Micron Technology Incorporated

Crossroads Systems, Inc.

David Boles

Alexander Tomlinson

Steven Moyer

John M Groves

Neelima Premsankar

Gaurav Sanjay Ramdasi

Nabeel Meeramohideen Mohamed

Gregory Becker

Greg A Becker

Sudheer Kurichiyath

Tony M Brewer

David Andrew Roberts

Jacob Mulamootil Jacob

Jack Basiago

H Bret Young

A prefix probe component receives a request to perform a cursor operation to search for one or more data elements of a key-value data store, the request comprising a key identifier associated with the one or more data elements, and wherein the key-value data store comprises a tree structure with a plurality of nodes; traverses a portion of the plurality of nodes to identify data elements in the key-value data store that match the key identifier; determines whether a number of the data elements that match the key identifier satisfies a threshold condition; and responsive to determining that the number of data elements satisfies the threshold condition, performs the cursor operation for the data elements that match the key identifier.

Prefix probe for cursor operations associated with a key-value database system

Systems and techniques for reducing probabilistic filter query latency are described herein. A query for a probabilistic filter that is stored on a first media may be received from a caller. In response to receiving the query, cached segments of the probabilistic filter stored on a second media may be obtained. Here, the probabilistic filter provides a set membership determination that is conclusive in a determination that an element is not in a set. The query may be executed on the cached segments resulting in a partial query result. Retrieval of remaining data of the probabilistic filter from the first media to the second media may be initiated without intervention from the caller. Here, the remaining data corresponds to the query and data that is not in the cached segment. The partial query results may then be returned to the caller.

Reducing probabilistic filter query latency

A global lock is used to access a first set of data structures. An active transaction having a transaction start identifier is identified as a globally oldest active transaction associated with the first set of data structures. A first marker value of a first data structure of a second set of data structures is compared to the transaction start identifier to determine satisfaction of a first condition. In response to satisfying the first condition, the first data structure is accessed to identify a first set of data locks associated with one or more transactions each having a transaction completion identifier that satisfies a second condition when compared to the transaction start identifier. In response to satisfying the second condition, the first set of data locks is released.

Lock release management associated with a key-value database system

Systems, methods, and computer readable media for tracking memory deltas at a cache line granularity. The method includes receiving a base address for a physical memory region, receiving a list of empty log memory buffers associated with a delta logging session, and responsive to determining that a cache line associated with the physical memory region may be in a modified state, storing the modified cache line and metadata associated with the modified cache line in an active log memory buffer referenced by the list of empty log memory buffers. The method also includes determining that the active log memory buffer is full and appending a flag to the active log memory buffer, thereby marking the active log memory buffer as a full log memory buffer. The method also includes storing a list of full log memory buffers, wherein the list is visible to a host processor.

Tracking memory modifications at cache line granularity

A first request to initiate a memory access transaction is received by a processing device of a memory sub-system from a host system. One or more host data items are received from the host system. The one or more host data items are stored in a memory buffer residing on a volatile memory device of the memory sub-system. In response to an initiation of one or more operations to commit the memory access transaction, the one or more data items are migrated from the memory buffer to a persistent memory device.

Enabling memory access transactions for persistent memory

Disclosed in some examples are methods, systems, memory devices, and machine-readable mediums that allows an application thread to indicate an undo logging operation when calculations are beginning that may need to be rolled back if a crash or other failure occurs. During the undo logging operation, memory writes an identified memory are done to a copy and the original value is preserved. If the undo logging operation is committed, then the copy becomes the correct value and may then be subsequently used in place of the original, or the value stored in the copy is copied to the original. If the undo logging operation is abandoned, the copy is not preserved and the value goes back to the original.

Undo capability for memory devices

An example method of cursor seek operations using deleted record spans in memory sub-systems comprises: receiving, by a processing device, a request to search for a first data element of a key-value data store, the request specifying a first key associated with the first data element; determining that the first key is within a range of deleted data elements comprising a plurality of deleted data elements of the key-value data store; identifying a tail key associated with a tail element of the range of deleted data elements; identifying a second data element in the key-value data store, wherein the second data element is associated with a second key that follows the tail key in a specified order of keys; and providing the second data element in response to the request.

Cursor seek operations using deleted record spans

A first data structure lock to access a first data structure of a first set of data structures to perform an operation associated with a transaction is acquired. The operation associated with the transaction is executed, wherein the operation is one of inserting the transaction into the first data structure or removing the transaction from the first data structure. An oldest active transaction of the first data structure is identified. A globally oldest active transaction of the set of the data structures in view of the oldest active transaction is determined. A second set of data structures is accessed, the second set of data structures including information associated with completed transactions to identify a set of data locks associated with completed transactions each having a transaction completion identifier that satisfies a condition when compared to a transaction start identifier associated with the globally oldest active transaction. The set of data locks are released.

Lock management associated with a key-value database system

A key-value pair comprising a key and a value associated with the key is stored in a current segment of a data container. The key comprises a segment identifier identifying the current segment. The data container comprises a plurality of segments residing on a non-volatile memory device. Responsive to determining that a size of the current segment exceeds a threshold, a new segment is added to the container. The plurality of key-value pairs having respective current segment identifiers referencing one or more least recently created segments of the data container is deleted.

Using prefix-delete operations for data containers

Aspects of the present disclosure provide for operations of a key-value store (KVS) that uses journaling with storage of data (e.g., key data or value data of a key-value pair) in different types of data storage objects (e.g., data storage formats) selected based on one or more criteria.

Key-value store using journaling with selective data storage format

A KVS tree database and operations thereon are described herein. A KVS tree database is a multilevel tree that includes a base level and subsequent levels. The base level includes a heterogeneous kvset in a node, The heterogeneous kvset including entries for multiple KVS trees, such as a first entry for a first KVS tree and a second entry for a second KVS tree, The subsequent level includes a first node including a first homogeneous kvset for the first KVS tree and a second node including a second homogeneous kvset for the second KVS tree. Here, a homogeneous kvset includes nodes from only one KVS tree. The KVS tree database also includes a first determinative mapping of entries between the base level and the subsequent level and a second determinative mapping of entries between subsequent levels.

Growing community of inventors

Austin, TX, United States of America

David Boles