cloud data sharing platform project proposal · pserc summer workshop, july 2015 5 dave anderson,...

13
Carl Hauser & Anjan Bose (Washington State University) Ken Birman (Cornell University) Eugene Litvinov and Xiaochuan Luo (ISO New England) PSERC Workshop July 2015 1 Cloud Data Sharing Platform Project Proposal

Upload: others

Post on 10-Jul-2020

6 views

Category:

Documents


0 download

TRANSCRIPT

Carl Hauser & Anjan Bose (Washington State University)

Ken Birman (Cornell University)

Eugene Litvinov and Xiaochuan Luo (ISO New England)

PSERC Workshop July 2015 1

Cloud Data Sharing Platform Project Proposal

Cloud Data Sharing Platform Concept

PSERC Summer Workshop, July 2015

2

Benefits

ARPAe Project Review: Dec 18. 2013

3

A flexible and extensible platform for deploying new kinds of computational, data-intensive, historical and real-time analytics Automating management tasks associated with initiating

analysis tasks within the cloud

Strong consistency and security guarantees for shared data

Inexpensive continuous availability using remote backup locations

Working toward a demonstration that cloud technology can safely and securely host a critical infrastructure task

Background

ARPAe Project Review: Dec 18. 2013

4

Built on existing ARPA-E GridCloud Project Cloud management

Thousands of simulated (real-time) PMUs on the WECC system

Linear State Estimator (aka Phasor State Estimator)

and ISO-NE Phase-1 Demonstration 73 real (but recorded) PMU data streams

Network planning model

PSE

All cloud resources in a single AWS data center

Application

Tim

e A

lign

men

t

Co

mp

uta

tio

n

Res

ult

s

ARPA-E GridCloud Project

PSERC Summer Workshop, July 2015

5

Dave Anderson, Theodoros Gkountouvas, Carl Hauser

Data Collectors Replica 1

FWDN

FWDN-1

FWD1

…FWD2

Application

Tim

e A

lign

men

t

Co

mp

uta

tio

n

Res

ult

s

Time-Synchronized Data Sources Internet / VPN GridCloud running on Amazon Elastic Compute Cloud

Data Collectors Replica N

FWDN

FWDN-1

FWD1

…FWD2

Application

Tim

e A

lign

men

t

Co

mp

uta

tio

n

Res

ult

s

Results

GridCloud Demonstration (6K bus, 3 Replicas)

6

Managed by DMake and ISIS2

GridStat

Simulated Power System Replica 1

Re-

pla

yed

Dat

a Fi

les

PMUN P

PMUN-1 P

PMU1 P

P

…PMU2

S

FE

FE

FE

FE

GridStat (UDP)

GridCloud Replica 1

FWDN

FWDN-1

FWD1

…FWD2

SubstationSE1 P

SubstationSEM P

Control Center SERunning at 5Hz

Tim

e A

lign

men

t

S

Co

mp

uta

tio

n

Res

ult

s

Vis

ual

izat

ion

Simulated Power System Replica 2

Re-

pla

yed

Dat

a Fi

les

PMUN P

PMUN-1 P

PMU1 P

P

…PMU2

S

FE

FE

FE

FE

GridStat (UDP)

GridCloud Replica 2

FWDN

FWDN-1

FWD1

…FWD2

SubstationSE1 P

SubstationSEM P

Simulated Power System Replica 3

Re-

pla

yed

Dat

a Fi

les

PMUN P

PMUN-1 P

PMU1 P

P

…PMU2

S

FE

FE

FE

FE

GridStat (UDP)

GridCloud Replica 3

FWDN

FWDN-1

FWD1

…FWD2

SubstationSE1 P

SubstationSEM P

Time-Synchronized Data Sources Internet GridCloud running on Amazon Elastic Compute Cloud

Visualization Viewer (RDP)

Visualization

Local Machine

4632 Streams

4632 Streams

4632 Streams

291 Substations

291 Substations

291 Substations

175

275

375

475

575

0 100 200 300La

ten

cy (

ms)

Time (Seconds)

175

275

375

475

575

0 100 200 300

Late

ncy

(m

s)

Time (Seconds)

0

2000

4000

6000

8000

10000

0 100 200 300

Late

ncy

(m

s)

Time (Seconds)

175

275

375

475

575

0 100 200 300

Late

ncy

(m

s)

Time (Seconds)

6k Bus System, 144 instance:

Replica Latency Reduction8

Graphs: Latency – Length of time until all necessary data is available

R2 has a likely bad Network

Interface Card that we were

unable to replace prior to

preparation of this graph

0

50

100

150

200

250

175 195 215 235 255 275 295 315 335 355 375 395 415 435 455 475 495 515 535 555 575

Best Replica

0

50

100

150

200

250

175 195 215 235 255 275 295 315 335 355 375 395 415 435 455 475 495 515 535 555 575

Best Replica

Replica 3

Replica 2

Replica 1

6K Bus System, 144 instance:

Replica Latency Reduction9

Graphs: Number of times a particular latency occurs

ISO-NE Phase-1: Heterogeneous Data

10

In Phase 1 we added support any configuration of C37.118

(~24B-65,500B frames), with new challenges:

Framing of data within TCP streams using C37.118 header info

Application (PSE) configuration changes: need to know how each

stream is configured

Also support C37.118 PDC frames via the CloudRelay; PDC

frames can contain many PMUs worth of measurements

which need to be demuxed for use in GridCloud

Component Model of ISO-NE Phase 1

GridStat

Cloud Relay

GridCloud Replica 1

FWDN

FWDN-1

FWD1

…FWD2

SubstationSE1 P

SubstationSEM P

Control Center SERunning at 5Hz

Tim

e A

lign

men

t

S

Co

mp

uta

tio

n

Res

ult

s

Vis

ual

izat

ion

GridCloud Replica 2

FWDN

FWDN-1

FWD1

…FWD2

SubstationSE1 P

SubstationSEM P

GridCloud Replica 3

FWDN

FWDN-1

FWD1

…FWD2

SubstationSE1 P

SubstationSEM P

GridCloud running on Amazon EC2 or VPC

73 Streams

Managed by DMake and ISIS2

73 Streams

73 Streams

TCP or TLS Client

Handler

C37 PDC Stream

Demuxer

Replayed Data

Serving data from 73 PMUs

ISO-NE Internet/VPN

TCP

DataArchive

DataArchive

DataArchive

Key Elements of Supplemental Proposal w/ISO-NE

PSERC Summer Workshop, July 2015

12

Implement Encryption on various links Remote backup sites in other AWS regions; region->region

failover Make PSE output available as data streams that can be

consumed by other applications, in the cloud, or outside the cloud

Document impact of Encryption Network delays Network outages Cloud outages and node overloads

Where next?

ARPAe Project Review: Dec 18. 2013

13

Applications using the platform Current demonstrations: PSE and data archive

Unexplored strength of the cloud platform: dynamic deployment of computational resources

Expressing and enforcing data sharing policies What is each sharing entity willing to share with which

other participants

Acquiring and using real-time and archival topology information

See Anjan Bose or Carl Hauser to discuss further ([email protected]; [email protected])

Discussion

PSERC Summer Workshop, July 2015

14