Happy 2019

Happy 2019
Happy New Year

Hello, une nouvelle année commence pour nous tous comme toujours beaucoup de bonnes choses et à bientôt pour ceux qui croiseront ma route et même à ceux qui ne seront qu’ici. Amusez vous, profitez et de belles choses pour ces nouveaux jours.
—-
Hello, a new year begins for us all as always many good things and soon for those who will cross my path and even those who will be only here. Have fun, enjoy these beautiful things for these new days.

Optical amplification 100G and more for distance > 10km

In optical communication network, signal travels through fibers in every large distances without significant attenuation. However, when it comes to the distance up to hundreds of kilometers, to amplify the signal during transit becomes rather essential. In this case, an optical fiber amplifier is required to achieve signal amplification in long distance optical communication. This article aims to give a brief introduction to the most deployed fiber [amplifier— Erbium doped fiber amplifier (EDFA)](https://www.4fiber.com/wdm-optical-network/edfa.html?lipi=urn%3Ali%3Apage%3Ad_flagship3_pulse_read%3BzkwpjuCWQbyowJSDogVyyg%3D%3D).

What is EDFA?

An EDFA is an optical or IR repeater that amplifies a modulated laser beam directly, without opto-electronic and electro-optical conversion. Generally speaking, it is an optical repeater device that is used to boost the intensity of optical signals being carried through a fiber optic communications system.

Working Principle of EDFA

EDFA serves as a kind of optical amplifier which is doped with the rare earth element erbium so that the glass fiber can absorb light at one frequency and emit light at another frequency. An external semiconductor laser couples light into the fiber at infrared wavelengths of either 980 or 1480 nanometers. This action excites the erbium atoms. Additional optical signals at wavelengths between 1530 and 1620 nanometers enter the fiber and stimulate the excited erbium atoms to emit photons at the same wavelength as the incoming signal. This action amplifies a weak optical signal to a higher power, effecting a boost in the signal strength. The following picture shows 13dBm output C-band 40 channels booster EDFA for DWDM Networks.

![EDFA](/content/images/2018/03/EDFA.jpeg)

The Advantages of EDFA

The EDFA obtains the advantages of high gain, wide bandwidth, high output power, high pumping efficiency, low insertion loss, and it is not sensitive to the polarization state.

It provides in-line amplification of signal without requiring electronics, and the signal does not need to be converted to electrical signal before amplification. The amplification is entirely optical.
It provides high power transfer efficiency from pump to signal power.
The amplification is independent of data rate.
The gain is relatively flat so that they can be cascaded for long distance use. On the debit side, the devices are large. There is gain saturation and there is also the presence of amplified spontaneous emission (ASE).
The Applications of EDFA

The EDFA was the first successful optical amplifier and a significant factor in the rapid deployment of fiber optic networks during the 1990s. By adopting it in conventional optical digital communication system applications, we can save a certain amount of optical repeaters. Meanwhile, the distance relay could also be increased significantly, which is vital for the long-haul fiber optic cable trunking systems. The EDFA is usually employed in these circumstances:

EDFA can be employed in the high-capacity and high-speed optical communication system. It offers a constructive and ideal solution for handling low sensitivity of receivers and short transmission distances because of a lack of OEO repeater.

In addition, EDFA can be adopted in long-haul optical communication system, such as land trunk optical transmission system and the submarine optical fiber cable transmission system. It helps to lower construction cost dramatically by reducing the quantity of regenerative repeaters.

Moreover, EDFA can also be employed in wavelength-division multiplexing (WDM) system, especially dense wavelength-division multiplexing (DWDM) system. It enables the problems of insertion loss to be solved successfully and reduces the influences of chromatic dispersion.

Conclusion

By far, being the most advanced and popular optical amplifier, EDFA has been widely adopted in the optical fiber communication networks. Featured by flat gain over a large dynamic gain range, low noise, high saturation output power and stable operation with excellent transient suppression, it surely will capture a rather vital and indispensable position in optical communication in the near future.

Sample [EDFA products](www.4fiber.com)

Other documentations

https://www.packetlight.com/products/infrastructure-products/pl-1000il-optical-amplifier

Optimizing and Scaling on a Leaf-Spine Architecture

Posted by (and sources): Mike Peterson on May 05, 2016
Source: her/ici

Description

The Internet of Things (IoT) and the proliferation of virtualization have caused traffic between devices in the data center to grow. Referred to as “east-west traffic,” this term accounts for traffic going back and forth between servers in a data center.

When you run lots of east-west traffic through a topology designed for north-south traffic (traffic that enters and exits the data center), devices connected to the same switch port may contend for bandwidth – and end-users experience poor response time.

If hosts on one access switch need to quickly communicate with hosts on another access switch, uplinks between the access layer and aggregation can be a point of congestion. A common three-tier network design may worsen the issue, constraining the location of devices like virtual servers.

Moving to a Leaf-Spine Architecture

That’s where leaf-spine architecture comes in, scaling horizontally through the addition of spine switches. This two-layer topology allows devices to be exactly the same number of segments away.

With each leaf switch connecting to each spine connection, the number of spine switches is limited to the number of uplink ports on the leaf. The most common leaf switches come with only four 40G QSFP+ uplink ports, limiting your network to only four spine switches. This starts to limit network scalability.

One way to achieve more scale is to break the 40G SR4 channel into four 10G duplex channels, turning the four 40G uplink ports into 16 available uplinks. This increases the number of spine switches that can be a part of the mesh network to 16, providing four times the scalability.

Scaling Networks: 10G vs. 40G

Let’s use an example to compare scaling in leaf-spine architecture between 10G and 40G networks.

With 40G uplinks, the number of spine switches is fixed at four, based on the leaf having four uplinks. Typically, each spine has a total of four line cards. These line cards come with 36 40G ports per line card. The total number of available ports to connect to leaf switches is 144; each leaf has 48 ports to connect to network devices, allowing for a maximum of 6,912 computers to connect to the 40G mesh network.

When you scale out on a 10G network, scaling is increased by a factor of four. Each 40G uplink is broken into four 10G channels, allowing for 16 spine switches. With four line cards, and 36 40G ports per line card split into 10G legs, there are a maximum of 576 leaf switches (144 ports x 4). With each leaf having 48 ports, you can connect 27,648 computers – four times the scaling throughout the mesh network.

10G Channels: Potential Obstacles

Moving to four 10G channels in leaf-spine architecture introduces a new concern: Latency (the amount of time it takes for a packet of information to travel from point A to point B) increases because the pipes are split into smaller lanes. The smaller the lanes, the slower the traffic. Although throughput remains the same, latency increases.

One of the biggest challenges to implementing a mesh network is cabling. Mesh networks require LC patch cords to create a cross-connect, ensuring that all leaf switches and spines are properly connected. A cross-connect is created in the main distribution area (MDA), creating several cabling issues: insertion loss, maintaining polarity, increase in cable counts, etc. Rack challenges include density, required U space and power availability.

To create the 10G channel, a complex cross-connect must be created. Each eight-fiber MPO port on the switch is broken up into an LC duplex connection; 144 MPOs become 576 LC duplex connections per switch, for a total of 18,432 LC duplex ports (both sides of the cross connect). To connect the 10G channels to each leaf and spine, a total of, 9,216 LC duplex patch cords are needed. As a result, additional channels for MACs (moves, adds and changes), cable routing and space constraints are possible.

This essentially breaks an MPO into four lanes and makes an LC connection. Each lane is combined with lanes from other spines and converted back into an eight-fiber MPO (Base-8) with four channels from four different spine switches. Cable management, space utilization, documentation and labeling become extremely difficult to troubleshoot and maintain.

Shuffle Cassettes Save Space and Reduce Complexity

There’s a new leaf-spine architecture solution available that drastically reduces the amount of space needed, as well as the number of cables in the MDA: Belden shuffle cassettes.
These cassettes eliminate the need to create a cross-connect to separate 40G channels into 10G, and recombine to connect to each leaf, handling lane reassignments internally. Each shuffle cassette has four MPOs in and out; each leaf requires four shuffle cassettes.

COMPARING TOTAL MODULES

<table><tbody><tr><th>Traditional MPO-LC-MPO</th><th>Belden Shuffle Cassette</th><th>*Savings*</th></tr><tr><td>704 modules</td><td>416 modules</td><td>*288 modules*</td></tr><tr><td>176U space</td><td>104U space</td><td>*72U (roughly 1.6 racks)*</td></tr><tr><td>9,216 patch cords</td><td>2,304 patch cords</td><td>*6,912 patch cords*</td></tr></tbody></table>

By utilizing the same connector, reducing connections and standardizing on components across the channel, Belden’s shuffle cassettes allow for scaling in leaf-spine architecture, reduce the opportunity for human error, speed up deployment time and reduce time spent on MACs. By using a shuffle cassette that fits into any Belden housings, you reclaim valuable floor space.

Automation for network

Ansible on network world

Presentations

Ansible (Python)

Replay of Ansible presentation made by Francois 

Ansible tutorial
Network automation with ansible

First configuration of network devices

Zero Touch Provisioning

ZTP overview

Network definition language

looks a lot like cmdb light, same approach to config and design generation language defines objects and comiled to fill db they generate complete templates  
Robotron top down network management at facebook scale

Talk 2: Wedge100 + Backpack: From the Leaf to the Spine Zhiping Yao + Xu Wang, Facebook

Other tools (not Ansible)

Opencontrail

[multi-hypervisor-support-using-opencontrail]

(http://www.opencontrail.org/multi-hypervisor-support-using-opencontrail/)  
[qualisystems](http://www.qualisystems.com/)

Napalm

[napalm automation]

(https://github.com/napalm-automation/napalm)
[Napalm @spotify on Github](https://github.com/spotify/napalm)

More for DevOps Chef (ruby)

[chef]

(https://learn.chef.io/)  
[Puppet](https://puppet.com/) (not used @Criteo)

Python for Network

[https://pynet.twb-tech.com/](https://pynet.twb-tech.com/)

Criteo Tools for network diff between 2 configuration files (Cisco/Arista) :  
[https://github.com/criteo/netcompare?files=1](https://github.com/criteo/netcompare?files=1)

Why use Docker / Ansible in front of Puppet / Chef :  
[http://thenewstack.io/are-docker-users-migrating-to-ansible-and-away-from-puppet-and-chef/](http://thenewstack.io/are-docker-users-migrating-to-ansible-and-away-from-puppet-and-chef/)

# Monitoring / Graphs

* Time series DB
[Opentsdb](http://opentsdb.net/)
[Grafana](http://play.grafana.org/#/dashboard/db/grafana-play-home?panelId=6&fullscreen&edit)

* Front-end
[Graphite](https://graphiteapp.org/)
[Elastic kibana](https://www.elastic.co/fr/products/kibana)
[ElasticSearch](https://www.elastic.co/fr/products/elasticsearch)

# Virutalenv (VM/libvirt/container/…)

If you want to test some apps/stuff you can use one of this « tools »

– Docker : [https://www.docker.com/](https://www.docker.com/)
– Virtualenv : (more for dev) [http://virtualenv.readthedocs.org/en/latest/](http://virtualenv.readthedocs.org/en/latest/)
– Vagrant : (more for dev) [https://www.vagrantup.com/](https://www.vagrantup.com/)

# Other

[http://druid.io/](http://druid.io/)

## Videos and presentations

Storm usage at Criteo: [http://www.infoq.com/fr/presentations/storm-criteo](http://www.infoq.com/fr/presentations/storm-criteo)

Youtube Network Automation and Programmability Abstracation Layer [https://www.youtube.com/watch?v=93q-dHC0u0I](https://www.youtube.com/watch?v=93q-dHC0u0I)

<iframe allowfullscreen= »allowfullscreen » height= »314″ src= »//www.youtube.com/embed/93q-dHC0u0I » width= »560″></iframe>

@34:47 you will find Steve Feldman.

The only feldman I know is him ?

<iframe allowfullscreen= »allowfullscreen » height= »314″ src= »//www.youtube.com/embed/h8VWASQB8wk » width= »560″></iframe>

Blog : [https://pynet.twb-tech.com/blog/automation/cisco-ios.html](https://pynet.twb-tech.com/blog/automation/cisco-ios.html)

# Tools for DEV

[http://sourabhbajaj.com/mac-setup/Python/virtualenv.html](http://sourabhbajaj.com/mac-setup/Python/virtualenv.html)  
[http://hetland.org/writing/instant-python.html](http://hetland.org/writing/instant-python.html)

# Tools

[https://www.esev.com/blog/post/2015-01-pgp-ssh-key-on-yubikey-neo/](https://www.esev.com/blog/post/2015-01-pgp-ssh-key-on-yubikey-neo/)

# A Trier:

[https://github.com/digitalocean](https://github.com/digitalocean)  
[https://www.digitalocean.com/company/blog/](https://www.digitalocean.com/company/blog/)  
[http://xplproject.org.uk/](http://xplproject.org.uk/)  
[http://networkstatic.net/programming/](http://networkstatic.net/programming/)  
[https://www.opennetworking.org/blog/?Itemid=316](https://www.opennetworking.org/blog/?Itemid=316)  
[http://northboundnetworks.com/collections/flowmaker](http://northboundnetworks.com/collections/flowmaker)  
[https://herdingpackets.net/2014/02/06/using-the-cisco-csr1000v-in-gns3-with-virtualbox/](https://herdingpackets.net/2014/02/06/using-the-cisco-csr1000v-in-gns3-with-virtualbox/)  
[http://highscalability.com/](http://highscalability.com/)

Network BGP on TOR

Layer 3 design with spine and leaf

Routing

* prefix list automation:
[Bgpq3](https://github.com/snar/bgpq3)

# Network design

### Google new network design
Read this paper: [Conferences sigcomm](http://conferences.sigcomm.org/sigcomm/2015/pdf/papers/p183.pdf)

### Facebook DC design

Information about 1st design @FB:
http://networkingnerd.net/2013/09/23/your-data-center-isnt-facebook-and-thats-just-fine/

Facebook DC design Next Gen:
Introducing data-center fabric the next generation facebook DC network

A video presentation about L3 spine and leaf @FB (useful demo @2’22 »)

[https://www.youtube.com/watch?v=mLEawo6OzFM&t=142]

Pictures of FB DC:
[Photo tour new facebook data-center in iowa (2014)](http://www.datacenterknowledge.com/archives/2014/11/20/photo-tour-new-facebook-data-center-in-iowa/)

Servers:
http://si.wsj.net/public/resources/images/BN-FO246_1114fb_WN_20141114145545.jpg
FB servers
### LinkedIn DC design plus L3

DC design spine and leaf and water-cooling at LinkedIn  
[http://www.networkworld.com/article/3161184/data-center/linkedin-pumps-water-down-to-its-server-racks-uses-an-interesting-spine-and-leaf-network-fabric.html](http://www.networkworld.com/article/3161184/data-center/linkedin-pumps-water-down-to-its-server-racks-uses-an-interesting-spine-and-leaf-network-fabric.html)

And from linkedin blog:
[https://engineering.linkedin.com/blog/2016/03/project-altair–the-evolution-of-linkedins-data-center-network](https://engineering.linkedin.com/blog/2016/03/project-altair–the-evolution-of-linkedins-data-center-network)

### Other company L3 design

An old article from Metadata blog:
[http://muratbuffalo.blogspot.fr/2010/11/portland-scalable-fault-tolerant-layer.html](http://muratbuffalo.blogspot.fr/2010/11/portland-scalable-fault-tolerant-layer.html)

### Tools for network BGP and design

[http://www.pica8.com/company/press-releases/pica8-adds-bgp-support-extends-routing-to-the-data-center-edge.php](http://www.pica8.com/company/press-releases/pica8-adds-bgp-support-extends-routing-to-the-data-center-edge.php)

[http://www.cisco.com/c/en/us/products/collateral/switches/nexus-5000-series-switches/whitepaperc11-522337.html](http://www.cisco.com/c/en/us/products/collateral/switches/nexus-5000-series-switches/whitepaperc11-522337.html)

IETF draft for L3 in the DC:  
[https://tools.ietf.org/html/draft-ietf-rtgwg-bgp-routing-large-dc-00](https://tools.ietf.org/html/draft-ietf-rtgwg-bgp-routing-large-dc-00)

Presentation from Nanog about  

[https://www.youtube.com/watch?v=GNe_mIxQZcc](https://www.youtube.com/watch?v=GNe_mIxQZcc)

From Arista:  
[https://www.youtube.com/watch?v=TLbzvbfWmfY](https://www.youtube.com/watch?v=TLbzvbfWmfY)

LoadBalancer

How to load balance applications in a L3 DC (Replay of a meetup)  
[https://www.youtube.com/watch?v=MKgJeqF1DHw](https://www.youtube.com/watch?v=MKgJeqF1DHw)

Hardware

[http://www.qantas.com.au/travel/airlines/our-network/au/en](http://www.qantas.com.au/travel/airlines/our-network/au/en)

Latex premier documents

Utilisation de latex, mise en place d’un Docker avec tous les outils :

Dockerfile:
<pre class= »prettyprint linenums »><code>
#This is a comment FROM debian:jessie-slim
#FROM debian:jessie MAINTAINER Jerome Baudet <jerome@baudet.io>
RUN echo « deb http://ppa.launchpad.net/ansible/ansible/ubuntu trusty main » >> /etc/apt/sources.list
RUN apt-key adv –keyserver keyserver.ubuntu.com –recv-keys 93C4A3FD7BB9C367
RUN apt-get update && apt-get install –fix-missing -y sudo curl texlive-latex-base texlive-xetex latex-xcolor texlive-math-extra texlive-latex-extra texlive-fonts-extra texlive-bibtex-extra fontconfig lmodern preview-latex-style texlive-latex-recommended tipa prosper preview-latex-style cabal-debian pandoc pandoc-data texlive-doc-fr sshpass openssh-client openssh-server vim vim-latexsuite
RUN apt-get autoremove
RUN apt-get clean
RUN mkdir /var/run/sshd
RUN mkdir /root/.ssh
COPY authorized_keys /root/.ssh/
RUN sed -i ‘s/PermitRootLogin prohibit-password/PermitRootLogin yes/’ /etc/ssh/sshd_config
#SSH login fix. Otherwise user is kicked off after login
RUN sed ‘s@sessions*requireds*pam_loginuid.so@session optional pam_loginuid.so@g’ -i /etc/pam.d/sshd
ENV NOTVISIBLE « in users profile »
RUN echo « export VISIBLE=now » >> /etc/profile
#will ease up the update process
#updating this env variable will trigger the automatic build of the Docker image
#ENV PANDOC_VERSION « 1.19.2.1 »
#install pandoc #
RUN cabal update && cabal install pandoc-${PANDOC_VERSION} #
EXPOSE 2200
VOLUME [« /data »] CMD [« /usr/sbin/sshd », « -D »]
#ENTRYPOINT /bin/bash
</pre></code>

What You Need to Know About Alien Crosstalk Today

![Alien Crosstalk](http//info.belden.com/webadmin/blog/images/alien-crosstalk.jpg « Alien Crosstalk »)

Contact | Subscribe
– – – – – –

The industry has been predicting the growth of 10GBASE-T for years, and it’s finally happening. More networks are planning 10G migrations. Why? Due to demand from more advanced devices, users and applications.

But new concerns come into play with this Ethernet standard. Alien crosstalk – the interference caused by wire pairs in one cable inducing noise into other wire pairs in adjacent cables – is the transmission parameter that most significantly impacts 10GBASE-T performance.

 

## What is Alien Crosstalk?

Alien crosstalk is a combination of alien near-end crosstalk (NEXT) and alien far-end crosstalk (FEXT); the noise source originates from a common mode signal that is converted onto the differential mode signal through some type unbalance on cable and components.

 

## Why is Alien Crosstalk Bad for Today’s Applications?

In high-speed, high-bandwidth applications – used today to accommodate more users and more devices – alien crosstalk can cause many problems. In a cable bundle, it’s possible that cable pairs in one cable pick up interference from pairs of another cable. The digital signal processors (DSPs) used in [10GBASE-T architectures](http://www.ethernetalliance.org/subcommittees/10gbase-t/faqs/) can’t remove unpredictable exterior noise. Noise sensitivity increases at higher frequencies, such as 500 MHz, which is the highest frequency of Category 6A cabling.

This interference isn’t just a nuisance; it has the potential to shut your entire network down – which leads to unplanned downtime, financial losses, a productivity nosedive and unhappy users.

 

## What Causes Alien Crosstalk?

Alien crosstalk originates from a common mode signal that is converted onto the differential mode signal through some type of unbalance on the cable and components. Category 6A cabling and components are designed and tested to reduce alien crosstalk to a level low enough that it does not interfere with the differential mode signal.

Make sure all components are [Category 6A](http://www.belden.com/blog/datacenters/index.cfm?page=5). This is especially important to note for patch cords. When high-quality, Category 6 cable is tested, it may pass Category 6A patch cord requirements because only near-end crosstalk and return loss are measured. When these patch cords are bundled and placed in a Category 6A channel, however, the channel fails alien crosstalk requirements. Why? Because Category 6 cable isn’t designed to handle high data-speed requirements.

Don’t over-tighten cable ties. Cable ties that increase tension can force cables together and impact alien crosstalk. When using cabling ties to dress the cabling, make sure the ties are free to rotate after tightening.

 

## What Do Standards Say About Alien Crosstalk?

[[Poll Goes Here]]Alien crosstalk reduces cabling’s operational bandwidth due to increased channel noise levels; as a result, ANSI/TIA standards state that [Category 6A cabling](http://www.belden.com/blog/datacenters/Installing-Category-6A-The-Future-is-Now.cfm) best meets the demands of 10G. (According to current standards, Class E and Category 6 cabling aren’t recommended for new 10GBASE-T installations over 37m.)

Here are a few reasons why Category 6A is the best choice:

– Achieving [10GBASE-T](http://www.hpctoday.com/best-practices/10-things-to-know-before-deploying-10-gigabit-ethernet/) over copper requires 500 MHz bandwidth and full duplex transmission, which Category 6A provides
– It must follow stringent ISO/IEC and ANSI/TIA transmission-parameter regulations
– It uses specially designed jackets and cross-webs that physically separate internal twisted pairs from external twisted pairs, ensuring low alien crosstalk
– It provides a guaranteed alien crosstalk margin above minimum TIA-568-C.2 requirements
– Category 6A specifications allow compliant cabling for 10GBASE-T transmission

Belden’s Category 6A cabling solutions provide a simple design without compromising performance or quality, ensuring robust performance and reliable networks. Learn more about our [10GSX cables](http://info.belden.com/10gxs), which feature a smaller diameter, a rounder jacket, a smaller bend radius, fewer twists and easy-to-remove barrier tape.

<div style= »text-align: center; »>[![IoT WP CTA](http//info.belden.com/webadmin/blog/images/IoT WP CTA_86060.png « IoT WP CTA »)](http://info.belden.com/ecos/iot-convergence-wp)</div>Source: Spine and Leaf (1st) test