pygnmi 15. Overview of nornir_pygnmi

Hello my friend,

Are you looking for building network automation at scale leveraging the future-proof model-driven network automation? Besides attending our zero-to-hero network automation training and network automation with nornir, we suggest you to take a look at nornir_pygnmi, the new plugin we have created for Nornir to simplify management of network devices with gNMI.

1
2
3
4
5
No part of this blogpost could be reproduced, stored in a 

retrieval system, or transmitted in any form or by any 

means, electronic, mechanical or photocopying, recording, 

or otherwise, for commercial purposes without the 

prior permission of the author.

Is GNMI a Good Interface for Network Automation?

Yes, it is. GNMI is one of the most recent interfaces created for the management plane, which allows you to manage the network devices (i.e., retrieve configuration and operational data, modify configuration) and collect the streaming or event-driven telemetry. Sounds like one-size-fits-all, isn’t it? On top of that, GNMI supports also different transport channels (i.e., encrypted and non-encrypted), which makes it suitable both for lab testing and for production environment. You may feel that we are biased to gNMI, and you are right. Actually, that is a Continue reading

Tools 11. Running Prometheus Exporters on Dis-aggregated Data Center Switches (Cisco NX-OS, NVIDIA Cumulus Linux, Arista EOS)

Hello my friend,

Monitoring is a key component of any IT and network systems, as it gives us possibility to know what is going on in those systems. That is important at least for two reasons:

• Observing actual status of the system counters
• Analyze trends over the time period

Let’s see how we can attempt to normalize collection of such a data with Prometheus from dis-aggregated network switches empowered by Linux.

1
2
3
4
5
No part of this blogpost could be reproduced, stored in a 

retrieval system, or transmitted in any form or by any 

means, electronic, mechanical or photocopying, recording, 

or otherwise, for commercial purposes without the 

prior permission of the author.

What Is the Relation Between Monitoring and Automation?

The answer would be “closed-loop automation” which is also often sold as IBN (Intent-Based Networking). It works as the collected data is automatically compared against certain patterns or thresholds and in case of threshold violation, some automated action can be triggered. It can be as simple as a notification via a mail, a Slack channel or a Telegram chat; however, it can be also a more complicated, where some automated workflow is triggered to perform corrective configuration Continue reading

Automation 16. How to Prepare Cisco Nexus 9000 (Cisco NX-OS) to Be Automated with NETCONF and GNMI. Complete Guide.

Hello my friend,

When I booted up my laptop in the morning I was aiming to write something simple and easy: it is a summer and I don’t want to spend the whole Sunday on writing a blogpost. However, the topic, which is supposed to be an easy ride, appeared to be a bit more complicated. Well, it is even better then, as the official documentation to the topic is unfortunately vague.

1
2
3
4
5
No part of this blogpost could be reproduced, stored in a 

retrieval system, or transmitted in any form or by any 

means, electronic, mechanical or photocopying, recording, 

or otherwise, for commercial purposes without the 

prior permission of the author.

Shall I use NETCONF or GNMI for automation?

The answer to this question, as usual, it depends. The first fact to think about is whether your platform supports any of these protocols. For example, the platform we will talk about today started supporting GNMI relateively recently and, to my surprise, it supports quite a legacy version of GNMI. At the same time, NETCONF, being an older technology compared to GNMI, are supported more robust. On the other hand, there are platforms, such as Nokia Continue reading

Automation 15. The Good, The Bad and the Ugly of Model-Driven Network Automation Featuring Cisco, Nokia, and OpenConfig YANG

Hello my friend,

All of us (definitely me, at least) are always thrilled hearing news from network vendors on their implementation of model-driven interfaces for network management. Having spent years automating network devices in a text-based paradigm (i.e., from CLI-based automation to full fledged configuration rendering with a replacement), I’m a firm believer that model-driven approach based on YANG modules and protocols such as GNMI, NETCONF, and RESTCONF, is a proper way to go. Recently we disclosed the development we are doing in terms of network topology visualization with DANT. And today we’d like to share lessons learned based on that experience.

We planned to write this blogpost for a few weeks if not months, but due to various reasons it was delayed. We are delighted to finally post it, so that you can get some useful ideas how you can build your own CI/CD pipeline with GitHub, probably the most popular platform for collaborative software development.

1
2
3
4
5
No part of this blogpost could be reproduced, stored in a 

retrieval system, or transmitted in any form or by any 

means, electronic, mechanical or photocopying, recording, 

or otherwise, for commercial purposes without the 

prior permission  Continue reading

pygnmi 14. Using skip-verify for self-signed certificates

Hello my friend,

lab testing is one of the crucial parts of any new technology introduction. At the same time, we know that in labs we typically have a lower security requirements. For example, we use SSL certificates, but we don’t have proper PKI and, therefore, certificates are self-signed.

1
2
3
4
5
No part of this blogpost could be reproduced, stored in a 

retrieval system, or transmitted in any form or by any 

means, electronic, mechanical or photocopying, recording, 

or otherwise, for commercial purposes without the 

prior permission of the author.

Is GNMI a Good Interface for Network Automation?

Yes, it is. GNMI is one of the most recent interfaces created for the management plane, which allows you to manage the network devices (i.e., retrieve configuration and operational data, modify configuration) and collect the streaming or event-driven telemetry. Sounds like one-size-fits-all, isn’t it? On top of that, GNMI supports also different transport channels (i.e., encrypted and non-encrypted), which makes it suitable both for lab testing and for production environment. You may feel that we are biased to gNMI, and you are right. Actually, that is a reason why we created pyGNMI library, Open Continue reading

Random. How to Choose Your Desktop Operating System for Network Automation Development

Hello my friend,

After writing quite long and complicated previous blogpost about CI/CD with GitHub, I need some therapy to write something light and chill. I decided to choose the setup of the working space for development and utilisation of the network automation and, in general, network design and operations. Though I don’t pretend to be absolutely objective and unbiased, as it is simply not possible, I intend to share some observations I did from my own experience and discussions with our network automation students, which I hope will be interesting for you.

1
2
3
4
5
No part of this blogpost could be reproduced, stored in a 

retrieval system, or transmitted in any form or by any 

means, electronic, mechanical or photocopying, recording, 

or otherwise, for commercial purposes without the 

prior permission of the author.

Why Is It Important?

During our Zero-to-Hero Network Automation Trainings, and other trainings as well, we talk a lot about choice of tools to build automation solutions: they shall be fit for purpose and easy to use. However, in addition to that, you should also feel a fun, when you utilise them. It may sound odd, as we are Continue reading

Automation 14. Deep dive into Building CI/CD for Network Automation and Software Development with GitHub Actions

Hello my friend,

We planned to write this blogpost for a few weeks if not months, but due to various reasons it was delayed. We are delighted to finally post it, so that you can get some useful ideas how you can build your own CI/CD pipeline with GitHub, probably the most popular platform for collaborative software development.

1
2
3
4
5
No part of this blogpost could be reproduced, stored in a 

retrieval system, or transmitted in any form or by any 

means, electronic, mechanical or photocopying, recording, 

or otherwise, for commercial purposes without the 

prior permission of the author.

Regards

A lot of lessons about building the CI/CD pipelines and importance of unit testing and linting checks I learned from a colleague of mine, Leigh Anderson, whom I’m very grateful for that.

CI/CD Overview

CI/CD is an approach, which is very often used in software development, and discussed outside of that area. It stands for:

• CI (Continuous Integration) is a process, where the created software (for sake of simplicity, any piece of code) is getting ready to be deployed.
• CD (Continuous Deployment) is a process, where the software, which is ready for deployment, is actually deployed Continue reading

6 Years Creating Content for You

Hello my friend,

Typically on this date, June the 5th we celebrate the birthday of our company, Karneliuk.com. It started with a blog back in 2016 and since then we are constantly creating, what we believe is, interesting and useful educational content in the area of network technologies and network automation. We thank you a lot for being with us all this time!

Technically, we started blog earlier than June the 5th, but on that date we published our first blogpost about interconnecting Cisco IOS XR and Nokia SR OS VMs, which defined the course of the blog and the direction for the company – multivendorness. We breath multivendor network technologies every day in heterogeneous networks, which our team support for our companies and customers daily. And we build multivendor network automation to unleash the true potential of networks and IT systems being an enabler for applications and user services, rather than an obstacle draining time and money of organizations.

So, what have we done in the past twelve months? Let’s take a look.

Our Own Prometheus Exporter

One of the interesting experiences we’ve obtained in the software development was the development of Prometheus exporter to report trace Continue reading

Infrastructure 4. How to Run Cisco Nexus 9000v in Proxmox to Lab Cisco Data Centre

Hello my friend,

We use Proxmox in our Karneliuk Lab Cloud (KLC), which is a driving power behind our Network Automation and Nornir trainings. It allows to run out of the box the vast majority of VMs with network opening systems: Cisco IOS or Cisco IOS XR, Arista EOS, Nokia SR OS, Nvidia Cumulus, and many others. However, when we faced recently a need to emulate a customer’s data centre, which is build using Cisco Nexus 9000 switches, it transpired that this is not that straightforward and we had to spend quite a time in order to find a working solution. Moreover, we figured out that there are no public guides explaining how to do it. As such, we decide to create this blog.

1
2
3
4
5
No part of this blogpost could be reproduced, stored in a 

retrieval system, or transmitted in any form or by any 

means, electronic, mechanical or photocopying, recording, 

or otherwise, for commercial purposes without the 

prior permission of the author.

How Does KLC Help with Automation?

A lot of network automation trainings worldwide imply that a student has to build a lab his/her-own. Such an approach, obviously, is the easiest for Continue reading

Practical OpenStack #005. How Containers help to Build OpenStack

Hello my friend,

Maintaining application in Linux may be a challenging task, especially when you have a lot of them running on a single host. One of the problems contributing to it is a dependencies management. OpenStack, in its turn, consists of a huge amount of services, which needs to run together. Containers help to solve the problem of a dependencies’ management, and, therefore, helps to setup OpenStack. In this video you will learn how that happens

Don’t Forget to Learn How to Automate All the Things

Video to the Topic

Need Help? Contract Us

If you need a trusted and experienced partner to automate your network and IT infrastructure, get in touch with us.

P.S.

If you have further questions or you need help with your networks, we are happy to assist you, just send us a message. Also don’t forget to share the article on your social media, if you like it.

BR,

Anton Karneliuk

Practical OpenStack #004. Build Real World Setup

Hello my friend,

Believe or not, but the vast majority of the OpenStack Certified OpenStack Administrator (COA) trainings are not touching the architecture of the OpenStack clouds. Instead, they focus solely on the operational aspect, so that the COA certified engineer can manage OpenStack cloud. We believe it is not correct, as without how something is build it is very difficult to troubleshoot something or even guarantee a decent level of performance. Hence, we show today the full scale OpenStack cluster setup, which you can build yourself in your lab, as well as hyper converged cluster, which is suitable for building of 5G far edge locations.

Don’t Forget to Learn How to Automate All the Things

Video to the Topic

Need Help? Contract Us

If you need a trusted and experienced partner to automate your network and IT infrastructure, get in touch with us.

P.S.

If you have further questions or you need help with your networks, we are happy to assist you, just send us a message. Also don’t forget to share the article on your social media, if you like it.

BR,

Anton Karneliuk

Practical OpenStack #003. Why these videos

Hello my friend,

OpenStack isn’t new, since the first release on 21 October 2010, the 25 releases were already published bringing new capabilities and services to OpenStack users. Moreover, these days the micro services leveraging Kubernetes and containers are the main trend for application development, whereas virtual machines going in the past. That’s true to a degree; however, the OpenStack is a perfect platform to build a normalised infrastructure for your Kubernetes, which is by the way provide some key services to Kubernetes, such as load balancing.

Don’t Forget to Learn How to Automate All the Things

Video to the Topic

Need Help? Contract Us

If you need a trusted and experienced partner to automate your network and IT infrastructure, get in touch with us.

P.S.

If you have further questions or you need help with your networks, we are happy to assist you, just send us a message. Also don’t forget to share the article on your social media, if you like it.

BR,

Anton Karneliuk

Practical OpenStack #002. Why OpenStack

Hello my friend,

In the previous video you have learn about the core components of any cloud. It was quite straightforward when we remove all unneeded hype, isn’t it? At the same time, you may start thinking “Hey, that’s cool. How can I get it?” That’s a very good question. And in the new video we are covering the principles of interacting with the cloud from the user perspective and provide a quick overview of the existing public and private clouds as well as highlights, why we have chosen OpenStack as a cloud platform for Karneliuk projects and for these videos.

Don’t Forget to Learn How to Automate All the Things

Video to the Topic

Need Help? Contract Us

If you need a trusted and experienced partner to automate your network and IT infrastructure, get in touch with us.

P.S.

If you have further questions or you need help with your networks, we are happy to assist you, just send us a message. Also don’t forget to share the article on your social media, if you like it.

BR,

Anton Karneliuk

Practical OpenStack #001. Cloud Computing

Hello my friend,

There is still quite a bit of a hype around a term “cloud computing” these days. The whole industry of hardware manufacturers or software vendors claiming that they are products are “cloud native”, “cloud ready”, or “build for clouds”. Well, we don’t like clouds, we like sunny weather. Let’s cut through the hype of the “cloud computing” and get straight to what it is in its core.

Don’t Forget to Learn How to Automate All the Things

Video to the Topic

Need Help? Contract Us

If you need a trusted and experienced partner to automate your network and IT infrastructure, get in touch with us.

P.S.

If you have further questions or you need help with your networks, we are happy to assist you, just send us a message. Also don’t forget to share the article on your social media, if you like it.

BR,

Anton Karneliuk

Practical OpenStack #000. Introduction

Hello my friend,

For some time we were thinking whether to step into the cloud world or not. There are so many resources existing already, which provides a decent amount of information. However, closer look revealed the vast majority of such resources are covering either public clouds, such as AWS, GCP, and Azure, or VMware products. All of these areas are important and no doubts have their users. At the same time, OpenStack is at heart of some biggest online platforms and research centres. And it is Open Source. That was a sufficient reason for us to kick off these video series. Let’s get started.

Don’t Forget to Learn How to Automate All the Things

Video to the Topic

Need Help? Contract Us

If you need a trusted and experienced partner to automate your network and IT infrastructure, get in touch with us.

P.S.

If you have further questions or you need help with your networks, we are happy to assist you, just send us a message. Also don’t forget to share the article on your social media, if you like it.

BR,

Anton Karneliuk

Tools 10. Developing Our Own Cross-platform (AMD64/ARM32) Traceroute Prometheus Exporter for Network Monitoring using Python

Hello my friend,

This is the third and the last (at least for the time being ) blogpost about monitoring of the infrastructure with Prometheus, one of the most powerful and popular open source time series database and metrics collection framework. In today’s talk we’ll cover the build of our own Prometheus exporter, which performs trace route checks.

1
2
3
4
5
No part of this blogpost could be reproduced, stored in a 

retrieval system, or transmitted in any form or by any 

means, electronic, mechanical or photocopying, recording, 

or otherwise, for commercial purposes without the 

prior permission of the author.

Why to Automate Monitoring?

Many tools nowadays give you possibility not only to collect metrics, but also to act perform a simple )(or complex) analysis and act based on the result of such an analysis. So can Prometheus. With a help of the Alertmanager, it is possible to send a REST API request upon certain condition, which would trigger an automation activity or a workflow to act upon the business logic needed for the condition, such as remediation and/or configuration. This is why you need to know how the network automation works at a good level.

And we Continue reading

Events. My first MPLS World Congress: Impressions, Feelings, and Raise of AI.

Hello my friend,

This year I had an incredible opportunity to attend an event, which I wanted to attend for quite a bit back in past, when I was working for service providers (e.g., Vodafone, A1 BLR / Telekom Austria Group). The full name of the event is MPLS SDN and AI World Congress 2022, or simply #mplswc22.

1
2
3
4
5
No part of this blogpost could be reproduced, stored in a 

retrieval system, or transmitted in any form or by any 

means, electronic, mechanical or photocopying, recording, 

or otherwise, for commercial purposes without the 

prior permission of the author.

Thanks

I’m very helpful to my friends and colleagues Pau Nadeu Rabat and Jose Manuel Roman Fernandez Checa for inviting me to take part in the event. 

Brief History of the Event

MPLS SDN AI World Congress is one of the biggest events in the networking industry in the Europe. Originally it was called simple MPLS World Congress (hence, the name #mplswc22) and was aiming to bring together the biggest service providers (like the guys I worked for before, Deutsche Telekom, Orange, etc) and vendors of telecom equipment and software (like, Nokia, which I also Continue reading

Automation 13. Real-life Example of a Python Automation for a Network Audit with Nornir and Scrapli

Hello my friend,

Today we are going to discuss a real-life experience, how network automation helped us to save a lot of time and significantly improve a quality of the medium size data centre. You will learn about the problem, which audit was to solve and how Python leveraging Nornir and Scrapli solved it.

1
2
3
4
5
No part of this blogpost could be reproduced, stored in a 

retrieval system, or transmitted in any form or by any 

means, electronic, mechanical or photocopying, recording, 

or otherwise, for commercial purposes without the 

prior permission of the author.

Can Automation Help with Audits?

We, humans, are incredible creatures. We can create. We can write songs and compose music; we can invent new drugs and find new materials. We can develop new software and tools. However, in order to be able to do that, we need to have a free time and not to worry about anything. That’s why we need to rely on different tools, which can do routine tasks requiring a lot of concentration at least as good as we, humans, can. Probably, even better than we. Audit is one of such tasks, and in IT world it definitely Continue reading

Tools 9. Monitoring Availability of Customers via HTTP GET, ICMP, and DNS via Dockerised Prometheus

Hello my friend,

in the previous blogpost we’ve started the discussion how you can improve your customers’ experience in your network via better observability of the network health with Prometheus by means of periodic automated speedtest and iperf3 measurements. Albeit it is very important and useful, it doesn’t tell you if the customers’ or your own resources are available. By resource we mean any exposed service, such as web page, streaming service, etc. Today you will learn how to setup a monitoring with Docekrised Prometheus to make sure that you know for sure if the services are available for customers.

1
2
3
4
5
No part of this blogpost could be reproduced, stored in a 

retrieval system, or transmitted in any form or by any 

means, electronic, mechanical or photocopying, recording, 

or otherwise, for commercial purposes without the 

prior permission of the author.

Why to Automate Monitoring?

Monitoring allows to make you aware, what is the state of resources you are interested in. At a bare minimum, you shall be notified if the state of resources deviates from the acceptable value and, therefore, crosses some threshold. At the same time, this is just the first step. The end goal Continue reading

Tools 8. Monitoring Network Performance with Dockerised Prometheus, Iperf3 and Speedtest

Hello my friend,

in the time when the business is conducted online, it is vital to have a clear visibility into the health of your services and their performance, especially if they rely on the media or other components outside of your immediate control. Earlier in our blogpost we have covered how and why to use iperf3 for measurements of a performance between your hosts and speediest to measure a performance of an Internet connectivity. Today we’ll show how to automate this process with the help of Prometheus.

1
2
3
4
5
No part of this blogpost could be reproduced, stored in a 

retrieval system, or transmitted in any form or by any 

means, electronic, mechanical or photocopying, recording, 

or otherwise, for commercial purposes without the 

prior permission of the author.

How Can We Automate Monitoring?

Automation is not only about Ansible and Python. Knowing how you can properly use various applications, especially those great open source tools available on the market is a key to your success. At the same time, Ansible plays a key role in rolling out application these days, as it helps to ensure that deployment is done in a consistent way. Ansible is like Continue reading