AWS Lambda and #serverless. What’s all the fuzz about?

Slide 1

Slide 1 text

Joint work with Erik Wittern and Waldemar Hummer Josef Spillner AWS Lambda and #serverless. What’s all the fuzz about? Philipp Leitner [email protected] @xLeitix

Slide 2

Slide 2 text

Chalmers !2 https://icet-lab.eu @IcetLab We are hiring a postdoc!!

Slide 3

Slide 3 text

Chalmers !3 Function-as-a-Service?

Slide 4

Slide 4 text

Chalmers !4 Function-as-a-Service? Serverless?

Slide 5

Slide 5 text

Chalmers !5 Source: https://aws.amazon.com/lambda/

Slide 6

Slide 6 text

Chalmers !6 Source: https://aws.amazon.com/lambda/

Slide 7

Slide 7 text

Chalmers !7 Source: https://aws.amazon.com/lambda/

Slide 8

Slide 8 text

Chalmers !8 Source: https://aws.amazon.com/lambda/

Slide 9

Slide 9 text

Chalmers !9 Source: https://github.com/aws-samples/lambda-refarch-streamprocessing An Example: Real-Time Tweet Processing

Slide 10

Slide 10 text

Chalmers !10 Internally Container Runtime Events (e.g., client requests, system events, triggers…) User/tenant datastore Execution datastore e e e … Execution Queue Result data Function Controller Function datastore Function source code Function container Event Controller … Load balancer Authentication Rule engine Authorization Function container Function container Source: loosely based on IBM’s OpenWhisk architecture

Slide 11

Slide 11 text

Chalmers !11 No state

Slide 12

Slide 12 text

Chalmers !12 No state Hard cap on max. execution time

Slide 13

Slide 13 text

Chalmers !13 No state Hard cap on max. execution time Extreme tail latency

Slide 14

Slide 14 text

Chalmers !14

Slide 15

Slide 15 text

Chalmers !15 Methodology Interviews (n = 12) Analysis of Grey Literature (n = 50) Web Survey (n = 182)

Slide 16

Slide 16 text

Chalmers !16 Main Findings Compositional Application Model “Microservices on steroids”

Slide 17

Slide 17 text

Chalmers !17 Source: https://github.com/aws-samples/lambda-refarch-streamprocessing Recall: Real-Time Tweet Processing

Slide 18

Slide 18 text

Chalmers !18 Compositional Application Model

Slide 19

Slide 19 text

Chalmers !19 Main Findings Compositional Application Model Use Cases

Slide 20

Slide 20 text

Chalmers !20 Main Findings User-Facing 27 (29%) Backend 29 (31%) Both 38 (40%) But: many challenges in user-facing apps

Slide 21

Slide 21 text

Chalmers !21 Use Cases - Backend

Slide 22

Slide 22 text

Chalmers !22 Main Findings Compositional Application Model Use Cases Challenges

Slide 23

Slide 23 text

Chalmers !23 Challenges

Slide 24

Slide 24 text

Chalmers !24 The Future (as we see it) “Faster horses” Different Services vs (end user view) (provider view)

Slide 25

Slide 25 text

Chalmers !25 The Future (as we see it) Different Services “Function-as-a-Service as the assembly language of the cloud”

Slide 26

Slide 26 text

Chalmers !26 The Future (as we see it) Different Services “Function-as-a-Service as the assembly language of the cloud”

Slide 27

Slide 27 text

Chalmers !27 Learn More https://peerj.com/preprints/27005/