Time Appliances Project: Difference between revisions
Line 140: | Line 140: | ||
:- [https://ocp-all.groups.io/g/OCP-TAP/attachment/2/0/nsdi18-geng.pdf Stanford Paper] for reference only | :- [https://ocp-all.groups.io/g/OCP-TAP/attachment/2/0/nsdi18-geng.pdf Stanford Paper] for reference only | ||
:- [https://static.googleusercontent.com/media/research.google.com/en//archive/spanner-osdi2012.pdf Spanner Paper] for reference only | :- [https://static.googleusercontent.com/media/research.google.com/en//archive/spanner-osdi2012.pdf Spanner Paper] for reference only | ||
:- [https://arxiv.org/pdf/2002.04269.pdf On Time Synchronization Issues in Time-Sensitive Networks with Regulators and Nonideal Clocks] for reference only, raised by Elad | :- [https://arxiv.org/pdf/2002.04269.pdf On Time Synchronization Issues in Time-Sensitive Networks with Regulators and Nonideal Clocks] for reference only, raised by Elad | ||
:- [https://drive.google.com/file/d/11MFIk9WWJ-cmjj8uWifKQnefEKvdchZE/view?usp=sharing Accurate Network Clock Synchronization at Scale] for reference only, raised by Hesham | :- [https://drive.google.com/file/d/11MFIk9WWJ-cmjj8uWifKQnefEKvdchZE/view?usp=sharing Accurate Network Clock Synchronization at Scale] for reference only, raised by Hesham |
Revision as of 22:45, 24 November 2020
Welcome
- Welcome to the OCP Time Appliances Project wiki.
- This Project is open to the public and we welcome all those who would like to be involved.
- Time is a key element to get the highest efficiency in a distributed system. The performance of a distributed system depends on the synchronization of its elements. Several industries such as telecom, mobile, power, industrial, professional audio and video and many more have embraced the need for highly accurate and more importantly reliable distribution and synchronization of time and frequency across packet networks. Although the use case scenario for each of the industries is different, they all share one common thing and that is, time synchronization. Since there is a diverse need for time synchronization across different industries, driven from different use cases and applications, managing the needs of this industry chain becomes a challenge.
Time Appliances Project (TAP) aims to provide a platform to bring together, discuss, standardize and share technologies and solutions across industries with the datacenter applications and datacenter network infrastructure as the main interest. The project aims to bring together the community of datacenter operators, application developers, and equipment and semiconductor companies together to enable datacenter time-sensitive applications such as consistency in distributed systems, edge computing, AR/VR and IoT. These applications will greatly benefit from high accuracy, reliable, and scalable distribution and synchronization of time.
IEEE 1588 Precision Timing Protocol (PTP) and other synchronization methods have been adopted by various industries to maximize the efficiency of various distributed system use cases. Each use case scenario comes with a set of requirements and configurations. These configurations are collected as a ‘PTP profile’. Time appliances project aims to support the development of a PTP profile for datacenter applications and datacenter network infrastructure. The profile will cover time-sensitive applications over OCP-compliant and PTP-aware networking infrastructure such as network switches, network clocks, network interface cards, timing modules & connectors, etc. Additionally, the profile will address various requirements for high accuracy and reliable distribution and synchronization of time, such as expected performance, networking, software API, data models, deployment and telemetry. The project also aims at openness and interoperability through the use of open-source PTP software implementations for timing appliances.
Datacenter applications are the primary target of time appliances project. In addition, the project extends to related topics on time synchronization in academia, research and other industries. The time appliances project brings together the community and will be highly collaborative through technical meetings and collaboration with other OCP Projects including the Networking, Storage, Server, and Telco Projects.
- Disclaimer: Please do not submit any confidential information to the Project Community. All presentation materials, proposals, meeting minutes and/or supporting documents are published by OCP and are open to the public in accordance to OCP's Bylaws and IP Policy. This can be found on the OCP OCP Policies page. If you have any questions please contact OCP.
Mission Statement
- 1. Create specifications and references for Data Center Timing appliances, applications and networking infrastructure
- 2. Promote openness in Timing Appliances and interfaces through open-source implementations
Project Leadership
Co-leads:
- - Ahmad Byagowi (Facebook)
- - Elad Wind (NVIDIA)
Interim IC Rep:
- - Rajeev Sharma (OCP)
Workstreams
Project | Objective | POC | Meeting Schedule | |
---|---|---|---|---|
#1 | Open Source Grandmaster | Development of an open source grandmaster for DC and Edge systems | Oleg Obleukhov & Dotan Levi |
discussed during project call |
#2 | Data Center PTP Profile | Development of a PTP Profile tailored for data center applications | Michel Ouellette | discussed during project call |
#3 | Precision Time API | Time APIs to disseminate the time error (error bound) and bring accurate time to the user space | Georgi Chalakov | discussed during project call |
Overall | Project meeting coordinator | Kelvin Chukwu | please send all inquiries to Kelvin |
Get Involved
Documents
- - OCP Presentation Template - please contact Michael Schill or Archna Haylock for a copy
Release Candidates (RC) Docs
Name | Format | Version | Date | Contributor | Link |
---|---|---|---|---|---|
Open Source Grandmaster | MD | v0.1 | Nov-23, 2020 | OCP TAP | GitHub |
Data Center PTP Profile | MD | v0.1 | Nov-23, 2020 | OCP TAP | GitHub |
TAP GitHub
TAP Google Drive
Regular Project Calls
- - Wednesdays at 11am PST, starting on July 15th, 2020. Repeats every 2 weeks following that
- - Please contact Kelvin Chukwu to propose any agenda items
Join the meeting from your computer, tablet or smartphone:
https://global.gotomeeting.com/join/565185493
Dial in:
United States (Toll Free): +1 877 309 2073 United States: +1 571 317-3129 Access Code: 565-185-493
Recordings from Past Calls
- - #1 - July 15th, 2020 - Proposal to launch OCP-TAP Incubation Project
- - #2 - July 29th, 2020
- - #3 - August 12th, 2020 - TAP Vision
- - #4 - August 26th, 2020 - Open Grandmaster workstream goals and draft spec
- - #5 - September 9th, 2020 - PTP workstream goals and draft spec
- - #6 - September 23rd, 2020 - Practical Use Cases of Synchronized Clocks
- - #7 - October 7th, 2020 - Timing card implementation
- - #8 - October 21st, 2020 - Starter's guide to ptp4l
- - #9 - November 4th, 2020 - Time Sync in TSNs
- - #10 - November 18, 2020 - Computer Timekeeping and Synchronization
Presentations & Videos
- - OCP TAP - Vision
- - TAP Presentation
- - Impact of Oscillator Noise on PTP Time Error – Part 1
- - Impact of Oscillator Noise on PTP Time Error – Part 2
- - ELPROMA TAP Remarks Oct 7 2020
- - Introduction to PTP Oct 21 2020
- - Time Sync in TSN Nov 4 2020
- - Computer Timekeeping and Synchronization Nov 18 2020
References & External Links
- - Spanner, TrueTime & The CAP Theorem by Eric Brewer, Google
- - Practical Uses of Synchronized Clocks in Distributed Systems by Barbara Liskov
- - Stanford Paper for reference only
- - Spanner Paper for reference only
- - On Time Synchronization Issues in Time-Sensitive Networks with Regulators and Nonideal Clocks for reference only, raised by Elad
- - Accurate Network Clock Synchronization at Scale for reference only, raised by Hesham
- - Exploiting a Natural Network Effect for Scalable, Fine-grained Clock Synchronization for reference only, raised by Hesham
- - SIMON: A Simple and Scalable Method for Sensing, Inference and Measurement in Data Center Networks for reference only, raised by Hesham
- - New Guidelines for Inclusiveness for reference only, raised by Michael