Serverless AWS reInvent 2019 recap

Tuesday, Jan 28, 2020
1. Intro & Activity Update
2. Community Open Mic
3. Jonathan Dion: “The state of
Serverless after AWS re:Invent
2019”
4. Networking
1
ServerlessToronto.org Meetup Agenda

Serverless is not just about the Tech:
2
Serverless is New Agile & Mindset
Serverless Dev (gluing
other people’s APIs and
managed services)
We're obsessed by
creating business value
(meaningful MVPs,
products) and helping
Startups
We build bridges
between Serverless
Community (“Dev leg”),
and Front-end & Voice-
First folks (“UX leg”),
and empower UX
developers
Achieve agility NOT by
“sprinting” faster (like in
Scrum), but by working
smarter (by using
bigger building blocks
and less Ops)

We wrapped up last decade nicely…
3
Wardley Maps show how things
evolve (x-axes) and move up the
value chain (y-axes), then evolve
again… moving up.
Innovation happens, then evolves,
goes mainstream, becomes a
commodity. E.g.
• Electricity => 20th century reality
• Computers => e.g. Internet
• Internet => e.g. Cloud
• Cloud => e.g. Machine Learning,
Serverless/Utility Computing
• ML/Serverless => e.g. Voice First
Technologies
=> e.g. Killer Robots ?
Tech predictions for the next one?

Venue Sponsor
5
As Certified B Corporation, Myplanet is purpose-driven and
creates benefit for all stakeholders, not just shareholders!

Knowledge Sponsor
6
Get your raffle tickets… and GOOD LUCK!
1. Go to www.manning.com
2. Select eBook or Video title you like
3. Add it to the shopping cart
4. Raffle winner will send me email address used
5. For Manning staff to move to your dashboard

Upcoming 2020 #ServerlessTO Meetups
8
1. Using Data Science & Serverless Python (Zappa) to
Find an Apartment in Toronto – Ian Whitestone, Data
Scientist at Shopify ** FEB 27 **
2. Intro to PySpark – Python Data Analysis at scale in the
Cloud – Jonathan Rioux, Lead Data Scientist at EPAM
Systems & author of PySpark in Action book
3. Introduction to Google BigQuery – Matt Welke, Software
Developer at GroupBy Inc
4. Serverless with Pivotal Cloud Foundry – Adib Saikali,
Principal Platform Architect at VMware
5. Fivetran – Data Pipelines, Reinvented – Replicate your
data into the Cloud Warehouse of your choice

Community Open Mic
9
Your 10 sec. pitch ☺
- Looking for work?
- Offering work?
About You – because without you, there would be no meetups!

Feature Talk
Jonathan Dion, Senior Technical
Evangelist at AWS
10

Jonathan Dion
Senior Developer Advocate
Amazon Web Services
@jotdion
linkedin.com/in/jotdion
Serverless Toront o

General Availability
LEARN MORE CON-326R - Running Kubernetes Applications on AWS Fargate
Introducing
The only way to run serverless Kubernetes containers securely,
reliably, and at scale
Amazon EKS for AWS Fargate

Some limitations
• Maximum of 4 vCPU and 30 GB memory per pod
• No persistent volumes or file systems
• You cannot run Daemonsets, Privileged pods or pods that use
HostNetwork or HostPort
• Application Load Balancer only
• Pods must match a Fargate profile at the time that they are scheduled
in order to run on Fargate
• No public IP address

Receive lower rates
automatically. Easy to use
with recommendations in
AWS Cost Explorer
Significant
savings of up to 72%
Flexible across instance family,
size, OS, tenancy or AWS
Region; also applies to AWS
Fargate & soon to AWS
Lambda usage
LEARN MORE CMP210: Dive deep on Savings Plans
Announced
Simplify purchasing with a flexible pricing model that offers savings of up
to 72% on Amazon EC2, Amazon ECS, AWS Fargate & AWS Lambda usage
Savings Plans

AWS Serverless Application Repository

Lambda execution model
Synchronous
(push)
Asynchronous
(event)
Stream
(Poll-based)
/order
Amazon API
Gateway
Lambda
function
Amazon
DynamoDB
Amazon
Kinesis
changes
AWS Lambda
service
function
Amazon
SNS
Amazon
S3
reqs
Lambda
function
Amazon
SQS + FIFO
NEW!!!

Async enhanced controls
When you know that retries won’t succeed, or would like to
remove the old requests that haven’t been invoked, use these
new controls:
• Event Age (Default/Maximum: 6 hours, Minimum: 60s)
• Retry Attempts (Default/Maximum: 2, Minimum: 0)
These new controls are not breaking changes.

Introducing Lambda Event Destinations
For asynchronous invocations, capture
success or failure
• Record contains details about the request
and response in JSON format
• Contains more information than data sent
to a DLQ
• Can send both outcomes to same
destination
or
• Can send success to one destination,
failure to another
AWS Lambda
AWS Lambda
Amazon EventBridge
Amazon SNS
Amazon SQS

Use Lambda Event Destinations for:
• Batching events to process with SQS or EventBridge
• Visibility and notification using EventBridge or SNS
• Handling failure with additional context
• Building simple workflow by chaining Lambda functions

Announcements for streamed event sources
• Enhanced controls on stream processing
• How many retries on a record: MaximumRetryAttempts (0 to 10,000)
• Skip processing old records: MaximumRecordAgeInSeconds (60 seconds to 7 days)
• Trace down the bad records: BisectBatchOnFunctionError
• Send batch info and request context of failed records to SNS/SQS: On-failure
destination
• Batch Window: allows to batch up to 300 seconds before
invocation (instead of data every seconds)
• Invocation criteria: 6MB payload, Maximum Batch Window, Maximum Batch Size
• Parallelization Factor: allows to have up to 10 batches on 10
instances per Kinesis shard (instead of 1 instance per shard)

The function lifecycle
Bootstrap
the runtime
Start your
code
Full
cold start
Partial
cold start
Warm
start
Download
your code
Start new
Execution
environment
AWS optimization Your optimization

Lambda + VPC, no longer a cold-start pain point!
 Before: 14.8-sec duration
After: 933 ms duration →
NEW!!!

Lambda + VPC, no longer a cold-start pain point!
 Before: 14.8 sec duration
After: 933ms duration →
NEW!!!

L E A R N M O R E SVS401 - Optimizing your serverless applications
Provisioned Concurrency on AWS Lambda
Introducing
• Keeps functions initialized and hyper-ready, ensuring
start times stay in the milliseconds
• Builders have full control over when provisioned
concurrency is set
• No code changes are required to provision concurrency
on functions in production
• Can be combined with AWS Auto Scaling at launch

What Happens on a Cold Start?
Client
Invoke
AWS Lambda
Service
Create new
Execution
Environment
& Download
Code
Bootstrap
the runtime Invoke the
Handler

Prior to Provisioned Concurrency…
• Pre-warm your function with the concurrent invocations
• Don’t ping more often than every 5 minutes
• Pass in a test payload that can be identified as such
• Create handler logic that replies accordingly without running the whole function
exports.handler = async (event) => {
// if a warming event
if (await warmer(event)) return 'warmed’
// else proceed with handler logic
return 'Hello from Lambda’
}

AWS Lambda Provisioned Concurrency
• Greater control over the start up time
of your Lambda functions.
• Functions stay initialized and hyper-
ready to respond in double-digit
milliseconds.
• No code changes necessary
• Ideal for interactive applications such as
web or mobile backends
• Good for customers that are building:
• Applications that have strict latency SLAs
• Applications that support high-velocity traffic bursts

Use Cases: Strict Latency SLAs
• Applications that:
• Have direct interaction with end-users
• Leverage languages that have a slower cold start time or require large deployment packages
• Lambda will:
• Pre-create and initialize the sandboxes as if it was a normal cold start
• The provisioning speed defaults to standard burst scaling rate (3K at once then +500 per
minute)

Use Cases: High-Velocity Traffic Bursts
• Applications that:
• Serve content such as ads during a live stream
• Mobile applications such as games
• Marketing blitzes or flash sales
• Lambda will:
• Provisioning scheduled in advance
• Schedule in advance to allow for scale-up
• Provision in increments of 5 minutes
• Invocations above the provisioned concurrency are handled by on-demand Lambda

Provisioning Concurrency
• Provisioning scheduled in
advance
• Schedule in advance to
allow for scale-up
• Provision in increments of
5 minutes
• Invocations above the
provisioned concurrency
are handled by on-demand
Lambda
0
500
1000
1500
2000
2500
0
500
1000
1500
2000
2500
Traffic (TPS) Provisioned concurrency

Lambda runtimes
New Runtimes
• Based on the Amazon Linux 2 execution environment
• Node.js 12
• Java 11
• Python 3.8
Change to current Runtimes
• requests library removed from botocore used by boto3 (Python SDK)

Business workflow is rarely sequential start to finish
Multiple decision paths
Need to handle failure
Multiple step processes

AWS
Step Functions

Introduction to Step Functions
All work in your state
machine is done by
tasks
You can directly
call other AWS
services
Use Amazon
CloudWatch to
monitor your
workflows
Retries and fallbacks
are available to you
States are elements
in your state
machine

AWS Step Functions + Lambda
“Serverless” workflow management with
zero administration:
• Makes it easy to coordinate the
components of distributed applications
and microservices using visual workflows
• Automatically triggers and tracks each
step and retries when there are errors, so
your application executes in order and as
expected
• Logs the state of each step, so when
things do go wrong, you can diagnose and
debug problems quickly
Choice
Start
ExtractImageMetadata
CheckJobStatus
Rekognition
ImageTypeCheck
NotSupportedImageType
End
Thumbnail
AddRekognizedTags
Tasks
Failure capture
Parallel tasks

Step Functions: Integrations
Simplify building workloads such as order processing, report
generation, and data analysis
Write and maintain less code; add services in minutes
More service integrations:
AWS Step
Functions
Amazon Simple
Notification
Service
(Amazon SNS)
Amazon Simple
Queue Service
(Amazon SQS)
Amazon
SageMaker
AWS Glue AWS Batch Amazon Elastic
Container Service
AWS Fargate Amazon EMR
NEW!!!

Simpler integration, less code
With serverless polling With direct service integrationStart
End
AWS
Lambda
functions
Start
End
No
Lambda
functions

AWS Step Functions Express Workflows
Introducing
Orchestrate AWS compute, database, and messaging services at rates
greater than 100,000 events/second, suitable for high-volume event
processing workloads such as IoT data ingestion, streaming data
processing and transformation.
L E A R N M O R E API321: Event-Processing Workflows at Scale with AWS Step Functions

Which one is right my use case?
Standard Workflows
• IT automation
• report generation
• order processing
• payment and billing
• machine learning model training
• ELT and big data orchestration
• media processing
Express Workflows
• event driven microservice
orchestration
• high volume data processing
• IoT data ingestion
• order / cart validation
• payment reconciliation

Standard vs. Express Workflows
Standard Express
Console Step Functions Step Functions
State machine
definition
Amazon States Language Amazon States Language
Documentation Step Functions Step Functions
Pricing Priced per state transition
Priced by the number of executions
you run, their duration, and memory
consumption

Standard Express
Maximum
duration
365 days 5 minutes
Start execution
Bucket Size: 1,300
Refill Rate per Second: 300
Bucket Size: 6,000
Refill Rate per Second: 6,000
State transition
Bucket Size: 5,000
Refill Rate per Second: 1,500
Unlimited
Execution
semantics
Exactly-once workflow step
execution
At-least-once workflow step
execution

Standard Express
Executions
Executions are persisted and have
ARNs
Executions are not persisted except
as log data
Execution
history
Stored in Step Functions, with
tooling for visual debugging in the
console
Sent to Amazon CloudWatch Logs
Service
integrations
Supports all service integrations and
activities
Supports all service integrations.
Does not support activities.
Patterns Supports all patterns
Does not support Job-run (.sync) or
Callback (.waitForTaskToken)

Standard + Express Workflows
Express Workflows can be nested within a Standard Workflow, enabling
customers to use Express Workflows for short periods of task execution
and a Standard Workflow for long periods of task execution or waiting.
EndStart
Standard Workflow
Express Workflow
Validate
Image
Approval
Notification
Approval
Received
Store
Metadata
Rekognition Add
Tags
Thumbnail

Customers are ditching monolithic applications
Monolithic application
Does everything
Shared release pipeline
Rigid scaling
High impact of change
Hard to adopt new technologies
Microservices
Does one thing
Independent deployments
Independent scaling
Small impact of change
Choice of technology
When the impact of change is small, release velocity can increase

How microservices communicate

Amazon EventBridge
Native integrations with SaaS providers
17 target services
Easily build event-driven architectures
Fully managed, pay-as-you-go
Amazon
EventBridge

Amazon EventBridge event bus
Rule 1
Rule 2
Rule 3
Event Source
Amazon EventBridge overview
Targets

Event Source TargetsAmazon EventBridge event bus
Rule 1
Rule 2
Rule 3
Amazon EventBridge overview

Anatomy of an event
{
"version": "0",
"id": "adeacade-c34c-ce58-c4a0-74f106398c4e",
"account": ”123456789012",
"region": "us-east-1",
"time": "2019-12-02T21:46:19Z",
"source": "order-service",
"detail-type": "New Order",
"resources": [],
"detail": {
"orderId": "cfb2ae566f9b",
"customerId": "C12345",
...
}
}

How do you manage event types?

Amazon EventBridge Schema Registry
Introducing
Store event structure - or schema - in a shared central location, so it’s
faster and easier to find the events you need. Generate code bindings
right in your IDE to represent an event as an object in code.
Preview
LEARN MORE API320 – Building event-driven architectures with Amazon EventBridge

Schema Registry and Discovery pricing
Schema Registries
Free!
Schema Discovery
Free tier Additional usage

Amazon Relational Database Service (Amazon RDS)
Managed relational database service with a choice of popular database engines
Easy to
administer
Easily deploy and maintain
hardware, OS, and DB
software; built-in
monitoring
Performant
& scalable
Scale compute
and storage with a few
clicks; minimal downtime
for your application
Available
& durable
Automatic Multi-AZ data
replication; automated
backup, snapshots,
and failover
Secure &
compliant
Data encryption at rest
and in transit; industry
compliance and
assurance programs

Amazon Aurora
Enterprise database at open source price
Delivered as a managed service
Amazon Aurora
Speed and availability of high-end commercial databases
Simplicity and cost-effectiveness of open source databases
Drop-in compatibility with MySQL and PostgreSQL
Simple pay as you go pricing

You have some choices to make . . .
Provision for peak
Continuously monitor and
manually scale up or down
-or-
Expensive
Hard
• Requires experts
• Risks outages
Provision less than peak User (business) impact
-or-

Aurora Serverless architecture
• Starts up on demand, shuts
down when not in use
• Scales up/down
automatically
• No application impact when
scaling
• Pay per second, 1 minute
minimum
Warm pool
of instances
Application
Database storage
Scalable DB capacity
Request routers

You asked for a simpler access model for Aurora
Serverless
• Modern applications on AWS Lambda require networking
• Access restricted within Amazon Virtual Private Cloud (Amazon VPC)
• Database connection limits

Amazon RDS Data API solves these challenges
Millions of
IoT/mobile devices RDS Data API
API
endpoint
Aurora
Serverless
Access through simple web interface
• Public HTTPS endpoint for easier access
• No client configuration required
• No persistent connections required
• Java Client library for a developer-friendly
experience (New!)
AWS Command
Line Interface
(AWS CLI)
AWS
Lambda
AWS
Cloud9
AWS
AppSync

Many applications, including serverless apps, have a
large number of open DB connections and high
connection open/close rate, exhausting DB resources
Modern apps can have
1000s of DB connections,
exhausting DB resources
Custom failure handling:
Code can contain security
risks like DB credentials
Self-managed proxy servers
help manage DB load, but
are difficult to deploy

Amazon RDS Proxy
Introducing
Fully managed, highly available database proxy feature for Amazon
RDS. Pools and shares connections to make applications more
scalable, more resilient to database failures, and more secure.
Public Beta

Amazon RDS Proxy – How it works

Connection pooling
• Sets of system variables, sets of user-defined variables, calls of locking
functions, tables locks, creates of temporary tables, prepares statement
prepare call
RDSProxy
Connection pooling

Seamless and faster failovers
Fast, seamless failover
RDSProxy
Application connections preserved during failovers
Detects failovers and connects to standby quicker, bypassing DNS caches
Up to 66% faster failover times

Improved application security
• Enforce IAM authentication with your relational databases
• Centrally manage database credentials using AWS Secrets Manager

No passwords embedded in code
const signer = new AWS.RDS.Signer({ username: process.env.username,
hostname: process.env.endpoint, region: 'us-east-2', port: 3306, })
const token = await signer.getAuthToken()
const connection = mysql.createConnection({ host: process.env.endpoint,
user: process.env.username, password: token, database:
process.env.database, }) connection.connect()

RDS Proxy regions and engines
• RDS Proxy is available in:
• Asia Pacific (Tokyo)
• EU West (Ireland)
• US East (N. Virginia)
• US East (Ohio)
• US West (Oregon)
• RDS Proxy supports the following
engines:
• RDS MySQL 5.6 and 5.7
• Aurora MySQL 5.6 and 5.7
• RDS and Aurora PostgreSQL coming soon!

Achieve up to 67% cost reduction and 50% latency reduction compared
to REST APIs. HTTP APIs are also easier to configure than REST APIs,
allowing customers to focus more time on building applications.
Reduce application costs by
up to 67%
Reduce application latency by
up to 50%
Configure HTTP APIs easier
and faster than before
HTTP APIs for Amazon API Gateway
Introducing
Preview

REST API vs HTTP API
REST API HTTP API
Authorizers AWS Lambda, IAM, Cognito
Native OpenID Connect/Oauth
2.0, Cognito (JWT issuer)
Integration
HTTP proxy, Lambda proxy,
HTTP, AWS services, Private
integration, Mock
HTTP proxy, Lambda proxy
API Management
Usage plans, API Keys,
Custom domain names
Custom domain names
Extra security
features
Client certificates, AWS WAF,
Resource policies
None

REST API vs HTTP API
REST API HTTP API
Development
Cache, Transformation,
Validation, Test invocation
CORS, Automatic deployments,
Default stage and route
API Type
Regional, Edge-optimized,
Private
Regional
Monitoring
Access logs to CloudWatch logs
and Kinesis Data Firehose,
Execution logs, CloudWatch
metrics, AWS X-Ray
Access logs to CloudWatch
logs, CloudWatch metrics

Create HTTP APIs
OpenAPI 3.0 AWS Serverless
Application Model
integration
AWS::ApiGatewayV2::Api
AWS CloudFormation
AWS::ApiGatewayV2::Api

Latency example

Amplify for iOS & Android
Introducing
Open source libraries and toolchain that enable mobile developers to
build scalable and secure cloud powered serverless applications.
L E A R N M O R E MOB317 - Speed up native mobile development with AWS Amplify

Amplify DataStore
New Feature
Multi-platform (iOS/Android/React Native/Web) on-device persistent
storage engine that automatically synchronizes data between
mobile/web apps and the cloud using GraphQL.
L E A R N M O R E MOB402: Build data-driven mobile and web apps with AWS AppSync

Server-side Encryption using CMK
• Key Management Service integration of Customer managed key

Hot keys identification
Identify frequently access keys and database traffic trends
Preview

Adaptive capacity
• Handles imbalanced workloads by isolating frequently accessed items
automatically
• Shifts data across partitions to rebalance
• Main table and Global Secondary Indexes
• On by default
• No additional cost

2020 and Beyond
Excelling in Service Fundamentals
Availability, latency, security, scale
Equipping Developers
Improving tools for monitoring, troubleshooting, auditing
Adding Efficiency to Dev Patterns
New patterns through events, workflows, and APIs

Jonathan Dion
Senior Developer Advocate
Amazon Web Services
@jotdion
linkedin.com/in/jotdion
Thank you!
All re:Invent 2019 sessions: bit.ly/36HeYl6

Serverless AWS reInvent 2019 recap

Recommandé

Recommandé

Contenu connexe

Tendances

Tendances (20)

Similaire à Serverless AWS reInvent 2019 recap

Similaire à Serverless AWS reInvent 2019 recap (20)

Plus de Daniel Zivkovic

Plus de Daniel Zivkovic (20)

Dernier

Dernier (20)

Serverless AWS reInvent 2019 recap