Outline

Welcome … as we collectively explore a new frontier

Overview∙ Scoping / Context,∙ Terminologies / Topologies∙ Considerations / Facets, and∙ Exclusions Edge Device Component Stack Scenario Walk-throughs Prospective Takeaways ( your collective homework )

Overview

Scoping / Context

How you could spend a career on this topic ...Some Applicable Contextual References∙ Industry 4.0 - “smart factory, fourth industrial revolution”∙ Industrial Internet of Things - “evolution of distributed control system with higher levels of automation”∙ Consortiums / Working Groups∙ Linux Foundation Edge OpenFog EdgeX Foundry Everyware Software Framework ∙ OpenStack related : Edge Computing Group Fog Edge Massively Distributed Clouds StarlingX ∙ Kubernetes related : Kubernetes IoT Edge KubeEdge KNative K3s

∙ etc., etc. …

∙ Yet, for this session, we will try to focus on a coherent edge computing strategy implemented to address the large number of devices, workloads and the respective data/management/model flows

Terminologies / Topologies

Trying to be (at least internally consistent herein) for this “connected world view” ...The common layers of compute/networking/storage resources:∙ IoT or IIoT clusters (focusing on the interaction with these devices)∙ countless types of devices that can measure or control environments

∙ Edge / Gateways∙ smaller, form factor systems, that may have distinct I/O capabilities and

are often environmentally hardened to aggregate connectivity with IoTclustered devices

∙ Core∙ either another tier up of more resource-rich collections of systems

or in data-center-based locales∙ Cloud∙ either public, hybrid or on-premise collections of resources

Considerations / Facets

The main aspects this presentation focuses on are ...The triad of∙ Data-Driven Process and Flows∙ sources, landscape, piplelines, governance, consumption

∙ Management of Devices, Operating Systems, Workloads ∙ provision, configure, monitor, update/secure infrastructure and applications

∙ Model / Workload to actually perform the computing/analytic functions∙ utilize, create, propagate

∙ Note : the color coding of the triad will be used throughout

Exclusions

Setting expectations for items beyond the scope of this presentation ...Skipping:∙ The majority of actual IoT devices themselves, even though they∙ provide the Data, and∙ are often the target of Model / Workload generated control outputs

∙ Any edge platform system architectures besides x86_64 or aarch64∙ Coverage of the immense catalog of possible applications / workloads (only a few cited as examples)

Edge Device Component Stack

Edge Computing Platform (1)

Dissecting the layers ...Countless Hardware Choices

● ADLINK Embedded and Industrial IoT Gateway Solutions● Dell Edge Gateway Systems●●●●●●●●

Edge Computing Platform (2)

Dissecting the layers ...Countless Operating System Choices

● BYO/DIY ( Yocto )● Distributions:

● special purpose, real-time versions● community-based Linux● enterprise-based Linux ( SUSE JeOS )

● Hosting workloads via:● baremetal deployments● on top of hypervisors ( KVM )● using container runtime engines w/ or w/o Kubernetes● or combinations of any/all of the above

Edge Computing Platform (3)

Dissecting the layers ...Countless Workload Deployment Options

● Via simple binaries / packages● Bundled in a VM● Bundled in a container image● or combinations of any/all of the above

∙ So, nearly every known data center technology∙ in a tiny, yet powerful and remote environment

Scenario Walk-throughs

Starting Simple

Score for Proof-of-Concept approach ...IoT Devices Connected to Edge Node

● Simple, direct access connectivity model● Data

● polled from IoT device APIs● little retention or propagation

● Management● mostly one-off, ad-hoc, DIY

● Model / Workload● simple DIY workloads● reactive control functions

Starting Simple

Score for Proof-of-Concept approach ...IoT Devices Connected to Edge Node

● Simple, direct access connectivity model● Data

● polled from IoT device APIs● little retention or propagation

● Management● mostly one-off, ad-hoc, DIY

● Model / Workload● simple DIY workloads● reactive control functions

Concerns:Difficult to Scale / ReplicateData RetentionRecoverySecurity

Going Beyond Simple

Suggestions to address phase concernsIoT Devices Connected to Edge Node

● Simple, direct access connectivity model● Data investigate / consider

● data pipeline / flow disciplines● data governance principles / retention guidelines

● Management standardize on● edge node physical requirements● operating system footprints● workload hosting options

● Model assemble Workloads to● implement data pipeline actions● analyze, react and implement control functions

DataDiscovery

DataRefinement

DataEnrichment

DataGovernance

Begin Scaling

Building on Proof-of-Concept learnings ...Augmenting IoT Analytic / Connectivity / Visibility

● Adding PCIe, SIO connections● Data

● polled and direct from IoT devices● begin retention processes to Core / Cloud

● Management● remotely access● consider lifecycle / configuration management

● Model / Workload● develop analytic models on central data stores● create standard control workloads

Begin Scaling

Building on Proof-of-Concept learnings ...Augmenting IoT Analytic / Connectivity / Visibility

● Adding PCIe, SIO connections● Data

● polled and direct from IoT devices● begin retention processes to Core / Cloud

● Management● remotely access● consider lifecycle / configuration management

● Model / Workload● develop analytic models on central data stores● create standard control workloads

Concerns:Scale across so many aspectsData Propagation / Visibility / IntegrityReliability / ResiliencyAvailability / PerformanceServiceability / Security

Dealing With Scale

Suggestions to address concernsAugmenting IoT Analytic / Connectivity / Visibility

● Adding PCIe, SIO connections● Data focus on

● coordinating incremental / batch data flows● encryption at-rest and in-flight● consistent storage protocols

● Management implement● remote rollback / recovery methods● access controls (across triad)

● Model / Workload utilize● CI/CD development principles● repository / registry delivery network approach

Strive For A Managed State

Treated like part of a distributed data center ...Applying / extending standard pillars of IT

● Create and orchestrate process flows for● Data (leverage DataOps)

● ingest, transform, cleanse, match, consolidate,profile, metadata cataloging, govern

● consistent storage access protocols● Management (span *Ops)

● central admin, tiered, group-based configurations● infrastructure-as-code approach

● Model / Workload (leverage DevOps)● utilize tiers to enhance AI/ML/DL approach● DevOps create standard control workloads

Strive For A Managed State

Treated like part of a distributed data center ...Applying / extending standard pillars of IT

● Create and orchestrate process flows for● Data (leverage DataOps)

● ingest, transform, cleanse, match, consolidate,profile, metadata cataloging, govern

● consistent storage access protocols● Management (span *Ops)

● central admin, tiered, group-based configurations● infrastructure-as-code approach

● Model / Workload (leverage DevOps)● utilize tiers to enhance AI/ML/DL approach● DevOps create standard control workloads

Concerns:SecurityProvisioning / re-imagining / RecoveryConnectivity Issue (Low bandwidth / intermittent)

Summary

Prospective Takeaways

Our collective homework

● What would you like to see from SUSE for IoT?● How can you contribute to SUSE IoT strategy?

● Share your experiences with us● Community or working group involvement

Questions

Resources

Beyond the many embedded links∙ Stop by the SUSE Embedded booth (booth 19)∙ Ask about SUSE Manager hands-on guide (HO1426)

∙ A useful walkthrough that shows registering and updating nodes and creating/cloning/merging channels via GUI and command line

∙ Pertinent SUSECon Sessions● To review (already happened)

∙ BOV1327, DEV1181, HO1426, TUT1197, FUT1437, FUT1439● Upcoming

∙ TUT1314, TUT1364

Unpublished Work of SUSE LLC. All Rights Reserved.This work is an unpublished work and contains confidential, proprietary and trade secret information of SUSE LLC. Access to this work is restricted to SUSE employees who have a need to know to perform tasks within the scope of their assignments. No part of this work may be practiced, performed, copied, distributed, revised, modified, translated, abridged, condensed, expanded, collected, or adapted without the prior written consent of SUSE. Any use or exploitation of this work without authorization could subject the perpetrator to criminal and civil liability.

General DisclaimerThis document is not to be construed as a promise by any participating company to develop, deliver, or market a product. It is not a commitment to deliver any material, code, or functionality, and should not be relied upon in making purchasing decisions. SUSE makes no representations or warranties with respect to the contents of this document, and specifically disclaims any express or implied warranties of merchantability or fitness for any particular purpose. The development, release, and timing of features or functionality described for SUSE products remains at the sole discretion of SUSE. Further, SUSE reserves the right to revise this document and to make changes to its content, at any time, without obligation to notify any person or entity of such revisions or changes. All SUSE marks referenced in this presentation are trademarks or registered trademarks of Novell, Inc. in the United States and other countries. All third-party trademarks are the property of their respective owners.