karneliuk.com Archives - Page 2 of 7

Kubernetes 001. Building Cluster on Ubuntu Linux with Docker and Calico in 2022

Hello my friend,

The last Black Friday (in 2021) I’ve purchased some Kubernetes trainings from Linux Foundation with quite a huge discount. I’ve purchased and then has been doing nothing in their regard for quite a bit until I’ve received a notification in the beginning of the September that I have only 3 months to use the training and pass exam or it will expire. Moreover, I have 2 exams to pass, so… All my blogs for a bit will be dedicated to Kubernetes only, as I don’t have time to look into anything else for the time being. I will be using these blogs as a brain dump for me during preparation for CKA (Certified Kubernetes Administrator) and CKAD (Certified Kubernetes Application Developer). Whether I will manage it in 2,5 months, the time will show. Let’s get started.


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 Is Kubernetes Related to Network Automation?

Kubernetes is one of the most popular platforms to build Continue reading

Practical OpenStack #006. What is the step-by-step deployment plan

Hello my friend,

Once we are sorted with the theoretical part of how we are to deploy OpenStack, we are moving to a practical bit. Today we are going to explain the steps we are going to take in order to deploy OpenStack, the supporting storage and how they are to be integrated.

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

Building High-Available Web Services: Open Source Load Balancing Based on HAProxy + FRR and Origin Web Server Based on NGINX Connected to Arista EVPN/VXLAN. Part 2 – Configuration and Validation.

Hello my friend,

In the previous blogpost we have started discussion about the Open Source Load Balancing solution, which leverages HAProxy and FRR, which is connected to the data centre fabric running EVPN/VXLAN on Arista EOS switches and serving content from NGINX-based origin servers. In that blogpost we covered the architectural guidelines and design principles. Today we will cover the configuration and the validation of the solution.


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 More Important: Network Technologies or Network Automation?

The truth is that both disciplines are equally important. Knowledge and skills in network technologies will allow you to build great connectivity solutions to empower businesses across the globe and spread its capabilities as nowadays, perhaps, 99% of all business operations leverage IT and network technologies either entirely or at least partially. In its turn, network automation allows to make operational activities (e.g., configuration, troubleshooting, analysis, etc) much more precise and predictable, decreasing the probability of outages or Continue reading

Building High-Available Web Services: Open Source Load Balancing Based on HAProxy + FRR and Origin Web Server Based on NGINX Connected to Arista EVPN/VXLAN. Part 1.

Hello my friend,

Recently we’ve been working on an interesting (at least for me) project, which is an MVP of the highly available infrastructure for web services. There are multiple approaches existing to create such a solution including “simply” putting everything in Kubernetes. However, in our case we are building a solution for a telco cloud, which is traditionally not the best candidate for a cloud native world. Moreover, putting it to Kubernetes will require to build a Kubernetes cluster first, which is completely separate magnitude of the problem. Originally we were planning to write this blogpost the last weekend, but it took us a little bit longer to put everything together properly. Let’s see, what we are to share with 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.

You Are Not Talking about Automation Today, Aren’t You?

Yes, today’s blogpost is dedicated to the network technologies (to a huge mix of different network and infrastructure technologies, to be honest). That’s why there Continue reading

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?

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,


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.

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

« Previous 1 2 3 4 … 7 Next »