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.
2
3
4
5
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
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.
2
3
4
5
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.
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
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.
2
3
4
5
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.
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
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.
2
3
4
5
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.
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 is an approach, which is very often used in software development, and discussed outside of that area. It stands for:
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.
One of the interesting experiences we’ve obtained in the software development was the development of Prometheus exporter to report trace Continue reading
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.
2
3
4
5
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.
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
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
Video to the Topic
If you need a trusted and experienced partner to automate your network and IT infrastructure, get in touch with us.
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
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.
Video to the Topic
If you need a trusted and experienced partner to automate your network and IT infrastructure, get in touch with us.
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
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.
Video to the Topic
If you need a trusted and experienced partner to automate your network and IT infrastructure, get in touch with us.
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
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.
Video to the Topic
If you need a trusted and experienced partner to automate your network and IT infrastructure, get in touch with us.
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
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.
Video to the Topic
If you need a trusted and experienced partner to automate your network and IT infrastructure, get in touch with us.
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
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.
Video to the Topic
If you need a trusted and experienced partner to automate your network and IT infrastructure, get in touch with us.
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
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.
2
3
4
5
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.
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
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.
2
3
4
5
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.
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.
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
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.
2
3
4
5
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.
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
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.
2
3
4
5
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.
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
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.
2
3
4
5
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.
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
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.
2
3
4
5
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.
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
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.
2
3
4
5
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.
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
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.
2
3
4
5
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.
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