Abstract: The various embodiments described herein include methods, devices, and systems for reading and writing data from a database table. In one aspect, a method includes: (1) initiating a read transaction to read from a first non-key column of a row in the database table, the database table having a plurality of rows, each row comprising a primary key and a plurality of non-key columns, the initiating including: (a) determining that a write transaction holds a lock on a second non-key column of the row in the database table, and (b) determining that no lock is held on the first non-key column; and (2) in response, concurrently reading data from the first non-key column and writing a new column value to the second non-key column; where each non-key column includes a last-write timestamp that indicates when the last write occurred for the respective non-key column.

Abstract: The various embodiments described herein include methods, devices, and systems for reading and writing data from a database table. In one aspect, a method includes: (1) initiating a read transaction to read from a first non-key column of a row in the database table, the database table having a plurality of rows, each row comprising a primary key and a plurality of non-key columns, the initiating including: (a) determining that a write transaction holds a lock on a second non-key column of the row in the database table, and (b) determining that no lock is held on the first non-key column; and (2) in response, concurrently reading data from the first non-key column and writing a new column value to the second non-key column; where each non-key column includes a last-write timestamp that indicates when the last write occurred for the respective non-key column.

Abstract: The various embodiments described herein include methods, devices, and systems for reading and writing data from a database table. In one aspect, a method includes: (1) initiating a read transaction to read from a first non-key column of a row in the database table, the database table having a plurality of rows, each row comprising a primary key and a plurality of non-key columns, the initiating including: (a) determining that a write transaction holds a lock on a second non-key column of the row in the database table, and (b) determining that no lock is held on the first non-key column; and (2) in response, concurrently reading data from the first non-key column and writing a new column value to the second non-key column; where each non-key column includes a last-write timestamp that indicates when the last write occurred for the respective non-key column.

Abstract: The subject matter described herein provides techniques to ensure that queries of a distributed database observe a consistent read of the database without locking or logging. In this regard, next-write timestamps uniquely identify a set of write transactions whose updates can be observed by reads. By publishing the next-write timestamps from within an extendable time lease and tracking a “safe timestamp,” the database queries can be executed without logging read operations or blocking future write transactions, and clients issuing the queries at the “safe timestamp” observe a consistent view of the database as it exists on or before that timestamp. Aspects of this disclosure also provide for extensions, done cheaply and without the need for logging, to the range of timestamps at which read transactions can be executed.

Abstract: The various embodiments described herein include methods, devices, and systems for reading and writing data from a database table. In one aspect, a method of reading and writing data from a database table, includes: (1) initiating a write transaction to write data to a first non-key column of a row of the database table, the database table having a plurality of rows, each row comprising a primary key and a plurality of non-key columns; (2) locking the first non-key column; and (3) in accordance with a determination that the second non-key column is not locked, initiating a read transaction to read data from the second non-key column, where initiation of the read transaction occurs prior to completion of the write transaction.

Abstract: The subject matter described herein provides techniques to ensure that queries of a distributed database observe a consistent read of the database without locking or logging. In this regard, next-write timestamps uniquely identify a set of write transactions whose updates can be observed by reads. By publishing the next-write timestamps from within an extendable time lease and tracking a “safe timestamp,” the database queries can be executed without logging read operations or blocking future write transactions, and clients issuing the queries at the “safe timestamp” observe a consistent view of the database as it exists on or before that timestamp. Aspects of this disclosure also provide for extensions, done cheaply and without the need for logging, to the range of timestamps at which read transactions can be executed.

Abstract: The various embodiments described herein include methods, devices, and systems for reading and writing data from a database table. In one aspect, a method of reading and writing data from a database table, includes: (1) initiating a write transaction to write data to a first non-key column of a row of the database table, the database table having a plurality of rows, each row comprising a primary key and a plurality of non-key columns; (2) locking the first non-key column; and (3) in accordance with a determination that the second non-key column is not locked, initiating a read transaction to read data from the second non-key column, where initiation of the read transaction occurs prior to completion of the write transaction.

Abstract: The various embodiments described herein include methods, devices, and systems for reading and writing data from a database table. In one aspect, a method of reading and writing data from a database table, includes: (1) initiating a write transaction to write data to a first non-key column of a row of the database table, the database table having a plurality of rows, each row comprising a primary key and a plurality of non-key columns; (2) locking the first non-key column; and (3) in accordance with a determination that the second non-key column is not locked, initiating a read transaction to read data from the second non-key column, where initiation of the read transaction occurs prior to completion of the write transaction.

Abstract: The subject matter described herein provides techniques to ensure that queries of a distributed database observe a consistent read of the database without locking or logging. In this regard, next-write timestamps uniquely identify a set of write transactions whose updates can be observed by reads. By publishing the next-write timestamps from within an extendable time lease and tracking a “safe timestamp,” the database queries can be executed without logging read operations or blocking future write transactions, and clients issuing the queries at the “safe timestamp” observe a consistent view of the database as it exists on or before that timestamp. Aspects of this disclosure also provide for extensions, done cheaply and without the need for logging, to the range of timestamps at which read transactions can be executed.

Abstract: The various embodiments described herein include methods, devices, and systems for reading and writing data from a database table. In one aspect, a method of reading and writing data from a database table, includes: (1) initiating a write transaction to write data to a first non-key column of a row of the database table, the database table having a plurality of rows, each row comprising a primary key and a plurality of non-key columns; (2) locking the first non-key column; and (3) in accordance with a determination that the second non-key column is not locked, initiating a read transaction to read data from the second non-key column, where initiation of the read transaction occurs prior to completion of the write transaction.

Abstract: A method reads and writes data from a database table. Each row in the table has a primary key and multiple non-key columns. Each non-key column has one or more column values, and each column value has an associated timestamp that identifies when the column value was stored. The timestamps associated with the column values in each non-key column provide a unique order for the column values. A read transaction is initiated to read from a first non-key column of a first row. A write transaction is in progress that is updating a second non-key column of the first row, where the second non-key column is distinct from the first non-key column. The write transaction holds a lock on the second non-key column of the first row. The method concurrently reads the data from the first non-key column and writes a new column value to the second non-key column.

Abstract: An ad serving system accepts a request for one or more ads, determines a set of eligible ads from among ads having a guaranteed serving reservation and ads with uncommitted serving. This may be done using information from the accepted request, wherein the set of eligible ads may include at least one ad having a guaranteed serving reservation and at least one ad with uncommitted serving. A set of ads to serve is determined from the set of eligible ads using an arbitration process, and the determined set of ads is served. Ad serving guarantees may be generated and/or desired ad serving guarantees may be checked using an inventory test and/or a price test.

Abstract: A method reads and writes data from a database table. Each row in the table has a primary key and multiple non-key columns. Each non-key column has one or more column values, and each column value has an associated timestamp that identifies when the column value was stored. The timestamps associated with the column values in each non-key column provide a unique order for the column values. A read transaction is initiated to read from a first non-key column of a first row. A write transaction is in progress that is updating a second non-key column of the first row, where the second non-key column is distinct from the first non-key column. The write transaction holds a lock on the second non-key column of the first row. The method concurrently reads the data from the first non-key column and writes a new column value to the second non-key column.

Abstract: An on-line advertising system provides advertisements to users. At least one candidate advertisement is identified in response to an advertisement trigger. At least one key is identified relating to features of the candidate advertisement, the advertisement trigger, or a combination of the candidate advertisement and the advertisement trigger. Candidate advertisement performance is estimated based on historical data relating to the key.

Abstract: A method and apparatus for surface approximation without cracks. In one or more embodiments, a surface to be rendered is split into multiple adjacent regions (e.g., subdivision surfaces or patches). A data structure is associated with each boundary edge between two regions for storage of adjacency information in the form of a sequence of tessellation vertices. Adjacency information for a given edge is written into the data structure when an adjacent region is first tessellated. When the remaining adjacent region is tessellated, the adjacency information is read from the data structure and used to achieve tessellation without cracks. In one embodiment, visible cracks due to T-vertices are prevented by forming an overlap of adjacent regions at the location of each T-vertex. This technique allows regions to be tessellated and rendered without any advance knowledge of how the adjacent regions will be split.

Abstract: A method and apparatus for rendering shadows. A pre-rendering process implements a two dimensional array or map of depth-based functions, such as a visibility function in z. During rendering of an object scene, these functions are accessed via lookup operations to efficiently determine the function value for a sample point at a given depth. The use of visibility functions allows for partial light attenuation effects. Each visibility function is computed by filtering multiple transmittance functions obtained by casting sample rays from a light source onto an object scene. The visibility function is implemented as a sequence of vertices. Colored shadows are modeled by vertices comprising a depth value and separate visibility function values for red, green, and blue light at a given depth value. Compression is achieved by minimizing the number of vertices needed to represent a visibility function within a desired error tolerance.

Abstract: A method and apparatus for rendering shadows. A pre-rendering process implements a two dimensional array or map of depth-based functions, such as a visibility function in z. During rendering of an object scene, these functions are accessed via lookup operations to efficiently determine the function value for a sample point at a given depth. The use of visibility functions allows for partial light attenuation effects. Each visibility function is computed by filtering multiple transmittance functions obtained by casting sample rays from a light source onto an object scene. The visibility function is implemented as a sequence of vertices. Colored shadows are modeled by vertices comprising a depth value and separate visibility function values for red, green, and blue light at a given depth value. Compression is achieved by minimizing the number of vertices needed to represent a visibility function within a desired error tolerance.

Abstract: A method and apparatus for surface approximation without cracks. In one or more embodiments, a surface to be rendered is split into multiple adjacent regions (e.g., subdivision surfaces or patches). A data structure is associated with each boundary edge between two regions for storage of adjacency information in the form of a sequence of tessellation vertices. Adjacency information for a given edge is written into the data structure when an adjacent region is first tessellated. When the remaining adjacent region is tessellated, the adjacency information is read from the data structure and used to achieve tessellation without cracks. In one embodiment, visible cracks due to T-vertices are prevented by forming an overlap of adjacent regions at the location of each T-vertex. This technique allows regions to be tessellated and rendered without any advance knowledge of how the adjacent regions will be split.

Abstract: A method and apparatus for rendering shadows. A pre-rendering process implements a two dimensional array or map of depth-based functions, such as a visibility function in z. During rendering of an object scene, these functions are accessed via lookup operations to efficiently determine the function value for a sample point at a given depth. The use of visibility functions allows for partial light attenuation effects. Each visibility function is computed by filtering multiple transmittance functions obtained by casting sample rays from a light source onto an object scene. The visibility function is implemented as a sequence of vertices. Colored shadows are modeled by vertices comprising a depth value and separate visibility function values for red, green, and blue light at a given depth value. Compression is achieved by minimizing the number of vertices needed to represent a visibility function within a desired error tolerance.

Abstract: A method and apparatus for rendering shadows. A pre-rendering process implements a two dimensional array or map of depth-based functions, such as a visibility function in z. During rendering of an object scene, these functions are accessed via lookup operations to efficiently determine the function value for a sample point at a given depth. The use of visibility functions allows for partial light attenuation effects such as partially blocking surfaces, semi-transparent surfaces and volumetric elements, to be accurately modeled over a range of z. Each visibility function is computed by filtering multiple transmittance functions obtained by casting sample rays from a light source onto an object scene. The visibility function is implemented as a sequence of vertices, each comprising a depth (z) value and corresponding function value. Colored shadows are modeled by vertices comprising a depth value and separate visibility function values for red, green and blue light at the given depth value.