Message-Oriented Middleware for Edge Computing Applications

Post on 22-Jan-2018

50 views 1 download

Transcript of Message-Oriented Middleware for Edge Computing Applications

Message-Oriented Middleware

for Edge Computing Applications

Thomas Rausch

Distributed Systems GroupTU Wien

http://dsg.tuwien.ac.at

2

3

Amazon AWS IoT

4

The Internet

Decoupled messaging

5

Subscribe to topic:/robot/arm

6

Publish “move x,y,z” to/robot/arm

7

Topic Subscriber

/robot/arm robot-client

/robot/arm ...Subscriber:robot-client

Publisher

Subscriber:...

Topic/robot/arm

8

Trust?!

Availability?!

End-to-end latency: ??ms!

9

IoT Edge Cloud

Edge Computing

S. Nastic, T. Rausch, et al. 2017. A Serverless Real-Time Data Analytics Platform for Edge Computing. IEEE Internet Computing 21, 4 (2017), 64–71.

G. Lewis, et al. 2014. Tactical Cloudlets: Moving Cloud Computing to the Edge. In 2014 IEEE Military Communications Conference. IEEE, 1440–1446.

Proximity Scale

10

IoT Edge Cloud

Edge Computing

S. Nastic, T. Rausch, et al. 2017. A Serverless Real-Time Data Analytics Platform for Edge Computing. IEEE Internet Computing 21, 4 (2017), 64–71.

G. Lewis, et al. 2014. Tactical Cloudlets: Moving Cloud Computing to the Edge. In 2014 IEEE Military Communications Conference. IEEE, 1440–1446.

Proximity Scale

11

Data centers

Internet

Mobileedge

Extremeedge

IoT

Cloudlets

Message broker

Cloud-based MOM

Edge-enabled MOM

Edge-Enabled MOM

12

Sound familiar?NCA’05

DEBS’05

MW’05

PODC’07

13

What’s new?

● The Cloud– Centralized on-demand resources

● Edge resources– Decentralized, heterogeneous, and volatile resources

(Mobile devices, IoT gateways, Cloudlets)

● Internet of Things– Plethora of devices and protocol standards

● A new form of elasticity: Cloud to Edge– Osmotic Computing

M. Villari, et al. 2016. Osmotic Computing: A new Paradigm for Edge/Cloud Integration. IEEE Cloud Computing

14

Data centers

Internet

Mobileedge

Extremeedge

IoT

Cloudlets

Challenges

15

Data centers

Mobileedge

Extremeedge

IoT

Proximity awareness

16

Data centers

Mobileedge

Extremeedge

IoT

Elasticity from the Cloud to the Edge

Runtimeprovisioningof brokers

Networkreconfiguration

17

Data centers

Mobileedge

Extremeedge

IoT

Privacy& Trust

18

State of the Art – Research & Prototypes

19

Messaging Model

● MQTT: ISO standardized pub/sub protocol– TCP/IP– Small code footprint, low-bandwidth design– Has become a common platform for IoT [1]

● Simple protocol– CONNACK– SUBSCRIBE(t), SUBACK(t)– PUBLISH(t,msg)– ...

[1] A. Al-Fuqaha, et al. 2015. Internet of Things: A Survey on Enabling Technologies, Protocols, and ApplicationsIEEE Commun. Surv. Tutorials, vol. 17, no. 4, pp. 2347–2376, 2015.

20

Messaging Model

● Message delivery guarantees (MQTT “QoS”)– (0) At most once– (1) At least once– (2) Exactly once

● Maintaining guarantees despite distribution

21

State of the Art – The Real World

● Static topic bridging for proximity-based messaging(Mosquitto & HiveMQ)

Local Broker Central BrokerShared Topic

Internal Topics

Staticconfig

Staticconfig

Edge Cloud

Steve. 2017. Mosquitto MQTT Bridge-Usage and Configuration http://www.steves-internet-guide.com/mosquitto-bridge-configuration/

22

State of the Art – The Real World

ScalAgent. JoramMQ, a distributed MQTT broker for the Internet of Things . White paper. 2014.

23

The EMMA Project

ClientGW

c

GW

c

Broker

Broker

Device

Mon

itorin

g pr

otoc

ol

Controller

Bridging Tables

Gateway

Bridging Table

Reconfigura- tion Engine

REST Endpoints

Network monitor

Buffer out

Buffer in

MQTT Server

https://git.dsg.tuwien.ac.at/emma

Architecture Bridging & Reconfiguration

T. Rausch, S. Nastic, S. Dustdar. 2018. EMMA: Distributed QoS-Aware MQTT Middleware for Edge Computing Applications . IC2E’18.

24

Prototype Evaluation

us-east eu-west

eu-central

BrokerGateway BrokerGateway

Controller

... ...

Gateway

Broker

98 ms 27 ms

85 ms

25

Goals for short/mid-term future

● Elicit algorithmic challenges and novelties● Message delivery guarantees (MQTT QoS)● Elastic diffusion of brokers to the Edge

– Control heterogeneity of Edge resources– Efficient monitoring and network reconfiguration

Cloud

Dipl.-Ing. (MSc)

Thomas RauschResearch Assistant

TU WienDistributed Systems GroupArgentinierstraße 8/184-1, 1040, Vienna, AustriaT: +43 1 58801 184 838E: rausch@dsg.tuwien.ac.atdsg.tuwien.ac.at/staff/trausch