10 years of Unified Edge Cloud

  • Last Update:2019-12-23
  • Version:001
  • Language:en

Agenda

  • Nexedi
  • Problems
  • Solutions
  • Architecture
 

Nexedi

 

Nexedi - Profile

Nexedi World Map
  • Largest Free Software Publisher in Europe
  • Founded in 2001 in Lille (France) - 35 engineers worldwide
  • Enterprise Software for mission critical applications
  • Build, deploy, train and run services
  • Profitable since day 1, long term organic growth
Nexedi is probably the largest Free Software publisher with more than 10 products and 15 million lines of code. Nexedi does not depend on any investor and is a profitable company since day 1.

Nexedi - Clients

Nexedi References and Map

nexedi.com/success

Nexedi clients are mainly large companies and governments looking for scalable enterprise solutions such as ERP, CRM, DMS, data lake, big data cloud, etc.  

Nexedi - Free Software Stack

Nexedi Software Stack

stack.nexedi.com

Nexedi software is primarily developed in Python, with some parts in Javascript.

SlapOS: (one of) the first edge cloud systems

https://www.cio.com/article/2417512/servers/vifib-wants-you-to-host-cloud-computing-at-home.html

We've been deploying edge computing at Nexedi since around 2008 with SlapOS.

Everyone has a different notion of edge computing. We tried to define them in this article: Five Evolutions of Cloud Computing "https://www.nexedi.com/NXD-Blog.Five.Cloud.Evolution". However, Edge Computing is kind of buzzword that covers many old ideas of distributed computing that recently became more widely accepted.

SlapOS Success Case: Teralab

https://www.nexedi.com/success/slapos-IMT-Documents.Teralab.Success.Case

Problems

 

Problems to Solve

  • IaaS: 13 (OVH, Rapid.Space, AWS, Alicloud, Azure, on-premise, etc.)
  • OS versions: 9 (Debian, CentOS, SuSE, Ubuntu, Android, etc.)
  • Sites: 10 (Lille, Tokyo, Munich, Paris, Shanghai, etc.)
  • Servers et VMs: 267 (incl. 25 HTTP front-ends)
  • Projects: 50
  • Processes per project: 15 (MariaDB, python, Apache, HAProxy, etc.)
  • Domain names: 5000
  • Instances: 1900 (incl. 404 developer instances)
  • Test suites: 112
  • Test results: 293,229
  • Resiliency: 3 copies on 3 sites

Ensure the same result everywhere in a deterministic way → full automation

Cut costs → full portability

 

Solutions

 

Cost: Open Compute Project - OCP

 

Cost: Rapid.Space

https://rapid.space

https://beta.rapid.space/ is a high performance, low cost cloud infrastructure that provides:

  • big servers;
  • CDN;
  • IoT buffering.

It is available in China  in addition to Europe. It is based on SlapOS and Open Compute Project (OCP) hardware, the same as the one used by Facebook.

Resiliency: Home Servers

automated deployment of "smart factory box" for some European automotive company in new factories located in Africa and Asia (all kinds of services can be remotely deployed starting from ERP, CDN and soon MES)

Full Automation: WebRunner IDE

  • Web based IDE (to develop all projects that Nexedi sells to big customers)

Architecture

 

Design Goals: Automate Service Lifecycle

  • catalog (of services)
  • build
  • test
  • ordering
  • provisioning
  • configuration
  • orchestration
  • monitoring
  • issue tracking
  • accounting
  • billing
  • disaster recovery

We wanted our solution to cover all aspects of the lifecycle of service:

  • catalog of available services (appstore);
  • build;
  • ordering;
  • provisioning;
  • configuration;
  • orchestration;
  • monitoring;
  • issue tracking;
  • accounting;
  • billing;
  • disaster recovery (incl. ability to rebuild everything after 10 years):

Design Goals: Unify Service Description

  • no matter where
  • no matter when
  • no matter what
  • no matter which distro or OS (POSIX)
  • no matter which version of distro
  • no matter which architecture
  • no matter complexity
  • with real time, high peformance and resiliency support
  • and at lowest possible cost

What mattered to Nexedi when SlapOS was created is that whatever service we would deploy, we wanted to be able to deploy it fully automatically using the same "service descriptor language", no matter:

  • where (data centre, on premise, inside an airplane, on a smartphone, inside a sensor, etc.);
  • when (today, in 5 years, in 10 years, etc.);
  • what (database service, VM service, application sever service, smart sensor processing service, data buffering service, etc.);
  • which distro (Debian, Ubuntu, Fedora, CentOS, FreeBSD, SuSE, RedHat, Arch, etc.);
  • which version of distro (2016, 2017, 2018, etc.);
  • which architecture (bare metal, VM, x86, ARM, etc.);
  • complexity (unitary service, orchestration of dozens services);
  • real time constraints (no constraint, hard real time).

Why don't you use OpenStack?

We often get questions such as:

  • why don't you use virualisation?
  • why don't you OpenStack?
  • why don't you use Docker?
  • etc.

We usually answer: because they don't work according to industrial grade standards, even now, and because they did not exist in 2008.

The meaning of "does not work" is a bit different for Nexedi and for most open source developers. In Nexedi, we want systems that "work always in the same way and for a very long time", rather than systems that "work sometimes" and are easy to install with a beautiful web site. We want this kind of predictability for everything (build, ordering, etc.). We care more about improving our software than community or documentation. If a solution works but is rejected by most community, we ignore community because we have to serve our customers first (our business model is based on customers, not on venture capital).

The article "Are Linux containers stable enough for production and why Nexedi uses SlapOS nano-containers instead ?" (https://www.nexedi.com/NXD-Blog.Docker.SlapOS.Nano.Container.Elbe) explains for example why we do not use Docker or LXC containers and why we do not plan to use them for ourselves. Until recently, it was very difficult to find anyone who would agree with us (just like OpenStack 10 years ago). But more and more people now understand the problems of binary portability with Linux kernel and its consequence on Docker/LXC containers.

So, we might use "kernel namespaces" with SlapOS.

SlapOS could easily support docker/LXC type containers; we actually already implemented it. But those docker/LXC containers will only work if some strict conditions are met: host OS/Kernel and guest OS/Kernel should be same for example. Sadly most developers do not understand those conditions and do not respect them. It is thus difficult to provide something that works according to our standards.

Why don't you use LXC/Docker?

  • Not portable across Linux distros
  • Not supported outside Linux
  • Still a bit unpredictable on Linux
    ~ $ free -h
        

https://www.nexedi.com/NXD-Blog.Docker.SlapOS.Nano.Container.Elbe

 

Everything is a Service

  • A database service
  • A kvm service
  • A routing service
  • An HTTP cache service
  • An ERPservice
  • etc.
 

SlapOS: Service Operation Automation

    
            ~ $ slapos request mariadb my_db
            ~ $ slapos request kvm my_vm
            ~ $ slapos request re6st-registry my_registry
            ~ $ slapos request cdn my_cdn
            ~ $ slapos request erp5 my_erp
          
        
 

Service Unification from Edge to Space

What mattered to Nexedi when SlapOS was created is that whatever service we would deploy, we wanted to be able to deploy it fully automatically using the same "service descriptor language", no matter:

  • where (data centre, on premise, inside an airplane, on a smartphone, inside a sensor, etc.);
  • when (today, in 5 years, in 10 years, etc.);
  • what (database service, VM service, application sever service, smart sensor processing service, data buffering service, etc.);
  • which distro (Debian, Ubuntu, Fedora, CentOS, FreeBSD, SuSE, RedHat, Arch, etc.);
  • which version of distro (2016, 2017, 2018, etc.);
  • which architecture (bare metal, VM, x86, ARM, etc.);
  • complexity (unitary service, orchestration of dozens services);
  • real time constraints (no constraint, hard real time).

Design Goals: Security and Resiliency

  • insecure network
  • unstable network
  • unstable hardware
  • unstable electricity
  • vanishing code sources

http://iwgcr.org/

And we wanted our solution to be take into account "real world" features of public infrastructures which we had observed and made statistics of:

  • insecure network (anyone can spy it);
  • unstable network (packets are lost, connectivity is lost);
  • unstable hardware (any component can crash);
  • unstable electricity (electricity shortage is always possible).

The article "Downtime statistics of current cloud solutions" (http://iwgcr.org/wp-content/uploads/2013/06/IWGCR-Paris.Ranking-003.2-en.pdf) should give a good overview of the lack of resiliency of cloud, networking and electricity no matter who is the supplier.

Promise Based Minimalist Architecture

  • Master: ERP5 (promise definition, ordering, provisioning, accounting, billing, issue tracking)
  • Slave: buildout (promise execution, build, instantiation, configuration, monitoring)

https://en.wikipedia.org/wiki/Promise_theory

So, we used buildout (http://docs.buildout.org/en/latest/) as the base for our service descriptor language and ERP5 to keep track of "service lifecycle" after we found out that any edge or cloud system can be made of two components: a devops and an ERP (see "SlapOS: SlapOS: A Multi-Purpose Distributed Cloud Operating System Based on an ERP Billing Model" https://ieeexplore.ieee.org/document/6009348). 

For resiliency, we based all our design on the idea that resiliency must be implemented with software and should rely on redundant infrastructure on redundant sites with redundant suppliers. However each site or hardware does not need to be redundant.

This approach was quite successful. By sticking to a very simple and minimal architecture, we could achieve with a small budget what huge community projects such as OpenStack still fail to achieve after 10 years. And we could do much more, because our architecture was more generic.

Nano Containers

  • declarative
  • bare metal
  • multiple versions
  • multiple instances
  • no superuser needed
  • portable across Linux distributions (unlike Docker)
  • portable to other POSIX OS (Android, FreeBSD, etc.)
  • (option) source cache (encouraged)
  • (option) binary cache
  • (option) virtualisation
  • (option) name spaces
  • (option) containerisation (discouraged)
 

re6st

re6st

re6st was created to fix problems of current Internet through an IPv6 overlay network.

In today's Internet, latency is usually sub-optimal and telecommunication providers provide unreliable transit. There are lots of network cuts. DPI systems introduce sometimes data corruption in basic protocols (ex. TCP). Governments add censorship and bogus routing policies, in China for example. There is no way to ensure that two points A and B on the Internet can actually interconnect. The probability of connectivity fault is about 1% in Europe/USA and 10% inside China. It is too much for industrial applications.

Without re6st, SlapOS (or any distributed container system) can not work. If one has to deploy 100 orchestrated services over a network of edge nodes with a 1% probability of faulty routes, the overall probability of failure quickly becomes too close to 100%. There is therefore no way to deploy edge without fixing the Internet first.

This is very easy to understand in China. But it is also true in Europe and USA (maybe not yet in Japan).

re6st routing provides one solution to that. re6st is available in China (license: 中华人民共和国增值电信业务经营许可证:沪A1-20140091). Nexedi has the right to provide global low latency high resiliency IPv6 network for IoT.

In addition to re6st, we use buffering to that we do not lose data sent by edge nodes (gateways or sensors) in case of application server failure for example:

Both re6st and fluentd are used in all IoT deployments done by Nexedi and based on SlapOS.

References

You can find more articles related to SlapOS: