Lee Calcote

November 2018

calcotestudios.com/talks

from containers to functions

Electrical and Computer Engineering

The University of Texas at Austin

http://calcotestudios.com/talks

Lee Calcote
October 2017

Electrical and Computer Engineering

The University of Texas at Austin

eth0

eth1

container

network namespace

l0

loopback 0

host network namespace

Lee Calcote

linkedin.com/in/leecalcote

@lcalcote

gingergeek.com

lee@calcotestudios.com

clouds, containers, functions, applications, and their management

calcotestudios.com/talks

github.com/leecalcote

visit layer5.io for more

compliments of NGINX

gingergeek.com

What is Cloud Native?

• Containerized. Each part (applications, processes, etc) is packaged in its own container. This facilitates reproducibility, transparency, and resource isolation.
   
• Dynamically orchestrated. Containers are actively scheduled and managed to optimize resource utilization.
   
• Microservices oriented. Applications are segmented into microservices. This significantly increases the overall agility and maintainability of applications.

@lcalcote

@lcalcote

Development Process

Application Architecture

Deployment and Packaging

Application Infrastructure

Agile

Waterfall

DevOps

N-Tier

Monolithic

Microservices

Cloud

Containers

Physical Servers

Virtual Servers

Data Center

Hosted

Evolution to Cloud Native

λ

Functions

Serverless

Events

SRE

bare metal

     AND

          virtual machines

               AND

                    containers

                         AND

                              unikernels

                                   AND

                                        functions

@lcalcote

We hold these truths to be self-evident...

From Engines to Orchestrators

@lcalcote

Disclaimer: I'm a Docker Captain and organize Docker Austin.

Containers are the “Fastest Growing Cloud Enabling Technology”

- 451 Research

"By 2020, more than 50% of global organizations will be running containers in production."

- Gartner

System Containers

• Like a VM
• Full OS image
• Multiple processes

Application Containers

• Single process

• Use namespaces to deal with resource isolation for a single process.

• Use cgroups to manage resources for a group of processes.

Similarities:

@lcalcote

Types of Containers - Cloud Native way

[k uh n- tey -ner]

[ awr -k uh -streyt-or]

Definition:

@lcalcote

Core

Capabilities

• Cluster Management

  • Host Discovery

  • Host Health Monitoring

• Scheduling

• Orchestrator Updates and Host Maintenance

• Service Discovery

• Networking and Load-Balancing

• Stateful services

• Multi-tenant, multi-region

Additional

Key Capabilities

• Application Health & Performance Monitoring

• Application Deployments

• Application Secrets

@lcalcote

Nomad Architecture

@lcalcote

Docker Swarm 1.11 (Standalone)

@lcalcote

Kubernetes Architecture

@lcalcote

Mesos Architecture

@lcalcote

A high-level perspective of the container orchestrator spectrum.

@lcalcote

Microservices

The more, the more merrier?

Benefits

The first few services are relatively easy

Democratization of language and technology choice

Faster delivery, service teams running independently, rolling updates

@lcalcote

Challenges

The next 10 or so may introduce pain

Language and framework specific libraries

Distributed environments, ephemeral infrastructure, out-moded tooling

Which is why...

 I have a container orchestrator, right?

@lcalcote

The "layer 5" challenge

What do we need?

• Observability

• Logging
• Metrics
• Tracing

• Traffic Control

• Resiliency

• Efficiency
• Security

• Policy

@lcalcote

What is a Service Mesh?

a dedicated layer for managing service-to-service communication

@lcalcote

Missing: application lifecycle management, but not by much

Missing: distributed debugging; provide nascent visibility (topology)

Istio

@lcalcote

• Observability

• Resiliency

• Traffic Control

• Security

• Policy Enforcement

istio.io

github.com/istio

@IstioMesh

Observability

is what gets people hooked on service metrics

@lcalcote

Goals

• Metrics without instrumenting apps

• Consistent metrics across fleet

• Trace flow of requests across services

• Portable across metric backend providers

You get a metric!  You get a metric!  Everyone gets a metric!

Traffic Control

control over chaos

@lcalcote

• Traffic splitting
  • L7 tag based routing?
• Traffic steering
  • Look at the contents of a request and route it to a specific set of instances.
• Ingress and egress routing

Resilency

• Systematic fault injection

• Timeouts and Retries with timeout budget

• Circuit breakers and Health checks

• Control connection pool size and request load

content-based traffic steering

Istio Architecture

@lcalcote

Touches every packet/request in the system. Responsible for service discovery, health checking, routing, load balancing, authentication, authorization and observability.

Provides policy and configuration for services in the mesh.

Takes a set of isolated stateless sidecar proxies and turns them into a service mesh.

Does not touch any packets/requests in the system.

Istio Architecture

@lcalcote

Pilot

Auth

Mixer

Functions

λ

But why?

VM

VM

VM

VM

λ

The Promised Land

No compute cost when idle.

Flexible and precise scaling.

No provisioning, updating, and managing server infrastructure.

@lcalcote

with a few caveats

When to Use

• Asynchronous, concurrent, easy to parallelize into independent units of work
• Infrequent or has sporadic demand, with large, unpredictable variance in scaling requirements
• Stateless, ephemeral, without a major need for instantaneous cold start time
• Highly dynamic in terms of changing business requirements that drive a need for accelerated developer velocity

consider serverless when your workload is...

Serverless

Pain

Points

What is a Unikernel?

A library operating system

application

openGL

gtk

iconv

libgmp

libz

libstd++

libgcc

libc

kernel

libtls

a way of cross-compiling (existing) applications down to very small, lightweight, secure virtual machine

@lcalcote

application

Microservices are (intended to be) small, self-contained, single-purpose applications.

Unikernels cannot handle multiple processes, so forking is not allowed.

Unikernels can handle threads.

Are single user, but who needs multiple users?

Can statically link data into application.

Immutable Infrastructure

enforced

@lcalcote

Unik - Unikernel Compiler and Deployment

What is the CNCF?

• Foster growth and evolution of ecosystem
• Promote underlying technologies
• Provide stewardship for projects
• Make technologies accessible and reliable

@lcalcote

Disclaimer: I'm a Cloud Native Ambassador and TOC Contributor

a vendor-neutral foundation to...

170 Members

@lcalcote

Annual Report 2017

cncf.io

...a community of open source projects, including Kubernetes, Envoy and Prometheus.

Kubernetes and other CNCF projects are some of the highest velocity projects in the history of open source.

@lcalcote

@lcalcote

Cloud Native Interactive Landscape

Cloud Native Landscape

@lcalcote

Disclaimer: I organize the Austin CNCF meetup.

Members of CNCF Meetups

@lcalcote

Cloud Native Trail Map

Working Groups

@lcalcote

• Formed in June 2017 at the request of CNCF Technical Oversight Committee (TOC)
   
• Asked for state of tech/community & recommendations for possible involvement
   
• Most key Serverless players involved
   
• IBM, VMWare, Google, Red Hat, Huawei, Microsoft, SolarWinds, Docker, iguazio, Amazon, MasterCard, Pivotal, Serverless Inc., Clay Labs, The New Stack, A Cloud Guru, Platform9, Bitnami, Auth0, and Hyper 

Serverless WG

@lcalcote

• Describes & defined Serverless
• Highlights promising use cases and areas where functions have already proven value
• Differentiates Serverless from PaaS and Container Orchestration
• Describes the mechanics of a generic Serverless system
• Identifies potential future "harmonization" the WG could look at 

Serverless WG: White Paper

@lcalcote

Serverless White Paper

Creating a common model for event data, similar to CNI and CSI. Coalescing on a single format between a few proposals:

• Cloud-Native Event Mapping (CNEM) – (iguazio)

• CloudEvents – (Serverless, Inc.)

• Cloud Auditing Data Federation – (IBM, DMTF)

Serverless Landscape

@lcalcote

Solarwinds Participation

• Our products included in Cloud Native Landscape.

• Stewarding the Networking working group.

  • Our works highlighted in CNI project adoption.
• We are on the OpenTracing Specification Council.
• SolarWinds working within Serverless WG.
  • Published Serverless white paper. Creating CloudEvents.

• Participating in the OpenMetrics working group.

• Integrated with Prometheus, Istio, OpenTracing and OpenCensus.

• Run Kubernetes, Prometheus, Open Policy Agent.

• Deliver analysis:
  • container networking for ONIF.
  • Kubernetes deployment tooling.

@lcalcote

@lcalcote

from containers to functions

Lee Calcote

Thank you. Questions?

clouds, containers, functions,

applications and their management

linkedin.com/in/leecalcote

@lcalcote

gingergeek.com

calcotestudios.com/talks

github.com/leecalcote

lee@calcotestudios.com