Alibaba Group Holding Limited

Ping Zhou

Shu Li

Yu Du

One embodiment facilitates flash storage management. During operation, the system receives a request to write data to a non-volatile memory. The system writes, in a buffer in association with a controller, the data to a next available portion of a current physical page in a page stripe, wherein the page stripe includes physical pages which each correspond to a sequentially ordered die of the non-volatile memory. The system provides power-loss protection to the current physical page and a second physical page which indicates parity information for the page stripe, thereby providing power-loss protection to an increased number of partially full page stripes.

System and method for flash storage management using multiple open page stripes

Resource management includes: operating in a first mode, including executing a controller operation in connection with a plurality of local elements, wherein the controller operation is executed by a configurable resource; switching from the first mode to a second mode; and operating in the second mode, including: executing a data processing task in a first portion of the configurable resource; and executing the controller operation in connection with the plurality of local elements, in a second portion of the configurable resource.

High-volume, low-latency data processing in flexibly configured local heterogeneous computing environments

One embodiment facilitates atomicity assurance for storing data and metadata in a data stream. The system receives a first stream of data to be written to a storage device, wherein the first stream includes a plurality of I/O requests associated with data and corresponding metadata. In response to determining that residual data associated with a preceding I/O request of the first stream exists in a data buffer: the system appends, to the residual data, a first portion of data from a current I/O request to obtain a first page of data; the system writes a remainder portion of the current I/O request to the data buffer to obtain current residual data; and the system writes the first page of data to the storage device. Thus, the system thereby facilitates atomicity assurance for storing the data and corresponding metadata of each I/O request of the first data stream.

Method and system for facilitating atomicity assurance on metadata and data bundled storage

Disclosed are systems, methods, and apparatuses for providing a high-speed data path to storage devices. In one embodiment, a method is disclosed comprising receiving, by the processor, a data access command, the data access command specifying a location in memory to access data; issuing, by the processor, the data access command to the storage device via a first datapath, the first datapath comprising a non-block datapath; and accessing, by the processor, the non-volatile storage component through the first datapath and the memory, wherein the non-volatile storage component of the storage device is mapped to memory accessible by the processor.

Rapid side-channel access to storage devices

One embodiment facilitates global data placement in a storage device. The system receives a request to write first data. The system selects one of a plurality of physical media of the storage device to which to directly write the first data, based on a frequency of access, a block size, and a latency requirement of the first data, wherein the plurality of physical media includes a fast cache medium, a solid state drive, and a hard disk drive. In response to determining that the frequency of access of the first data is greater than a predetermined threshold, or in response to determining that the block size associated with the first data is not greater than a predetermined size and determining that the first data is an update to existing data, the system selects the fast cache medium. The system writes the first data to the selected one physical medium.

System and method for optimization of global data placement to mitigate wear-out of write cache and NAND flash

Using persistent memory for key-value storage is disclosed, including: receiving a put key operation including a key and a corresponding value; directly storing the key at a persistent memory media using a physical address that is determined based at least in part on the key, wherein the persistent memory media supports memory persistence and in-place updating; directly storing the value to the persistent memory media; receiving a get key operation including the key; and directly accessing the persistent memory media using the physical address that is determined based at least in part on the key to retrieve the value corresponding to the key.

Persistent memory for key-value storage

One embodiment facilitates mitigating write amplification in a phase change memory-based storage device. During operation, the system receives, by a controller of the storage device, data to be stored in a phase change memory (PCM) of the storage device. The system writes, by the controller, the data to a sector of a volatile memory of the same storage device to obtain an updated portion of the sector. In response to detecting a predetermined condition, the system writes the updated portion to the PCM.

Method and system for mitigating write amplification in a phase change memory-based storage device

One embodiment facilitates data destruction in a phase change memory-based storage device. During operation, the system detects, by the storage device, a power loss. Subsequent to the power loss, the system overwrites keys and configuration information used to transform data stored in the phase change memory, wherein the keys and the configuration information are stored in the phase change memory, thereby preventing the data from being recovered.

Method and system for data destruction in a phase change memory-based storage device

One embodiment described herein provides a data storage system. The storage system includes a communication interface for receiving a write command from a host of the data storage system and a first write cache coupled to the host via the communication interface. The write command includes to-be-written data and address information associated with the to-be-written data. The first write cache is configured to temporarily store the to-be-written data before sending the to-be-written data to a high-capacity storage device coupled to the host, and the first write cache is configured to update its content by performing byte-level in-place writing based on the received write command.

Method and system for implementing byte-alterable write cache

One embodiment of the present invention provides a system for a fine-grained power management. The system receives, by a first server, a task assigned based on a global power state, wherein the first server includes one or more drives, a drive includes one or more channels, and a channel includes one or more integrated circuits. The system places an inactive drive into a power-saving mode. The system places an inactive channel of an active drive into the power-saving mode by using a power switch associated with each integrated circuit in the inactive channel. The system places an inactive integrated circuit of an active channel into the power-saving mode by using a power switch associated with the inactive integrated circuit. The system updates a data structure storing the global power state based on a current power state of the first server, the drives, the channels, and the integrated circuits.

Growing community of inventors

Folsom, CA, United States of America

Ping Zhou