Hyper Open Edge Cloud

Rapid.Space in China

Rapid.Space Infrastructure Services in China
  • Last Update:2019-10-21
  • Version:001
  • Language:en

QR Code

 

Agenda

  • Nexedi
  • Track Record
  • Nexedi in China
  • Open Computing (OCP)
  • OCP Case Study
  • Open Networking
  • Open Networking Case Study

This presentation explains the different options of Rapid.Space in China. It is intended for international companies with presence in China that are looking for a resilient infrastructure with global presence.

We will first introduce Nexedi, its track record and its presence in China.

Then we will then the concepts of open compute project and open networking, both in terms of ecosystem and case study.

It should then become obvious that anyone could save tremendous amount of money by relying on Rapid.Space or similar approaches compated to traditional cloud and networking solutions.

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. We also have a bit of C and golang.

Nexedi - Rapid.Space Cloud

https://rapid.space

Rapid.Space (https://rapid.space/) is a high performance, low cost cloud infrastructure that provides:

  • big servers;
  • CDN;
  • IoT buffering.

It is available in Europe and soon in China through partner company. It is based on SlapOS and Open Compute Project (OCP) hardware, the same as the one used by Facebook, Yahoo! Japan, Booking.com, etc.

Everything about Rapid.Space is public.

Track Record

Nexedi focuses on entreprise projects of mid to large size. We have now nearly 20 years of entreprise IT.

Bip & Go

Big&Go is our largest project with 100.000 invoices generated per hour on a distributed database hosted on both Rapid.Space cloud and OVH cloud..

Automotive factory-in-a-box

Nexedi has implemented an edge computing box for an autonotive company. This edge computing box includes an ERP and MES that can be used to manage a small factory. All services are deployed automatically, including low latency IPv6 network and CDN.

Teralab Big Data Cloud

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

Teralab is a big data infrastructure which provides various services including AI  French government. It is powered by Nexedi's SlapOS software and ITRenew's OCP servers. See: https://www.nexedi.com/success/slapos-IMT-Documents.Teralab.Success.Case

Nexedi in China

 

RapidSpace.cn - Big Data Cloud

 

Grandenet.cn - Low Latency ADN/VPN

 

netk5.com.cn - Entreprise Services

  • Server migration
  • Network integration
  • Network management
 

China Legal Compliance

  • ICP
  • Appstore - www.officejs.cn
  • VPN (national) - www.grandenet.cn
  • CDN (CT/CU/CM + worldwide)
  • IDC - www.xunkongjian.cn
  • DNS server (CT/CU/CM + worldwide)
  • ZK architecture

RapiSpace.cn will block any illegal use in China.

 

China key figures

  • Capacity: 3200 OU servers, 1000 Gbps transit
  • OCP DC: 3 sites
  • Peering: 8 DC, 7 cities
  • Optimised international access: 3 peers
  • CDN: CT/CU/CM +  EU/USA/Japan
  • ...
 

Open Compute Project

 

OCP Server (Mitac, ITRenew)

open source hardware

everyone can copy

2X00€-3X00€ / server (24 core, 256 GB RAM, 4 TB SSD)

69€ / server / month

Used by Facebook, Yahoo Japan, Booking.com, etc.

Server itself costs about 2500€ fully equipped (256 GB RAM, 4 TB SSD, 40 vcores). This price is possible by using re-certified hardware. The hardware is based on OCP standard, the same standard as the one Facebook is using. We use Splitted Desktop (SDS) as our supplier. We actually launched Rapid.Space to also help SDS showcase their technology, since OCP only makes sense economically if one purchases a complete rack (it is incovenient to operate a single server).

OCP Ecosystem

  • Baidu
  • Booking.com
  • Facebook
  • Google
  • Huawei
  • IBM
  • Inspur
  • JD.com
  • Lenovo
  • Nexedi
  • Tencent
  • ...
 

OCP Cloud (SlapOS)

This is how Rapid.Space hardware platform looks like. It is mainly a console to order and monitor Rapid.Space services such as virtual machines or CDN.

OCP DC (Hydro66)

 

OCP Case Study

 

Monthly Price Example - 386 servers

 

Item Quantity Unit Price Total
CPU 7,720 74 571,280
RAM 77,200 6 463,200
Storage SSD 750,000 0.066 49,500
Total (MAX)     571,280
  • Servers
  • Cloud management system
  • DC, electricity
  • Networking (1000 Gbps shared, 27 Mbps / server)
 

Migration Cost - 386 servers

Item Price
P2V 617600
Recycling (estimate) -308800
  308800
 

Cloud Service Range

  Big VM Resilient VM Resilient DB Resilient PaaS CDN
High Performance Compute
High Performance Storage  
High Performance Network (✔)
Edge computing        
Scalable CPU, Storage & RAM  
Optional GPU    
Automatic disaster recovery with daily test   N/A
High Availability      
Typical applications Big applications (ERP, CRM)
Big data / BI (Jupyter, HADOOP)
HPC VDI
Small applications
Small virtual desktop
Software development
High performance MariaDB
High performance Datalake
Software development
Cloud native devops
Web acceleration
IoT buffering
 

DB Intensive Performance

  OCP qemu qcow2 OCP qemu /dev/sdb2 OCP qemu /dev/sdb OCP bare metal OVH OpenStack OVH Dedicated
Duration 85h 85h 12h55 9h51 16h22 8h27
VM Manager SlapOS SlapOS SlapOS N/A OpenStack N/A
Orchestrator SlapOS SlapOS SlapOS SlapOS SlapOS SlapOS
Database NEO + InnoDB NEO + InnoDB NEO + InnoDB NEO + InnoDB NEO + InnoDB NEO + InnoDB
Pystone (CPU test) 150.000 150.000 150.000 150.000 200.000 230.000
vCore 40 40 40 40 64 40
Xeon Generation v2 v2 v2 v2 N/A v3
Frequency 2.8/3.6 GHz 2.8/3.6 GHz 2.8/3.6 GHz 2.8/3.6 GHz 3.1 GHz 2.6/3.3 GHz
Storage 4 TB SATA SSD 4 TB SATA SSD 4 TB SATA SSD 4 TB SATA SSD 4 TB High Speed 4 TB NVME SSD
RAM 256 GB 256 GB 256 GB 256 GB 2 x 120 GB 256 GB
Monthly Price N/A N/A 195€ 195€ 2261€ 769€
Redundancy No No No No Yes No
 

Open Networking

 

Open Networking Switch (Edgecore)

open source hardware - everyone can copy

16 x 40 Gbps - X00-X000€

Used by Facebook, Microsoft Azure, Tesla in China

Replaces traditional routers at Tencent

Powered by Broadcom, Bigfoot, Cavium, Mellanox

We now need a switch. Ideally, we would prefer an unmanaged switch (cheaper, safer, more stable) because all networking is software based on SlapOS. But we would find for less than 500€ a 16 port 40 Gbps managed switch (also open hardware). With the right splitting cables, this is same as a 64 port 10 Gbps switch.

Open Networking Ecosystem

  • Bigfoot
  • Broadcom
  • Cavium
  • Delta
  • Dell
  • Edge-core
  • HPI
  • Ingrasys
  • Inventec
  • Lenovo
  • Mellanox
  • Netberg
  • Stordis
  • ...
 

Congestion Management - 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.

Open Networking Case Study

Item Quantity RMB / month / unit Total RMB / month
Site router 17 600 12,750
Site maintenance 17 900 15,300
100 Mbps Access 47 600-2,500 28,200-130,250
100 Mbps Optimised Transit 1 40,000 40,000
Total     96,250-198,300

At least one 100 Mbps Internet access per site

Two or more access from different suppliers on large sites

 

Conclusion

 

Cost Savings

  • Rapid.Space: 2 to 20 times less
  • Grandenet: 2 to 20 times less
 

Thank You

 
  • Nexedi SA
  • 147 Rue du Ballon
  • 59110 La Madeleine
  • France