BOSS: Building Operating System Services Stephen Dawson-Haggerty, Andrew Krioukov, Jay Taneja, Sagar Karandikar, Gabe Fierro, Nikita Kitaev, and David.

Slides:



Advertisements
Similar presentations
Distributed Systems Major Design Issues Presented by: Christopher Hector CS8320 – Advanced Operating Systems Spring 2007 – Section 2.6 Presentation Dr.
Advertisements

© 2007 Cisco Systems, Inc. All rights reserved.ICND1 v1.0—1-1 Building a Simple Network Exploring the Functions of Networking.
Total HVAC Control from anywhere Easy control that can save a fortune
Controlling Sutardja Dai Hall Andrew Krioukov Stephen Dawson-Haggerty, Jay Taneja David Culler.
C-Bus Now and in the Future
Gridlogix, Inc. Case Studies XML Symposium Anaheim, CA January 27, 2004 XML Web Services Case Studies in the HVAC Industry Tim Huneycutt,
Infrastructure Services for Physical Data Stephen Dawson-Haggerty.
Building Application Stack Andrew Krioukov Software Defined Buildings.
Sensor Network Platforms and Tools
TAC - Intelligent Infrastructure and Energy Efficiency.
Open Building Information Monitoring
Energy Management with the use of an Intelligent Building Management System.
MELS Project Wrap Up Rich Brown, David Culler, Stephen Dawson-Haggerty, Steven Lanzisera, Jay Taneja Computer Science Division, University of California,
Grids and Grid Technologies for Wide-Area Distributed Computing Mark Baker, Rajkumar Buyya and Domenico Laforenza.
Figure 1.1 Interaction between applications and the operating system.
Green Millennium Personalized Lighting Controls Andrew Krioukov, Arka Bhattacharya, Jason Trager, Domenico Caramagno David Culler.
Systems Analysis and Design in a Changing World, 6th Edition 1 Chapter 6.
Operating Systems: Principles and Practice
CSE 451: Operating Systems Winter 2010 Module 13 Redundant Arrays of Inexpensive Disks (RAID) and OS structure Mark Zbikowski Gary Kimura.
BOSS: Building Operating System Services Stephen Dawson-Haggerty
BOSS: Building Operating System Services Stephen Dawson-Haggerty, Andrew Krioukov, Jay Taneja, Sagar Karandikar, Gabe Fierro, Nikita Kitaev, and David.
Chapter 2 Operating System Overview Patricia Roy Manatee Community College, Venice, FL ©2008, Prentice Hall Operating Systems: Internals and Design Principles,
Intro Building automation describes the advanced functionality provided by the control system of a building. A building automation system (BAS) is an.
APC InfraStruxure TM Central Smart Plug-In for HP Operations Manager Manage Power, Cooling, Security, Environment, Rack Access and Physical Layer Infrastructure.
DISTRIBUTED SYSTEMS Principles and Paradigms Second Edition ANDREW S
Virtualization. Virtualization  In computing, virtualization is a broad term that refers to the abstraction of computer resources  It is "a technique.
B.Ramamurthy9/19/20151 Operating Systems u Bina Ramamurthy CS421.
Mesos A Platform for Fine-Grained Resource Sharing in the Data Center Benjamin Hindman, Andy Konwinski, Matei Zaharia, Ali Ghodsi, Anthony Joseph, Randy.
April 2000Dr Milan Simic1 Network Operating Systems Windows NT.
Advanced Computer Networks Topic 2: Characterization of Distributed Systems.
Distributed Computing Systems CSCI 4780/6780. Distributed System A distributed system is: A collection of independent computers that appears to its users.
Distributed Computing Systems CSCI 4780/6780. Geographical Scalability Challenges Synchronous communication –Waiting for a reply does not scale well!!
Distributed System Concepts and Architectures 2.3 Services Fall 2011 Student: Fan Bai
Embedded Design Using ARM For Strong Room Security System
VMware vSphere Configuration and Management v6
Jini Architecture Introduction System Overview An Example.
Introduction to HVAC Optimization
Operating-System Structures
Tanenbaum & Van Steen, Distributed Systems: Principles and Paradigms, 2e, (c) 2007 Prentice-Hall, Inc. All rights reserved DISTRIBUTED SYSTEMS.
Enabling Deep Demand Response With Supply-Following Loads Jay Taneja (with Prabal Dutta and David Culler) University of California, Berkeley.
Motivating Scenario 1: Networking 1M interacting people for 1,000 hours using at least 1 Gb/s for each user “Tele*, collaborative spaces, and reality TV.
1 CEG 2400 Fall 2012 Network Servers. 2 Network Servers Critical Network servers – Contain redundant components Power supplies Fans Memory CPU Hard Drives.
© 2010 VMware Inc. All rights reserved Why Virtualize? Beng-Hong Lim, VMware, Inc.
Building Automation System
Introduction to Operating Systems
Exploring the Functions of Networking
Distributed System 電機四 陳伯翰 b
Software Defined Networking (SDN)
Operating Systems : Overview
Operating Systems Bina Ramamurthy CSE421 11/27/2018 B.Ramamurthy.
Distribuerte Systemer Viktigere enn vi tror, vanskeligere enn det høres Komponenttorget ‘99 Trondheim Trygve Reenskaug Numerica Taskon Distaribuerte.
CSE 451: Operating Systems Winter 2009 Module 13 Redundant Arrays of Inexpensive Disks (RAID) and OS structure Mark Zbikowski Gary Kimura 1.
Mark Zbikowski and Gary Kimura
Operating Systems : Overview
Substation Automation System
CSE 451: Operating Systems Winter 2012 Redundant Arrays of Inexpensive Disks (RAID) and OS structure Mark Zbikowski Gary Kimura 1.
Operating Systems : Overview
Operating Systems : Overview
Operating Systems : Overview
Operating Systems : Overview
Operating Systems : Overview
Operating Systems : Overview
Operating Systems : Overview
Introduction To Distributed Systems
Personalized HVAC Control System
Chapter 2 Operating System Overview
Distributed Systems and Concurrency: Distributed Systems
‘‘ BUILDING AUTOMATION’’
Introduction and History
Presentation transcript:

BOSS: Building Operating System Services Stephen Dawson-Haggerty, Andrew Krioukov, Jay Taneja, Sagar Karandikar, Gabe Fierro, Nikita Kitaev, and David Culler Computer Science Division University of California, Berkeley © 2013 All rights reserved 4/5/13NSDI 2013: Lombard, IL1

4/5/13 NSDI 2013: Lombard, IL 2 Sutardja-Dai Hall UC Berkeley 93,000 sq. ft. with Digital Controls 73% of US electricity is consumed in buildings U.S. Energy Information Administration, /3 of building occupants are uncomfortable UC Berkeley CBE Study of 30,000 occupants >70% of large buildings have digital controls

4/5/13 NSDI 2013: Lombard, IL Temperature Sensors 312 Light Relays 12 Variable Speed Fans 138 Air Dampers 50 Electrical Sub-meters > 6,000 Sense and Control Points

4/5/13 NSDI 2013: Lombard, IL 4

Applications Ventilation Optimization: 17% energy savings Automated Fault Detection: % energy savings Occupant Lighting Controls 50-60% savings 4/5/13NSDI 2013: Lombard, IL5

Goals and Challenges Portability –Write once, run anywhere for buildings? –Current practice: hand-coded logic Fault tolerance –Partial failures of controllers –Network partitions –Current practice: really tough hardware Multiple processes –Concurrent applications and users –Current practice: none Federation –Multiple heterogeneous systems –Current practice: lots of stovepipes Scale Security & privacy 4/5/13 NSDI 2013: Lombard, IL 6

BOSS: Building Operating System Services HPL Hardware Abstraction Layer Auth. Trans. mgr. Time-series Security Abstraction Isolation + Scheduling HPL Control processes History Fault tolerance “Kernel” interface 7 4/5/13 NSDI 2013: Lombard, IL

Challenge: Portability Buildings are custom designed

4/5/13 NSDI 2013: Lombard, IL 9 Open area 450 Hardware Abstraction Physical view

4/5/13NSDI 2013: Lombard, IL10 VAV S4-21 Hardware Abstraction Systems View

4/5/13NSDI 2013: Lombard, IL11 Hardware Abstraction Controls view SDH.MEC-08.S4-21:DMPR COMD device: instance: 101 SDH.MEC-08.S4-21:VLV COMD device: instance: 102 Controller Air DamperReheat coil BACnet legacy solution: overload point names

Hardware Abstraction Layer 4/5/13NSDI 2013: Lombard, IL12 #VAV > $(120, 20)

Summary: Hardware Abstraction Layer Program applications in terms of relationships between system components –Computer systems tend to hide the physicality memory hierarchies, network topology –Unavoidable in buildings “it gets too hot on the sunny side” Allow for scale by avoiding hard-coding –“Run this in every room, except those on the north side” 4/5/13 NSDI 2013: Lombard, IL 13

BOSS: Building Operating System Services HPL Hardware Abstraction Layer Auth. Trans. mgr. Time-series Security Abstraction Isolation + Scheduling HPL Control processes History Fault tolerance “Kernel” interface 14 4/5/13 NSDI 2013: Lombard, IL

device controller head-end 15 Optimizer “transaction” manager

16 BOSS solution: “transactions”: write access to the building 4/5/13 NSDI 2013: Lombard, IL Writes to distributed resources Which interact in physical space Which are subject to failure Extend writes with – Priorities – Leases – Notifications – Reversion sequences

More BOSS sMAP Hardware Presentation Layer –30 Drivers, 30k data streams Archiver data storage service –500 writes/sec –Stream cleaning and processing Family of apps –Personal ventilation and lighting control –Electric grid-aware consumption 4/5/13NSDI 2013: Lombard, IL17 NameSensor TypeAccess MethodChannels ISO DataCAISO, NYISO, PJM, MISO, ERCOTWeb scrape1211 ACme devicesPlug-load electric meterWireless 6lowpan mesh344 EECS submetering projectDent Instruments PowerScout 18 electric metersModbus4644 EECS steam and condensateCadillac condensate; Central Station steam meterModbus/TCP13 UC Berkeley submetering feeds ION 6200, Obvius Aquisuite; PSL pQube, Veris Industries E30 Mosbus/Ethernet, HTTP4269 Sutardja Dai, Brower Hall BMSSiemens Apogee BMS, Legrand WattStopper, Johnson Control BMS BACnet/IP4064 UC Davis submetering feedsMisc., Schneider Electric IONOPC-DA34 (+) Weather feedsVaisala WXT520 rooftop weather station; Wunderground SDI-12, LabJack/Modbus, web scrape 33 CBE PMP toolkitDust motes; New York Times BMSCSV import; serial874

184/5/13NSDI 2013: Lombard, IL

Takeaways Applying computer systems design to buildings: lots of pieces, potential – Control systems – Mechanical systems – Occupants 30% electricity + steam savings, 60% lighting savings in test apps Many pieces at Control systems + CS future work – Making use of the torrent of data? – Compile/enforce constraints into the network? – How to verify applications are behaving? 4/5/13NSDI 2013: Lombard, IL19

Thank you HPL Hardware Abstraction Layer Auth. Trans. mgr. Time-series Security Abstraction Isolation + Scheduling HPL Control processes History Fault tolerance “Kernel” interface 204/5/13NSDI 2013: Lombard, IL

Write applications in terms of relationship between hardware elements 4/5/13NSDI 2013: Lombard, IL21

4/5/13NSDI 2013: Lombard, IL22

legacy solution: encode everything in point name SDH.MEC-08.S5-01.AIR_VOLUME building name MEC: Modular Equipment Controller #8S: VAV, 5: 5 th floor, 01: the 1 st one quantity being measured 234/5/13NSDI 2013: Lombard, IL

BOSS 4/5/13NSDI 2013: Lombard, IL24 1.Hardware presentation layer: sMAP 2.Hardware abstraction layer: device-specific logic 3.Time-series service: the archiver 4.Reliable control inputs: the transaction manager 5.Security: the authorization service a collection of services enabling portable, robust applications for the physical environment

4/5/13 NSDI 2013: Lombard, IL 25

26 writer 1 value: 69F 69 writer 2 value: 73F 73 4/5/13NSDI 2013: Lombard, IL No arbitration between applications Orphaned writes

Command Sequence 1.Set damper to 100% open 2.Set valve to 0% open 3.… wait 10 minutes 4.Reset to “whatever was happening before” What if… 1.#1 or #2 fail? 2.Client fails/becomes partitioned during #3? 3.Another application tries to do something? 4/5/13 NSDI 2013: Lombard, IL 27

28 writer 1 value: 69F priority: 3 lease: 3600s writer 2 value: 73F priority: 1 lease: 300s present value: 69cfmpresent value: 73cfm writer 2 clear BOSS solution: “transactions” priority array 1 16 writer 1 crashes 4/5/13NSDI 2013: Lombard, IL Extend writes with – Priorities – Leases – Notifications – Reversion sequences overridden! … writer 1 revert sequence runs 71