Anton Karneliuk

Author Archives: Anton Karneliuk

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

Automation 12. Automated EVPN Customer Deployment with Ansible, NETCONF, and NetBox. 6WIND version.

Hello my friend,

We’ve been preparing this blogpost for quite a while, but for various reasons it was put on the back burner. Now we finally are bringing this back to light. We’ll go over a practical use case of automation of 6WIND configuration with Ansible and NetBox relying NETCONF.

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.

Do I Need to Automate Everything?

The answer is, as usual: it depends. With our passion to automation, we would say: yes, definitely you should automate everything. But this is possible, only if you have unlimited resources (time, money, people). In reality, all the resources are limited and, moreover, may be even scarce. In such a case you would need to choose, where would you obtain the biggest leverage from automation. For example, some tasks are more frequent or time consuming than others. Clearly they are to be automated.

How to find them? Join our automation training and you will find that out!

We offer the Continue reading

Events. My experience from taking part in NANOG #84 Hackathon

Hello my friend,

collaboration, team work, team spirit… These and other similar words we are using quite often to show the importance of working together to achieve something big. Especially, if you have very limited time. This is a good description of the hackathon – the team effort to achieve something big (i.e., MVP of working product) in a short term (i.e., in 24 hours). Let’s see what’s that all about.

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 Was the Hackathon?

There are many hackathons worldwide. The trend we observe now is that many technical events have hackathon as a warm-up event or a part of the main program. In this case, the event was NANOG 84, which is one of the biggest gatherings for all Service Providers, Telco equipment vendors, and system integrators in North America. It is a massive opportunity to learn some interesting things from some leading companies, like their experience and lessons learned Continue reading

Automaton 11. Data Models for NVIDIA Cumulus 5.0. Collecting Config and Operational Data via REST API with Ansible.

Hello my friend,

Right before the New Year we have posted an article, where we introduced a new approach to manage Cumulus Linux, which became available in the latest release to the date Cumulus Linux 5.0. This approach relies on REST API, which makes the automation of this platform very comfortable. Today we’ll cover how to collect the configuration and ope data with the most popular network and infrastructure automation tools these days: Ansible and Python.

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.

Ansible or Python? Python or Ansible?

To be absolutely frank, there is no “or”. There is only “and” in such a question. Both Ansible and Python play crucial role in network and infrastructure automation. Both of them are widely used in various companies from small enterprise to gigantic tier-1 web companies (e.g., Google, Amazon, etc). We see different usage patterns, such as Ansible + AWX being a main toolkit, or Python + Stackstorm or Apache Airflow Continue reading

Automation 10. Installing NAPALM Community Driver for Nokia SRO OS and comparing NAPALM and OpenConfig/YANG.

Hello my friend,

In the previous blogpost we have compared the usage of NAPALM and OpenConfig YANG modules with NETCONF transport for Cisco IOS XR and Arista EOS. Those two operation systems are so called core operating systems for NAPALM and are included in its standard distribution. Besides them, other three (there are five core network operating systems in NAPALM) are Cisco IOS, Cisco Nexus, and Juniper Junos. You though can use NAPALM with other operating systems, if there are community drivers available. In today’s blogpost we cover installation and usage of NAPALM Community Driver for Nokia SR OS and its comparison against OpenConfig/YANG with NETCONF.

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 It Important to Know Python?

Recently we were engaged with one of our customers to help with OpenStack deployment. The first step for us was to build the lab, which could replicate the desired setup. The installation of OpenStack manually (there are means to automate that as Continue reading

Automation 9. Comparing data collection with NAPALM and OpenConfig/YANG over NETCONF in Python for Cisco IOS XR and Arista EOS

Hello my friend,

We hope you have aa good festive period, celebrated Christmas and New Year with your family and friends, and started the new year. Some time ago we’ve discussed a comparison of NAPALM and OpenConfig at a high level and promised to share some details. Today you will see the details of data collection with NAPALM and NETCONF with OpenConfig YANG modules.

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.

Where to Start with Automation?

Recently we have had an interesting webinar together with Rick Donato, my colleague from a Network Automation Industry. During the webinar we mapped various products existing in the automation (not only network, but in general) space and, you know what, there is indeed a huge amount of products. Therefore, when you are already a savvy in the automation world, you can benefit from them a lot. However, if you are just making first steps, you will be totally lost.

We want you to be a beneficiary Continue reading

1 2 3 6