Stanford Seminar - Nanosecond-level Clock Synchronization in a Data Center
Mendel Rosenblum
Stanford University
April 24, 2019
Computers connected by networks have long used protocols like Network Time Protocol (NTP) to adjust the clocks of the computers to show the same time of day. With machines connected by a high performance network in a data center, NTP can keep the clocks within a few hundred microseconds of each other, a clock skew that is fine for humans but not useful for distributed system software that can communicate between machines in a few tens of microseconds. In this talk I will describe the challenge of tightly synchronizing clocks in a data center and present Huygens, a new algorithm developed with my colleague Balaji Prabhakar's research group that can synchronized data center clocks within few 10s of nanoseconds. Like NTP, Huygens requires no special hardware and simply exchanges messages between machines but uses a number of techniques that exploit the high-performance networks in data centers. The techniques include sending many more probe messages and then using sophisticated machine learning techniques and global knowledge to compute the clock adjustment operations to keep the clock synchronized.
View the full playlist: https://www.youtube.com/playlist?list=PLoROMvodv4rMWw6rRoeSpkiseTHzWj6vu
Видео Stanford Seminar - Nanosecond-level Clock Synchronization in a Data Center канала stanfordonline
Stanford University
April 24, 2019
Computers connected by networks have long used protocols like Network Time Protocol (NTP) to adjust the clocks of the computers to show the same time of day. With machines connected by a high performance network in a data center, NTP can keep the clocks within a few hundred microseconds of each other, a clock skew that is fine for humans but not useful for distributed system software that can communicate between machines in a few tens of microseconds. In this talk I will describe the challenge of tightly synchronizing clocks in a data center and present Huygens, a new algorithm developed with my colleague Balaji Prabhakar's research group that can synchronized data center clocks within few 10s of nanoseconds. Like NTP, Huygens requires no special hardware and simply exchanges messages between machines but uses a number of techniques that exploit the high-performance networks in data centers. The techniques include sending many more probe messages and then using sophisticated machine learning techniques and global knowledge to compute the clock adjustment operations to keep the clock synchronized.
View the full playlist: https://www.youtube.com/playlist?list=PLoROMvodv4rMWw6rRoeSpkiseTHzWj6vu
Видео Stanford Seminar - Nanosecond-level Clock Synchronization in a Data Center канала stanfordonline
Показать
Комментарии отсутствуют
Информация о видео
Другие видео канала
Introduction to Precision Time Protocol (PTP)Clock synchronization and Manchester coding | Networking tutorial (3 of 13)Edge Computing PlatformGRCon17 - Symbol Clock Recovery and Improved Symbol Synchronization Blocks - Andy WallsOCPREG19 - Data Center Network Next Generation - presented by FacebookHow an atomic clock works, and its use in the global positioning system (GPS)Cloud Computing FundamentalsThe Surprising Secret of SynchronizationSecurity and Data Protection in a Google Data CenterSpecial Relativity: Simultaneity and Clock SynchronizationDATA CENTRE 101 | DISSECTING a SERVER and its COMPONENTS!Inside a Huge Data Center Filled with Apple Mac ComputersPTP Synchronization : An IntroductionTI Precision Labs - Ethernet: What is an Ethernet PHY?Time, Clocks and Ordering of Events in a Dist. System by Dan Rubenstein [PWL NYC]Creating new worlds: a journey through video game design | Peter Burroughs | TEDxMSUNTP Explained | Network Time Protocol | Cisco CCNA 200-301RHCSA RHEL 8 - Configure time service clientsWhat is Ethernet?