computer 501. Remote server 504 represents a machine that collects and stores useful and useful data for use by other computers, such as computer 501. For example, in the hypothetical case where computer 501 is designed and programmed to provide recommendations based on historical data, then this historical data may be provided to computer 501 from remote database 530 of remote server 504.

[0075] A public cloud 505 is any computer system available for use by multiple entities that provides on-demand availability of computer system resources and / or other computer functionality, particularly data storage (cloud storage) and computing power, without direct active management by users. Cloud computing typically leverages resource sharing to achieve coherence and economies of scale. Direct active management of the computing resources of the public cloud 505 is performed by the computer hardware and / or software of a cloud orchestration module 541. The computing resources provided by the public cloud 505 are typically implemented by virtual computing environments running on various computers comprising a host physical machine set 542, which is the universe of physical computers within and / or available in the public cloud 505. The virtual computing environments (VCEs) typically take the form of virtual machines from a virtual machine set 543 and / or containers from a container set 544. It is understood that these VCEs may be stored as images and may be transferred among and between various physical machine hosts, either as images or after instantiation of the VCEs. Cloud orchestration module 541 manages the transfer and storage of images, deploys new instantiations of VCE, and manages active instantiations of VCE deployments. Gateway 540 is a collection of computer software, hardware, and firmware that enables public cloud 505 to communicate over WAN 502.

[0076] Some further description of a virtualized computing environment (VCE) is now provided. A VCE can be stored as an "image." A new active instance of a VCE can be instantiated from the image. Two well-known types of VCE are virtual machines and containers. A container is a VCE that uses operating system-level virtualization. This refers to a feature of an operating system in which the kernel allows the existence of multiple isolated user space instances called containers. These isolated user space instances typically behave as actual computers from the perspective of programs running within them. A computer program running on a typical operating system can utilize all of the computer's resources, such as connected devices, files and folders, network shares, CPU power, and quantifiable hardware capabilities. However, a program running inside a container can only use the contents of the container and of the devices assigned to the container; this feature is known as containerization.

[0077] Private cloud 506 is similar to public cloud 505, except that the computing resources are available only for use by a single enterprise. While private cloud 506 is shown in communication with WAN 502, in other embodiments, the private cloud may be completely disconnected from the Internet and accessible only through a local / private network. A hybrid cloud is a composite of multiple clouds of different types (e.g., private, community, or public cloud types), often implemented by different vendors. While each of the multiple clouds remains a separate, discrete entity, the larger hybrid cloud architecture is bound together by standardized or proprietary technologies that enable orchestration, management, and / or data / application portability between the constituent clouds. In this embodiment, both public cloud 505 and private cloud 506 are part of a larger hybrid cloud.

[0078] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the various embodiments. As used herein, the singular forms "a," "an," and "the" are intended to include the plural forms unless the context clearly dictates otherwise. It should be further understood that the terms "includes" and / or "including," when used herein, specify the presence of stated features, integers, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof. In the preceding detailed description of exemplary embodiments of various embodiments, reference is made to the accompanying drawings (where like numbers represent like elements), which form a part of this specification, and in which are shown, by way of illustration, specific exemplary embodiments in which various embodiments may be practiced. These embodiments have been described in sufficient detail to enable those skilled in the art to practice the embodiments; however, other embodiments may be used, and logical, mechanical, electrical, and other changes may be made without departing from the scope of the various embodiments. In the foregoing description, numerous specific details have been set forth to provide a thorough understanding of the various embodiments. However, various embodiments may be practiced without these specific details. In other instances, well-known circuits, structures and techniques have not been shown in detail in order not to obscure the embodiments.

[0079] As used within this specification, different instances of the word "embodiment" do not necessarily refer to the same embodiment, but they can. Any data and data structures shown or described herein are merely examples; other embodiments may use different amounts of data, types of data, fields, number and types of fields, field names, number and types of columns, records, entries, or organizations of data. Furthermore, any data may be combined with logic such that a separate data structure may not be required. Therefore, the preceding detailed description should not be taken in a limiting sense.

[0080] While the description of various embodiments of the present disclosure has been presented for purposes of illustration, it is not intended to be exhaustive or limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope of the described embodiments. The terminology used herein has been selected to best explain the principles of the embodiments, practical applications, or technical improvements to the art found in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

[0081] While the present disclosure has been described with respect to specific embodiments, it is expected that variations and modifications thereof will become apparent to those skilled in the art, and it is therefore intended that the following claims be interpreted to cover all such variations and modifications that fall within the scope of the present disclosure.

[0082] Any advantages discussed in this disclosure are exemplary advantages, and there may be embodiments of the disclosure that achieve all, some, or none of any of the discussed advantages while remaining within the scope of the disclosure.

Claims

1. determining an input / output (I / O) rate associated with a containerized workload executing in a container environment, wherein the containerized workload interfaces with a storage system to store I / O data; determining a quiesce timeout for the containerized workload based at least in part on the I / O rate associated with the containerized workload and an amount of memory allocated to buffer the I / O data during execution of a backup operation without causing an I / O overflow; determining a storage unit specification that enables the backup operation to be performed within the quiesce timeout; and evaluating storage units available to the container environment and identifying a storage unit for the containerized workload based on the storage unit specification.

1. A computer-implemented method comprising:

2. Determining the I / O rate associated with the containerized workload includes: obtaining a historical I / O rate for the containerized workload; The computer-implemented method of claim 1 , further comprising:

3. Determining the I / O rate associated with the containerized workload includes: monitoring current I / O operations associated with the containerized workload to determine the I / O rate; The computer-implemented method of claim 1 , further comprising:

4. determining the quiesce timeout, estimating a first time for completing pending file system operations associated with the containerized workload; and estimating a second time for copying the I / O data from the memory allocated for a buffer to persistent storage. The computer-implemented method of claim 1 , further comprising:

5. migrating the containerized workload to the storage unit corresponding to the storage unit specification, wherein the containerized workload is moved from an originally assigned storage unit that was determined to have insufficient computing resources to enable backup processing within the quiesce timeout. The computer-implemented method of claim 1 further comprising:

6. determining that the storage units available to the container environment do not comply with the storage unit specifications to allow creation of a backup within the quiesce timeout; and allocating additional memory to the containerized workload to allow buffering of the I / O data associated with the containerized workload during the backup operation. The computer-implemented method of claim 1 further comprising:

7. The computer-implemented method of claim 1 , wherein a container orchestration system automates the allocation of the storage unit to enable creation of a backup within the quiesce timeout.

8. one or more computer-readable storage media storing program instructions; and one or more processors that, in response to execution of the program instructions, determining an input / output (I / O) rate associated with a containerized workload executing in a container environment, wherein the containerized workload interfaces with a storage system to store I / O data; determining a quiesce timeout for the containerized workload based at least in part on the I / O rate associated with the containerized workload and an amount of memory allocated to buffer the I / O data while performing a backup operation without causing an I / O overflow; determining a storage unit specification that enables the backup operation to be performed within the quiesce timeout; and Evaluating storage units available to the container environment and identifying a storage unit for the containerized workload based on the storage unit specification. one or more processors configured to A system comprising:

9. The program instructions configured to cause the one or more processors to determine the I / O rate associated with the containerized workload may include causing the one or more processors to: Monitoring current I / O operations associated with the containerized workload to determine the I / O rate. The system of claim 8 further configured to:

10. The program instructions configured to cause the one or more processors to determine the I / O rate associated with the containerized workload may include causing the one or more processors to: Obtaining historical I / O rates for the containerized workload The system of claim 8 further configured to:

11. The program instructions configured to cause the one or more processors to determine the quiesce timeout may include causing the one or more processors to: Estimating a time to complete pending file system operations associated with the containerized workload; and Estimate the time to copy I / O data from the memory allocated to the containerized workload to storage managed by the file system. The system of claim 8 further configured to:

12. The program instructions may cause the one or more processors to: Migrating the containerized workload to the storage unit corresponding to the storage unit specification, wherein the containerized workload is moved from an originally assigned storage unit that was determined to have insufficient computing resources to enable backup processing within the quiesce timeout. The system of claim 8 further configured to:

13. The program instructions may cause the one or more processors to: determining that the storage units available to the container environment do not comply with the storage unit specifications to allow creation of a backup within the quiesce timeout; and Allocating additional memory to the containerized workload to allow buffering of the I / O data associated with the containerized workload during the backup operation. The system of claim 8 further configured to:

14. The system of claim 8 , wherein a container orchestration system automates the allocation of the storage unit to enable creation of a backup within the quiesce timeout.

15. one or more computer-readable storage media and program instructions collectively stored on the one or more computer-readable storage media, the program instructions causing one or more processors to: determining an input / output (I / O) rate associated with a containerized workload executing in a container environment, wherein the containerized workload interfaces with a storage system to store I / O data; determining a quiesce timeout for the containerized workload based at least in part on the I / O rate associated with the containerized workload and an amount of memory allocated to buffer the I / O data during execution of a backup operation without causing an I / O overflow; determining a storage unit specification that enables the backup operation to be performed within the quiesce timeout; and Evaluating storage units available to the container environment and identifying a storage unit for the containerized workload based on the storage unit specification. A computer program product configured to:

16. The program instructions configured to cause the one or more processors to determine the I / O rate associated with the containerized workload may include causing the one or more processors to: Monitoring current I / O operations associated with the containerized workload to determine the I / O rate.

16. The computer program product of claim 15, further configured to:

17. The program instructions configured to cause the one or more processors to determine the I / O rate associated with the containerized workload may include causing the one or more processors to: Obtaining historical I / O rates for the containerized workload 16. The computer program product of claim 15, further configured to:

18. The program instructions configured to cause the one or more processors to determine the quiesce timeout may include causing the one or more processors to: Estimating a time to complete pending file system operations associated with the containerized workload; and Estimate the time to copy I / O data from the memory allocated to the containerized workload to storage managed by the file system.

16. The computer program product of claim 15, further configured to:

19. The program instructions may cause the one or more processors to: Migrating the containerized workload to the storage unit corresponding to the storage unit specification, wherein the containerized workload is moved from an originally assigned storage unit that was determined to have insufficient computing resources to enable backup processing within the quiesce timeout.

16. The computer program product of claim 15, further configured to:

20. The program instructions may cause the one or more processors to: determining that the storage units available to the container environment do not comply with the storage unit specifications to allow creation of a backup within the quiesce timeout; and and allocating additional memory to the containerized workload to allow buffering of the I / O data associated with the containerized workload during the backup operation.

16. The computer program product of claim 15, further configured to:

Images