Abstract: Systems and methods providing object storage gateway for a client device with intermittent connectivity. In one implementation, data objects stored locally on the client device are identified. The data objects are associated with an application running on the client device. Filtered data are produced by filtering the data objects in view of a set of rules. Responsive to determining that a connection between the client device and the server device satisfies a quality condition, the filtered data are synchronized with the server device.

Abstract: Fragmented data on a storage device may be additionally stored in a cache. A cache weight for determining storage of a data unit can be determined. For example, a computing device can receive storage device characteristics from a storage device. A data unit comprising multiple fragments may be stored on the storage device. The computing device can receive data unit characteristics from the storage device. The computing device can determine a cache weight for the data unit. The computing device may output the cache weight for determining storage of the data unit in a cache.

Abstract: An example method may include receiving, in a virtualized execution environment, a data access request from a storage system, identifying, in view of a virtualized execution image associated with the virtualized execution environment, an application running in the virtualized execution environment, generating a cache classification that specifies whether data accessed by the application is suitable for cache compression, including, in the data access request, a tag indicating whether cached data is to be accessed in a compressed-memory cache, wherein the tag is determined in view of the cache classification, and sending, to a server of the storage system, the data access request. The application can be identified in view of metadata included in the virtualized execution image, where the metadata comprises one or more of an application name or an application version.

Abstract: An example method may include identifying one or more requested data items requested by a client system, identifying, in view of one or more requested data items, a plurality of predicted data items and, for each predicted data item, a respective probability that the predicted data item will be requested by a subsequent access request, identifying a plurality of cacheable data items, where the cacheable data items comprise one or more of the predicted data items, wherein each cacheable data item comprises a predicted data item that satisfies caching criteria, wherein the caching criteria are evaluated in view of the respective probability that the predicted data item will be requested by a subsequent access request, and storing each of the plurality of cacheable data items in a cache memory of a respective storage server at which the cacheable data item is located.

Abstract: Systems, methods, and apparatus are disclosed for management of data storage with stream based compressibility. In an example, an orchestrator including a storage controller managing one or more storage nodes executes on one or more processors. The orchestrator receives compression ratio measurements associated with attempts to compress data from a plurality of applications running on a plurality of service guests. The orchestrator receives a request for storing a first data stream from a first application running in a first service guest. Based on a compression ratio of previously stored data associated with the first application being less than a threshold, the orchestrator assigns a first compressibility setting to the first data stream. The first compressibility setting causes a storage node to store the first data stream without attempting to compress the first data stream.

Abstract: Data can be prefetched in a distributed storage system. For example, a computing device can receive a message with metadata associated with at least one request for an input/output operation from a message queue. The computing device can determine, based on the message from the message queue, an additional IO operation predicted to be requested by a client subsequent to the at least one request for the IO operation. The computing device can send a notification to a storage node of a plurality of storage nodes associated with the additional IO operation for prefetching data of the additional IO operation prior to the client requesting the additional IO operation.

Abstract: An example method may include receiving, in a virtualized execution environment, a data access request from a storage system, identifying, in view of a virtualized execution image associated with the virtualized execution environment, an application running in the virtualized execution environment, generating a cache classification that specifies whether data accessed by the application is suitable for cache compression, including, in the data access request, a tag indicating whether cached data is to be accessed in a compressed-memory cache, wherein the tag is determined in view of the cache classification, and sending, to a server of the storage system, the data access request. The application can be identified in view of metadata included in the virtualized execution image, where the metadata comprises one or more of an application name or an application version.

Abstract: Systems and methods are provided for managing write requests for drives in a cloud storage system. For example, a system can receive a plurality of write requests for writing a first set of data to a first drive of a plurality of drives. The first drive may be powered off. The system can write the first set of data to a cache in response to receiving the plurality of write requests. The system can determine that a number of the plurality of write requests exceeds a predetermined write request threshold. The system can power on the first drive in response to determining that the number of the plurality of write requests exceeds the predetermined write request threshold. The system can write the first set of data stored in the cache to the first drive.

Abstract: A technique for improving workload distribution by utilizing unused resources in a distributed computing system is described. In one example of the present disclosure, a system can determine that a computing entity of a distributed computing system includes an unused portion of a CPU capacity. The computing entity can have a first defined limit of the CPU capacity. The system can use the unused portion of the CPU capacity to improve a usage of a resource of the computing entity. The computing entity can have a second defined limit of the resource.

Abstract: Systems and methods are provided for managing write requests for drives in a cloud storage system. For example, a system can receive a plurality of write requests for writing a first set of data to a first drive of a plurality of drives. The first drive may be powered off. The system can write the first set of data to a cache in response to receiving the plurality of write requests. The system can determine that a number of the plurality of write requests exceeds a predetermined write request threshold. The system can power on the first drive in response to determining that the number of the plurality of write requests exceeds the predetermined write request threshold. The system can write the first set of data stored in the cache to the first drive.

Abstract: Systems and methods are provided for managing prefetching operations for read requests for drives in a distributed storage system. For example, a system can determine that a first drive of a plurality of drives is powered on. Prior to receiving a read request for reading a first set of data from the first drive, the system can enable a prefetching operation for prefetching the first set of data from the first drive to be written to a cache. The system may power off the first drive. The system may receive a read request for reading the first set of data from the first drive of a plurality of drives. In response to receiving the read request, the system may read the first set of data from the cache.

Abstract: A container image can be used to determine a caching algorithm for a software application. For example, a storage system can receive a context tag indicating an input/output (IO) pattern associated with a software application of a container. The context tag can be determined based on a container image of the container. The storage system can determine a caching algorithm for the software application based on the context tag. The storage system can apply the caching algorithm to the software application.

Abstract: A technique for dynamic scaling of a distributed computing system is described. In one example of the present disclosure, a system can include a base node configured to provide an access point to a distributed computing system and for servicing a first portion of requests and to generate at least one compute node based on a first load of the base node. The system can also include the at least one compute node of the distributed computing system for servicing a second portion of requests. The at least one compute node can be configured to generate an additional compute node for servicing a subset of the second portion of requests based on a second load of the at least one compute node.

Abstract: Fragmented data on a storage device may be additionally stored in a cache. A cache weight for determining storage of a data unit can be determined. For example, a computing device can receive storage device characteristics from a storage device. A data unit comprising multiple fragments may be stored on the storage device. The computing device can receive data unit characteristics from the storage device. The computing device can determine a cache weight for the data unit. The computing device may output the cache weight for determining storage of the data unit in a cache.

Abstract: Data can be prefetched in a distributed storage system. For example, a computing device can receive a message with metadata associated with at least one request for an input/output operation from a message queue. The computing device can determine, based on the message from the message queue, an additional IO operation predicted to be requested by a client subsequent to the at least one request for the IO operation. The computing device can send a notification to a storage node of a plurality of storage nodes associated with the additional IO operation for prefetching data of the additional IO operation prior to the client requesting the additional IO operation.

Abstract: Stateful workloads executing on temporarily available resources of a cloud computing system can be managed. For example, a computing device can receive, from a scheduled unit executing on a temporarily available resource of a compute node in a cloud computing system, a state measurement indicating a size of a state of the scheduled unit. The computing device can receive, from the compute node, a safe state capacity that is a size of data capable of being saved in response to the temporarily available resource being unavailable. The computing device can determine the state measurement for the scheduled unit exceeds the safe state capacity. In response to determining the state measurement exceeds the safe state capacity, the computing device can transmit a command to the scheduled unit to perform a state-reduction operation.

Abstract: Stateful workloads executing on temporarily available resources of a cloud computing system can be managed. For example, a computing device can receive, from a scheduled unit executing on a temporarily available resource of a compute node in a cloud computing system, a state measurement indicating a size of a state of the scheduled unit. The computing device can receive, from the compute node, a safe state capacity that is a size of data capable of being saved in response to the temporarily available resource being unavailable. The computing device can determine the state measurement for the scheduled unit exceeds the safe state capacity. In response to determining the state measurement exceeds the safe state capacity, the computing device can transmit a command to the scheduled unit to perform a state-reduction operation.