Scribd
Upload
43 views23 pages

ECS781P 12 Serverless

This document discusses serverless computing as the future of cloud computing. It begins by providing background on the current state of cloud computing, including cloud deployment models, service models, and essential cloud characteristics. It then discusses developments like big data processing and database as a service. Finally, it introduces serverless computing, where users submit functions that are executed without managing servers. Serverless computing offers autoscaling, pay-per-use billing, and is currently being used for web/API services, data processing, and more. However, challenges remain around state management, data access latency, and isolation.

Uploaded by

Yen-Kai Cheng
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
43 views23 pages

ECS781P 12 Serverless

This document discusses serverless computing as the future of cloud computing. It begins by providing background on the current state of cloud computing, including cloud deployment models, service models, and essential cloud characteristics. It then discusses developments like big data processing and database as a service. Finally, it introduces serverless computing, where users submit functions that are executed without managing servers. Serverless computing offers autoscaling, pay-per-use billing, and is currently being used for web/API services, data processing, and more. However, challenges remain around state management, data access latency, and isolation.

Uploaded by

Yen-Kai Cheng
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 23

ECS781P

CLOUD COMPUTING

THE FUTURE OF THE CLOUD?


SERVERLESS COMPUTING
Lecturer: Dr. Sukhpal Singh Gill
School of Electronic Engineering and Computer Science
Contents
• The current state of cloud
• Cloud developments
• Serverless computing

2
Cloud Deployment Models (NIST)
• Public cloud • Private cloud
• Sold to the public, mega- • Enterprise owned or leased
scale infrastructure • Tight security control
• ‘Infinite’ capability
• Metered usage

• Hybrid cloud
• Composition of public and
private clouds

3
Cloud Service Models (NIST)
• Depending on the types of resources offered by the cloud:
• Software as a Service (SaaS) - Gmail
• Use provider’s applications over a network
• Platform as a Service (PaaS) – Google App Engine
• Deploy customer-created applications to a cloud
• Infrastructure as a Service (IaaS) – AWS, GCP
• Rent processing, storage, network capacity, and other
fundamental computing resources

4
Cloud applications

Scalability Consistency

Communications
Application

Data

Maintainability Reliability

5
Essential Cloud Characteristics (from NIST)
• Pervasive network access
• Location independence
• High availability
• Resource pooling and partitioning
• Extensive use of virtualization
• Automated management for cloud clients
• Rapid elasticity

6
Managing current cloud applications
In current cloud infrastructure we have access to
resources on demand, but we still need to manage
them:
•How many resources do we need?
•What replication factor is right?
•Will I get better performance by load balancing?

7
Issues to address setting up a cloud env
1. Redundancy for availability, so that a single machine failure doesn’t
take down the service.
2. Geographic distribution of redundant copies to preserve the service
in case of disaster.
3. Load balancing and request routing to efficiently utilize resources.
4. Autoscaling in response to changes in load to scale up or down the
system.
5. Monitoring to make sure the service is still running well.
6. Logging to record messages needed for debugging or performance
tuning.
7. System upgrades, including security patching.
8. Migration to new instances as they become available.
https://arxiv.org/abs/1902.03383 8
Contents
• The current state of cloud
• Cloud developments
• Serverless computing

9
The Big Data way
Job
M R

Map
Reduce

Input Output
Map
Data Data
Reduce

Map

10
Resource management in Big Data
• Developers purely focus on writing the processing
functions

• The resource management elements (e.g. YARN)


automatically estimate how many resources are
needed to complete the job, and allocate free
resources to execute each task

11
DBaaS
• Database systems maintained by public cloud providers,
and offered as a service
• Excellent integration with cloud ecosystem, but vendor
lock-in

https://cloud.google.com/datastore/pricing 12
Contents
• The current state of cloud
• Cloud developments
• Serverless computing

13
What serverless proposes instead
• The cloud platform automatically takes care of
starting and stopping containers/applications
• Potentially scaling down to zero
• New instances must be up in milliseconds
• Change in billing model: pay per workload, rather
than pay per time
• FaaS: Serverless

14
FaaS (Function as a Service)
• FaaS: Cloud model in which users submit functions
that are executed in a serverless way by the cloud
• Programming model similar to reactive
• Pay per function execution, not per time
• Started by Amazon Lambda functions
• Currently in development for all major platforms
• Google Cloud functions
• Azure Functions
• Oracle FN // Apache OpenWhisk OSS

15
Architecture of a serverless cloud

https://arxiv.org/abs/1902.03383 16
Serverless and cloud computing

17
What is serverless being used for now

Percent Use Case Use Case


32% Web and API serving
21% Data Processing, e.g., batch ETL (database
Extract, Transform, and Load)
17% Integrating 3rd Party Services
16% Internal tooling
8% Chat bots e.g., Alexa Skills (SDK for Alexa AI
Assistant)
6% Internet of Things

https://serverless.com/blog/2018-serverless-community-survey-huge-growth-usage/ 18
Serverless challenges
• Stateless functions: provide storage solutions for
helping with function state
• Serverless-aware data access: low latency
read/write access to data
• Networking should be able to implement some
communication patterns in a much more efficient
manner
• Functions must be isolated to guarantee security

19
Recommended Reading
Cloud Programming Simplified: A Berkeley View on
Serverless Computing
•Eric Jonas, Johann Schleier-Smith, Vikram Sreekanti, Chia-
Che Tsai, Anurag Khandelwal, Qifan Pu, Vaishaal
Shankar, Joao Carreira, Karl Krauth, Neeraja
Yadwadkar, Joseph E. Gonzalez, Raluca Ada Popa, Ion
Stoica, David A. Patterson
•https://arxiv.org/abs/1902.03383

20
Further Reading
Mohammad Sadegh Aslanpour, A.N. Toosi, Claudio Cicconetti, Bahman Javadi, Peter
Sbarski, Davide Taibi, Marcos Assuncao, Sukhpal Singh Gill, Raj Gaire, Schahram Dustdar,
Serverless Edge Computing: Vision and Challenges, Proceedings of the 19th Australasian
Symposium on Parallel and Distributed Computing (AusPDC 2021), February 1–5, 2021,
Dunedin, New Zealand.
Link: https://dl.acm.org/doi/10.1145/3437378.3444367b

21
22

You might also like