公开(公告)号：US20240086327A1

公开(公告)日：2024-03-14

申请号：US18234633

申请日：2023-08-16

Applicant: Google LLC

Inventor： Balaram Sinharoy ,  Peter Hochschild

IPC: G06F12/0842 ,  G06F12/0891

CPC classification number: G06F12/0842 ,  G06F12/0891

Abstract: The present disclosure provides for automatically detecting errors, such as SDC, in a multi-core computing environment. For example, cores may run in an error detection mode, in which multiple cores repeat the same execution of instructions and the results are compared. Based on the results, it may be determined whether one of the cores is failing.

公开(公告)号：US20230216921A1

公开(公告)日：2023-07-06

申请号：US18174094

申请日：2023-02-24

Applicant: Google LLC

Inventor： Wilson Cheng-Yi Hsieh ,  Peter Hochschild

IPC: G06F12/084 ,  G06F9/52

CPC classification number: G06F12/084 ,  G06F9/526 ,  G06F2212/251

Abstract: Throughput is preserved in a distributed system while maintaining concurrency by pushing a commit wait period to client commit paths and to future readers. As opposed to servers performing commit waits, the servers assign timestamps, which are used to ensure that causality is preserved. When a server executes a transaction that writes data to a distributed database, the server acquires a user-level lock, and assigns the transaction a timestamp equal to a current time plus an interval corresponding to bounds of uncertainty of clocks in the distributed system. After assigning the timestamp, the server releases the user-level lock. Any client devices, before performing a read of the written data, must wait until the assigned timestamp is in the past.

公开(公告)号：US12041128B2

公开(公告)日：2024-07-16

申请号：US18174094

申请日：2023-02-24

Applicant: Google LLC

Inventor： Wilson Cheng-Yi Hsieh ,  Peter Hochschild

IPC: H04L67/1097 ,  G06F9/52 ,  G06F12/084 ,  G06F16/176 ,  G06F16/23 ,  H04L67/01 ,  H04L67/1095

CPC classification number: H04L67/1097 ,  G06F9/526 ,  G06F12/084 ,  G06F16/1767 ,  G06F16/2329 ,  H04L67/01 ,  H04L67/1095 ,  G06F2212/251

Abstract: Throughput is preserved in a distributed system while maintaining concurrency by pushing a commit wait period to client commit paths and to future readers. As opposed to servers performing commit waits, the servers assign timestamps, which are used to ensure that causality is preserved. When a server executes a transaction that writes data to a distributed database, the server acquires a user-level lock, and assigns the transaction a timestamp equal to a current time plus an interval corresponding to bounds of uncertainty of clocks in the distributed system. After assigning the timestamp, the server releases the user-level lock. Any client devices, before performing a read of the written data, must wait until the assigned timestamp is in the past.

公开(公告)号：US11953938B1

公开(公告)日：2024-04-09

申请号：US17740688

申请日：2022-05-10

Applicant: Google LLC

Inventor： Peter Hochschild ,  Alexander Lloyd ,  Wilson Cheng-Yi Hsieh ,  Robert Edman Felderman ,  Michael James Boyer Epstein

IPC: G06F1/12 ,  G01S19/01 ,  G06F11/16 ,  H04J3/06 ,  H04L7/00 ,  H04L43/106 ,  H04L67/10

CPC classification number: G06F1/12 ,  G01S19/01 ,  H04J3/0661 ,  H04J3/0667 ,  H04L7/0008 ,  H04L7/0012 ,  H04L7/0016 ,  H04L43/106 ,  H04L67/10 ,  G06F11/1675

Abstract: The present technology proposes techniques for generating globally coherent timestamps. This technology may allow distributed systems to causally order transactions without incurring various types of communication delays inherent in explicit synchronization. By globally deploying a number of time masters that are based on various types of time references, the time masters may serve as primary time references. Through an interactive interface, the techniques may track, calculate and record data relative to each time master thus providing the distributed systems with causal timestamps.

公开(公告)号：US10775837B1

公开(公告)日：2020-09-15

申请号：US15877845

申请日：2018-01-23

Applicant: Google LLC

Inventor： Peter Hochschild ,  Alexander Lloyd ,  Wilson Cheng-Yi Hsieh ,  Robert Edman Felderman ,  Michael James Boyer Epstein

IPC: G06F1/12 ,  H04L29/08 ,  H04L12/26 ,  H04L7/00 ,  G01S19/01 ,  G06F11/16

Abstract: The present technology proposes techniques for generating globally coherent timestamps. This technology may allow distributed systems to causally order transactions without incurring various types of communication delays inherent in explicit synchronization. By globally deploying a number of time masters that are based on various types of time references, the time masters may serve as primary time references. Through an interactive interface, the techniques may track, calculate and record data relative to each time master thus providing the distributed systems with causal timestamps.

公开(公告)号：US10042881B1

公开(公告)日：2018-08-07

申请号：US15358428

申请日：2016-11-22

Applicant: Google LLC

Inventor： Wilson Cheng-Yi Hsieh ,  Alexander Lloyd ,  Peter Hochschild ,  Michael James Boyer Epstein ,  Sean Quinlan

IPC: G06F17/00 ,  G06F7/00 ,  G06F17/30 ,  G06F9/46

Abstract: The present technology proposes techniques for ensuring globally consistent transactions. This technology may allow distributed systems to ensure the causal order of read and write transactions across different partitions of a distributed database. By assigning causally generated timestamps to the transactions based on one or more globally coherent time services, the timestamps can be used to preserve and represent the causal order of the transactions in the distributed system. In this regard, certain transactions may wait for a period of time after choosing a timestamp in order to delay the start of any second transaction that might depend on it. The wait may ensure that the effects of the first transaction are not made visible until its timestamp is guaranteed to be in the past. This may ensure that a consistent snapshot of the distributed database can be determined for any past timestamp.

公开(公告)号：US11601501B2

公开(公告)日：2023-03-07

申请号：US17189646

申请日：2021-03-02

Applicant: Google LLC

Inventor： Wilson Cheng-Yi Hsieh ,  Peter Hochschild

IPC: H04L67/1097 ,  H04L67/01 ,  H04L67/1095 ,  G06F16/23 ,  G06F16/176

Abstract: Throughput is preserved in a distributed system while maintaining concurrency by pushing a commit wait period to client commit paths and to future readers. As opposed to servers performing commit waits, the servers assign timestamps, which are used to ensure that causality is preserved. When a server executes a transaction that writes data to a distributed database, the server acquires a user-level lock, and assigns the transaction a timestamp equal to a current time plus an interval corresponding to bounds of uncertainty of clocks in the distributed system. After assigning the timestamp, the server releases the user-level lock. Any client devices, before performing a read of the written data, must wait until the assigned timestamp is in the past.

公开(公告)号：US11442925B1

公开(公告)日：2022-09-13

申请号：US16023326

申请日：2018-06-29

Applicant: Google LLC

Inventor： Wilson Cheng-Yi Hsieh ,  Alexander Lloyd ,  Peter Hochschild ,  Michael James Boyer Epstein ,  Sean Quinlan

IPC: G06F16/23 ,  G06F16/14 ,  G06F9/46 ,  G06F16/2458

Abstract: The present technology proposes techniques for ensuring globally consistent transactions. This technology may allow distributed systems to ensure the causal order of read and write transactions across different partitions of a distributed database. By assigning causally generated timestamps to the transactions based on one or more globally coherent time services, the timestamps can be used to preserve and represent the causal order of the transactions in the distributed system. In this regard, certain transactions may wait for a period of time after choosing a timestamp in order to delay the start of any second transaction that might depend on it. The wait may ensure that the effects of the first transaction are not made visible until its timestamp is guaranteed to be in the past. This may ensure that a consistent snapshot of the distributed database can be determined for any past timestamp.

公开(公告)号：US20250030763A1

公开(公告)日：2025-01-23

申请号：US18748656

申请日：2024-06-20

Applicant: Google LLC

Inventor： Wilson Cheng-Yi Hsieh ,  Peter Hochschild

IPC: H04L67/1097 ,  G06F9/52 ,  G06F12/084 ,  G06F16/176 ,  G06F16/23 ,  H04L67/01 ,  H04L67/1095

Abstract: Throughput is preserved in a distributed system while maintaining concurrency by pushing a commit wait period to client commit paths and to future readers. As opposed to servers performing commit waits, the servers assign timestamps, which are used to ensure that causality is preserved. When a server executes a transaction that writes data to a distributed database, the server acquires a user-level lock, and assigns the transaction a timestamp equal to a current time plus an interval corresponding to bounds of uncertainty of clocks in the distributed system. After assigning the timestamp, the server releases the user-level lock. Any client devices, before performing a read of the written data, must wait until the assigned timestamp is in the past.

公开(公告)号：US11789938B1

公开(公告)日：2023-10-17

申请号：US17875780

申请日：2022-07-28

Applicant: Google LLC

Inventor： Wilson Cheng-Yi Hsieh ,  Alexander Lloyd ,  Peter Hochschild ,  Michael James Boyer Epstein ,  Sean Quinlan

IPC: G06F16/23 ,  G06F16/14 ,  G06F9/46 ,  G06F16/2458

CPC classification number: G06F16/2379 ,  G06F9/466 ,  G06F16/148 ,  G06F16/2322 ,  G06F16/2477

Abstract: The present technology proposes techniques for ensuring globally consistent transactions. This technology may allow distributed systems to ensure the causal order of read and write transactions across different partitions of a distributed database. By assigning causally generated timestamps to the transactions based on one or more globally coherent time services, the timestamps can be used to preserve and represent the causal order of the transactions in the distributed system. In this regard, certain transactions may wait for a period of time after choosing a timestamp in order to delay the start of any second transaction that might depend on it. The wait may ensure that the effects of the first transaction are not made visible until its timestamp is guaranteed to be in the past. This may ensure that a consistent snapshot of the distributed database can be determined for any past timestamp.