Hello my friend,
quite often we are getting questions from various folks on LinkedIn or via private consultations about how start with network automation and, more importantly, whether to start with network automation at all. With this blogpost we’ll try to answer the most popular question.
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 simple: don’t wait. The amount of things you need to learn in order to succeed is huge and growing daily with new tools and platforms released. However, the good news is that we we provide you everything you need to start successfully in the role:
We offer the following training programs for you:
During these trainings you will learn the following topics:
Dear friend,
Today we’ll look into a question, which is raising quite often on various meetups related to network automation or various threads that is the addition of any automation components (say, NetBox) to the existing networking managing suit, where you already have some inventory (say, LibreNMS). We’ll take a look into multiple approaches and, as usual, will share some code snippets how you can do that.
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.
Integrating different automation tools is typically the most labor-intensive work, as it requires understanding of the tools you are integrating, their internal hierarchy and APIs capabilities. That is not even saying that you shall be proficient with a variety of technologies, such as REST API or GRPC API, JSON/Protobuf/XML, Postman and many other tools. That’s a lot, and may be not easy to start with.
The good news is that all these things you can learn at out Network Automation Training programs:
Dear friend,
After a bit of break caused by preparation to Kubernetes exams (we will continue blogs about Kubernetes as well) we are getting back to network and network automation topics. One of the interesting things, which is gradually emerging these days, is the possibility to manage multiple aspects of network devices (not only configuration or collection of operational data), such us issuing ping/traceroute checks, copying file, etc in a model-drive way (i.e., NETCONF, RESTCONF, GNMI with YANG). Today we are going to look into such a topic.
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.
It is exactly that. NETCONF/YANG all the things, my friend! Usage of model-driven network automation significantly improves the stability and manageability of the network due to much simpler way to perform all the operations remotely. You don’t need to scrape and parse CLI anymore; instead, you interact with network devices via programmable API, what makes it possible to integrate them Continue reading
Hello my friend,
It took a bit since our previous blogpost about the setup of the highly-available Kubernetes cluster with multiple control plane and worker nodes. We aimed to write the blogpost about the upgrade, but we will park it for now for two reasons:
Therefore, we decided to walk you through the main components used to build and publish your application in a cloud native way on Kubernetes. Let’s dive into that.
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.
These days there are interesting projects emerging with Kubernetes acting as a management plane for network devices. Kubernetes in such projects has a role of the entity, which distributes configuration to the worker nodes, which are either proxies for Continue reading
Hello my friend,
As mentioned in the previous blogpost, we continue looking into Kubernetes. In the previous blogpost we have built a simple cluster consisting of one control plane node, which is the one ruling the cluster, and two worker nodes, which are the ones hosting the customers’ workloads. Today we will add a few more nodes, both workers and control plane, to the cluster to convert it into a high available one.
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 absolutely are. One of the important things we have figured out is that automation is for sure the cross-platform technology. Therefore, whatever area you are coming from or heading two (networking, system administration, or cloud engineering), knowledge of automation frameworks and components, such as Ansible, Bash, Python, YAML, JSON, REST API, GRPC/GNMI is very beneficial and, in fact, is almost mandatory these days.
And in our Network Automation Trainings we have put it together in Continue reading
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.
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.
Kubernetes is one of the most popular platforms to build Continue reading
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.
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 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.
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 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
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.
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, 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
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.
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 Continue reading
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:
Let’s see how we can attempt to normalize collection of such a data with Prometheus from dis-aggregated network switches empowered by Linux.
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 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
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.
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 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
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