<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Robinson uche on Medium]]></title>
        <description><![CDATA[Stories by Robinson uche on Medium]]></description>
        <link>https://medium.com/@Robinson-uche?source=rss-d19a4cd86f3d------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*YwQrLqPepQ9K96RU5Dxd2g.jpeg</url>
            <title>Stories by Robinson uche on Medium</title>
            <link>https://medium.com/@Robinson-uche?source=rss-d19a4cd86f3d------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Fri, 20 Nov 2026 13:39:54 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@Robinson-uche/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Your Data and Attention]]></title>
            <link>https://medium.com/@Robinson-uche/your-data-and-attention-c4fb0cd8fbc4?source=rss-d19a4cd86f3d------2</link>
            <guid isPermaLink="false">https://medium.com/p/c4fb0cd8fbc4</guid>
            <dc:creator><![CDATA[Robinson uche]]></dc:creator>
            <pubDate>Thu, 14 May 2026 18:58:21 GMT</pubDate>
            <atom:updated>2026-05-14T18:58:21.359Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>The economy of you</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*wtJOJhzuC9lSaaI_iOAReA.jpeg" /><figcaption>Photo by <a href="https://unsplash.com/@choys_?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Conny Schneider</a> on <a href="https://unsplash.com/photos/a-blue-background-with-lines-and-dots-xuTJZ7uD7PI?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a></figcaption></figure><p>Money still exists. But it is no longer the only thing you spend.</p><p>In the future taking shape around us, two new currencies run parallel to every dollar transaction: your <strong>data</strong> and your <strong>attention</strong>. You spend them constantly, often without noticing, and unlike money, you cannot check a balance. You do not know how much you have given away, or to whom, or what it was worth.</p><p>This is not science fiction. It is already the architecture of the internet you use today, extended forward into a world with more surfaces, more sensors, more moments of contact between you and a system that is watching.</p><p><strong>Data as currency</strong></p><p>Every search, every purchase, every route you drive, every pause before a decision: these are transactions. The receiver is not a bank. It is a distributed network of platforms, data brokers, advertisers, insurers, and governments, each holding a different piece of a profile that, assembled, knows you better than most people in your life do.</p><p>In the near future, this becomes more explicit. Your health data negotiates your premiums. Your location history qualifies or disqualifies you for credit. Your political leanings, inferred rather than declared, determine what news you see, what prices you are quoted, what opportunities surface. The market for personal data is already enormous. What is coming is its maturation: from murky to systematic, from optional to ambient.</p><p>The troubling thing about data-as-currency is that you cannot spend it once and be done; it compounds. A single data point is worth little, but a decade of behavior, cross-referenced across platforms, builds something much more powerful: a probabilistic model of who you are, what you will do next, and what it will take to influence you.</p><p>You are not the owner of this model. You are its subject.</p><p><strong>Attention as currency</strong></p><p>Attention is scarcer than data, and therefore more valuable.</p><p>There are roughly sixteen waking hours in a day. Every platform, every brand, every political campaign, every piece of content is competing for a share of those hours, not to inform you, but to hold you. The longer you stay, the more data they collect, and the more data they collect, the more precisely they can hold you next time.</p><p>In this economy, boredom is the enemy of profit. Discomfort, mild and chronic, just enough to keep you scrolling, is a feature, not a malfunction. The content that survives is not the content that is most true or most useful; it is the content that is most sticky. Outrage is sticky. Fear is sticky. Desire is sticky. Nuance rarely trends.</p><p>As interfaces move from screens into wearables, into ambient environments, into systems that read your facial expressions and voice tone, the surface area of your attention expands. There is, in principle, no moment left that cannot be monetized, and the gap between a thought and an advertisement narrows.</p><p><strong>What this costs</strong></p><p>Every currency has an inflation problem when it is overextended. The inflation of attention is cognitive: a population chronically fragmented, less able to sustain the long focus that produces good thinking, good work, good relationships. The inflation of data is a gradual loss of the self as a private entity, an erosion of the interior life, the space where you are not a consumer or a demographic, but simply a person.</p><p>Neither of these costs shows up on a receipt.</p><p><strong>The question no one is asking loudly enough</strong></p><p>If your data and your attention are currencies, genuinely valuable, scarce, and consequential, then the question of consent is not a legal formality; it is a political one. Who decides the exchange rate? Who sets the terms? Who protects against exploitation in a market where one side has near-perfect information about the other and the other has almost none?</p><p>The answer, so far, is mostly the platforms themselves.</p><p>That is like asking the bank to write the consumer protection law.</p><p><strong>What remains</strong></p><p>There is a version of this future that is not dystopian. Attention directed well, toward beauty, toward depth, toward other people, is still one of the most powerful things a human being can do. Data, governed with real consent and real accountability, can produce medicine, climate models, and infrastructure that saves lives.</p><p>But neither of those futures arrives by default. They require deciding, collectively and individually, that the economy of you has limits, that some things are not for sale, and that the self is not an asset to be optimized by someone else’s algorithm.</p><p>The currency of the future is you. The only open question is who controls the mint.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c4fb0cd8fbc4" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[DevOps Project: Dockerized Flask Web App with GitHub Actions]]></title>
            <link>https://medium.com/@Robinson-uche/devops-project-dockerized-flask-web-app-with-github-actions-e89ec054c93f?source=rss-d19a4cd86f3d------2</link>
            <guid isPermaLink="false">https://medium.com/p/e89ec054c93f</guid>
            <dc:creator><![CDATA[Robinson uche]]></dc:creator>
            <pubDate>Fri, 28 Mar 2025 17:47:23 GMT</pubDate>
            <atom:updated>2025-03-28T21:55:19.181Z</atom:updated>
            <content:encoded><![CDATA[<p>You just finished learning some of the basic DevOps principles like containerization, CI/CD pipelines, and automation, and you want a simple project to get your hands dirty. Well, you’re in the right place!</p><p>What’s in It for You?</p><p>You’ll create a simple Flask web app with a single route, containerize it using Docker for portability and scalability, and set up a CI/CD pipeline with GitHub Actions to automate the build and push process of the Docker image. Let’s get into it!</p><h4><strong>CREATING YOUR FLASK WEB APP</strong></h4><p><strong>Install Python</strong><br>To install Python, click <a href="https://www.python.org/downloads/">here</a>. To check if it’s installed, run python3 --version</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/822/1*zyHxAdISfeSU73pNf6ZtFA.png" /></figure><p><strong>Create your virtual environment in your working directory and activate it</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*6xuDZGBBq8xzPmNlCAngsA.png" /></figure><p><strong>Install Flask using pip</strong></p><p>pip3 install flask</p><p><strong>Create a file app.py with this content</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*jkV3Uy4SYSBjNVyrOHeycw.png" /></figure><p>This script sets up a <strong>basic Flask web application</strong> that listens for incoming requests and responds with a simple message. It does the following:</p><ul><li><strong>Creates a Flask app instance</strong> to manage the web server.</li><li><strong>Defines a route (</strong><strong>/)</strong> that returns the message &quot;There&#39;s nothing you can&#39;t do!&quot; when accessed in a browser or via an API request.</li><li><strong>Runs the application</strong> on 0.0.0.0:5000 with debugging enabled, making it accessible locally and on a network.</li></ul><p><strong>Create and populate the requirements.txt file<br></strong>This file contains a list of all the apps dependencies</p><p>pip freeze &gt; requirements.txt</p><h4><strong>CONTAINERIZE THE APPLICATION</strong></h4><p><strong>Create your Dockerfile</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1002/1*o7ZZB7zCZ369wP3Rk4zyYg.png" /></figure><p>This <strong>Dockerfile</strong> defines how to containerize the Flask web application. Here’s what each line does:</p><ol><li>FROM python:3.12-slim – Uses a lightweight Python 3.12 image as the base.</li><li>WORKDIR /app– Sets /appas the working directory inside the container.</li><li>COPY requirements.txt . – Copies requirements.txtinto the container.</li><li>RUN pip3 install --no-cache-dir -r requirements.txt– Installs dependencies listed in requirements.txt.</li><li>COPY . . – Copies all application files into the container.</li><li>EXPOSE 5000 – Specifies that the app will listen on port 5000.</li><li>CMD [&quot;python3&quot;, &quot;app.py&quot;] – Defines the command to start the Flask app when the container runs.</li></ol><p><strong>Create a .dockerignore file</strong></p><p>It includes the virtual environment created earlier to exclude it when building the image</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*6H65aZ5YGVI9Rjddz9gLZA.png" /></figure><p><strong>Build and run the Docker container locally</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*n05R-vCeF-pX6qKgnAc7eA.png" /></figure><p>The command docker build -t flask-app .builds a Docker image named flask-app from the Dockerfile in the current directory (.).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*0IOaTE2yyIYuuiV0eBpevA.png" /></figure><p>The command docker run -p 5001:5000 flask-app starts a container from the flask-app image, mapping port <strong>5000</strong> inside the container to <strong>5001</strong> on your machine. This allows access via <a href="http://localhost:5001.">http://localhost:5001.</a></p><h4><strong>SETUP YOUR VERSION CONTROL</strong></h4><p><strong>Initialize a GitHub repository<br></strong>git init</p><p><strong>Create a .gitignore file<br></strong>echo “venv/” &gt;&gt; .gitignore</p><p><strong>Push your code to GitHub<br></strong>git add . <br>git commit -m “Initial commit” <br>git remote add origin <a href="https://github.com/your-username/your-repo.git">https://github.com/your-username/your-repo.git</a> <br>git push -u origin main</p><p><strong>Generate Dockerhub access token</strong></p><ul><li>Log in to Docker Hub.</li><li>Go to <strong>Account Settings</strong> &gt; <strong>Security</strong>.</li><li>Under <strong>Personal Access Tokens</strong>, click <strong>Generate Token</strong>.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*50bz8S4BHD8pH-PKIYAzhw.png" /></figure><p><strong>Create and store DOCKERHUB_USERNAME &amp; DOCKERHUB_TOKEN in github repo</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*jfwb1N_ZQ6c4XmJ33uHy1g.png" /></figure><p>Here’s what it should look like when you’re done</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*FoX7m6N0fu-X-dQOyNKzgQ.png" /></figure><h4><strong>IMPLEMENT CI/CD WITH GITHUB ACTIONS</strong></h4><p>In your working directory, create a .github/workflows dir. This is where you’d store your workflow YAML files. <br>Create a file called build-and-push.yml</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qz9RtSQjG-Vd7PNXdyXVpg.png" /></figure><p>This GitHub Actions workflow automates building and pushing a Docker image to Docker Hub whenever changes are pushed to the <strong>main</strong> branch.</p><p><strong>How It Works</strong></p><ol><li><strong>Triggers on Push:</strong> The workflow runs whenever code is pushed to the <strong>main</strong> branch.</li><li><strong>Runs on Ubuntu:</strong> It executes on the latest Ubuntu environment.</li><li><strong>Checks Out the Repo:</strong> The first step pulls the latest code from the repository.</li><li><strong>Logs into Docker Hub:</strong> It uses stored secrets (DOCKERHUB_USERNAME and DOCKERHUB_TOKEN) to authenticate securely.</li><li><strong>Builds and Pushes the Image:</strong> The final step builds the Docker image and pushes it to Docker Hub, tagging it as latest.</li></ol><p>To try it out, make a code change like adding a README file, commit changes, and push to your GitHub repo then navigate to the actions tab and watch the workflow run.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*y6VCs5D7SJArsO3oxTj40w.png" /></figure><p>Voila!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*IwBbvzCij15FnDD7T5EIvg.png" /></figure><p>You’ve successfully built a fully automated DevOps pipeline for a Flask web application! 🎉</p><p>In this project, you learned how to:<br> ✅ Containerize an application using Docker<br> ✅ Set up version control with GitHub<br> ✅ Automate builds and deployments with GitHub Actions</p><p>With this setup, any code changes automatically trigger a new Docker image build and push to Docker Hub, streamlining the deployment process. You can now expand on this by integrating Kubernetes for orchestration or setting up a cloud-based deployment.</p><p>Keep building, keep automating, this is just the beginning! 🚀</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e89ec054c93f" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Time We Nearly Deleted Our Entire Database]]></title>
            <link>https://medium.com/@Robinson-uche/the-time-we-nearly-deleted-our-entire-database-d8251ca0dddd?source=rss-d19a4cd86f3d------2</link>
            <guid isPermaLink="false">https://medium.com/p/d8251ca0dddd</guid>
            <dc:creator><![CDATA[Robinson uche]]></dc:creator>
            <pubDate>Wed, 12 Feb 2025 10:44:21 GMT</pubDate>
            <atom:updated>2025-02-12T10:44:21.238Z</atom:updated>
            <content:encoded><![CDATA[<h4>The horror of an unintended DROP TABLE in production.</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*EI_ooVlLX1OOrjcSYqf-Bg.jpeg" /><figcaption>Photo by <a href="https://unsplash.com/@targetfotografi?utm_content=creditCopyText&amp;utm_medium=referral&amp;utm_source=unsplash">Stephen Olatunde</a> on <a href="https://unsplash.com/photos/city-skyline-at-daytime-EBcKOqWzc9M?utm_content=creditCopyText&amp;utm_medium=referral&amp;utm_source=unsplash">Unsplash</a></figcaption></figure><p>“Omooooo!” Nnamdi’s scream echoed through the quiet office in Victoria Island. It was just past 4 PM on a Friday, that delicate time when Lagos traffic was about to turn from bad to impossible. But traffic was the least of their concerns now.</p><p>Nnamdi, the backend engineer at a Nigerian startup, had just watched in horror as his terminal executed a DROP TABLE command on their production database. The command was supposed to run on his local development environment, but he had accidentally left his terminal connected to production after fixing an urgent customer issue earlier.</p><p>“Chai! I’m finished!” he muttered, frantically typing commands to assess the damage. The database containing three years of payment transaction records for over 200,000 merchants was gone. Just like that.</p><p>His fingers trembling, he dialed Aisha, their DevOps lead who had stepped out to buy shawarma from the vendor down the street.</p><p>“Aisha, abeg come back quick-quick! There’s fire on the mountain”</p><p>“Kilode? Wetin happen?” Aisha could hear the panic in his voice.</p><p>“I… I just drop the transactions table for prod,” Nnamdi stammered.</p><p>The shawarma fell from Aisha’s hand. “You do wetin?! how?”</p><p>Back at the office, the situation room was chaos. Their CTO, Dr. Okonk, who everyone called “Prof,” had already been notified. He stood at the whiteboard, his usual calm demeanor masking the gravity of the situation.</p><p>“Okay, make we no panic,” Prof said, though sweat was forming on his forehead. “Aisha, when last did we run backup?”</p><p>Aisha was already at her laptop, fingers flying across the keyboard. Their backup system was supposed to run automatically every six hours. But when she checked the logs, her heart sank.</p><p>“Omo! The backup service has been failing silently for the past week. The disk space don full since and nobody noticed the alert emails wey dey go to spam folder.”</p><p>The room went quiet. Even the usual honking from Lagos traffic below seemed muted.</p><p>Then Prof spoke: “This is a teaching moment but rather than point fingers, let’s solve this issue first.”</p><p>They sprang into action:</p><ol><li>Aisha discovered they still had the binary logs from the database running</li><li>They quickly set up a new database instance</li><li>Using point-in-time recovery, they began restoring from their last good backup</li><li>They replayed the binary logs to catch up to the moment before the DROP</li><li>Nnamdi documented every step while Aisha worked her magic</li></ol><blockquote><strong>Binary logs</strong> <strong>(binlogs)</strong> are a type of database log that records all changes to a database in a binary format.</blockquote><blockquote><strong>Point-in-Time Recovery (PITR)</strong> is a database recovery technique that allows you to restore a database to a specific moment in the past, rather than just the last backup.</blockquote><p>Three hours later, fueled by the smokey jollof rice ordered from a nearby restaurant and countless bottles of malt, they had recovered 99.9% of the data. One small batch of transactions from the last 30 minutes before the incident was initially missing, but they found it in their message queue’s dead letter storage.</p><blockquote>A <strong>Dead Letter Queue (DLQ)</strong> is a feature in message queues that stores messages that could not be processed successfully after multiple attempts.</blockquote><p>The next Monday, Prof called an all-hands meeting. The conference room was packed as he outlined their new database management policies:</p><p>“First thing first, we need proper backup monitoring. No more silent failures. Secondly, we need to enforce strict access controls — ensuring that access to the production database is highly restricted and only authorized personnel can make changes. And better separation between production and development environments is obviously needed.”</p><p>They implemented:</p><ul><li>Automated backup verification with alerts to multiple channels</li><li>Database access through a bastion host only</li><li>Regular disaster recovery drills</li><li>Separate color schemes for production and development environment terminals</li><li>A new backup strategy with both hot and cold storage</li></ul><p>Aisha also added: “And my people, we go need proper documentation. No be only code we go backup, but the knowledge of how to recover when there’s trouble.”</p><p>Months later, at a tech event organized by <a href="http://pipeops.io/">Pipeops</a>, Prof shared their story. “You see,” he concluded, “when it comes to tech, backups are your safety net. As the saying goes, ‘Failing to prepare is preparing to fail.’ Always be ready.”</p><p>The incident changed the startup’s entire approach to database management. They made it a priority to monitor backups regularly, document recovery procedures, and conduct routine failure simulations to stay prepared for any future crisis.</p><p>Thanks for reading, I hope you enjoyed the read. Feel free to leave some claps ❤️</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d8251ca0dddd" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Deploy or Die Trying]]></title>
            <link>https://medium.com/@Robinson-uche/deploy-or-die-trying-b7f243eab484?source=rss-d19a4cd86f3d------2</link>
            <guid isPermaLink="false">https://medium.com/p/b7f243eab484</guid>
            <dc:creator><![CDATA[Robinson uche]]></dc:creator>
            <pubDate>Tue, 11 Feb 2025 12:32:31 GMT</pubDate>
            <atom:updated>2025-02-11T12:47:11.869Z</atom:updated>
            <content:encoded><![CDATA[<h4>When Everything Breaks the Night Before Launch</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ONc8tfb2xrOURSWK9TCrBg.jpeg" /><figcaption>Photo by <a href="https://unsplash.com/@davfts?utm_content=creditCopyText&amp;utm_medium=referral&amp;utm_source=unsplash">David Pupăză</a> on <a href="https://unsplash.com/photos/text-heNwUmEtZzo?utm_content=creditCopyText&amp;utm_medium=referral&amp;utm_source=unsplash">Unsplash</a></figcaption></figure><p>It was 2 AM in Lagos, and Chidi was wide awake, staring at his laptop. The glow of the screen reflected his frustration. His startup, QuickKudi, was supposed to launch its new feature, instant bill payments, by morning. But there was one problem: the deployment was failing.</p><p>His team had spent weeks coding, testing, and debugging. Everything worked locally, but once they tried deploying to AWS? Disaster. The pipeline kept breaking, dependencies were missing, and rollback scripts had turned into a war zone.</p><p>“Omo, if we don’t deploy this thing before morning, na our investor go deploy us commot from this startup,” his co-founder, Tolu, muttered.</p><p>Chidi sighed. He had spent the last two years deep in DevOps, but even he was beginning to hate YAML files. He scrolled through Twitter for a break and saw a tweet from a developer:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/462/1*0xOVDgEQW6aVIBLsqApe3g.png" /></figure><p>💡 “Tired of deployment wahala? Try <a href="https://www.linkedin.com/company/pipeops/"><strong>PipeOps (Accelerate Africa ‘24)</strong></a>! No-code cloud deployments in minutes. Let the<strong> </strong><a href="http://pipeops.io/"><strong>Pipeops.io</strong></a><strong> </strong>do the heavy lifting!”</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*JQkzrdOL9g5LQn_vE76ZQg.png" /></figure><p>PipeOps?</p><p>He clicked the link. Within minutes, he had signed up and connected his AWS account. He spun up a server, linked their GitHub repo, and with a deep breath, hit deploy. No need for complex YAML, no broken pipelines, just a clean UI guiding him through the setup.</p><p>“Make I test am sha,” he whispered.</p><p>15 minutes later…</p><p>“Deployment successful.”</p><p>Chidi blinked. He refreshed the dashboard. The new feature was LIVE. No server downtime, no manual interventions, just smooth deployment.</p><p>He turned to Tolu with a grin. “Guy, we don deploy. Make investor catch us na, we go tell am say PipeOps run am for us.”</p><p>Tolu’s eyes widened. “Which PipeOps?”</p><p>“The one that just saved our startup.”</p><blockquote><strong>Key take away</strong><br>If your startup is battling deployment stress, maybe it’s time to ditch the manual processes and let <a href="https://www.linkedin.com/company/pipeops/"><strong>PipeOps (Accelerate Africa </strong></a><strong>’24)</strong> handle the heavy lifting. No stress. Just vibes. 🚀</blockquote><p>Thanks for reading, I hope you enjoyed the read. Feel free to leave some claps ❤️</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b7f243eab484" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Chaos of Container Orchestration]]></title>
            <link>https://medium.com/@Robinson-uche/the-chaos-of-container-orchestration-8ca5ac89ecc5?source=rss-d19a4cd86f3d------2</link>
            <guid isPermaLink="false">https://medium.com/p/8ca5ac89ecc5</guid>
            <dc:creator><![CDATA[Robinson uche]]></dc:creator>
            <pubDate>Mon, 10 Feb 2025 21:55:26 GMT</pubDate>
            <atom:updated>2025-02-11T12:53:44.624Z</atom:updated>
            <content:encoded><![CDATA[<h4>It was supposed to be a routine deployment.</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*HHBT90TklIFndv3dnOiteg.png" /><figcaption>Image from <a href="https://www.virtasant.com/blog/k8-series-introduction-to-kubernetes">here</a></figcaption></figure><p>At Japa4ever Nigeria, a rapidly growing (you can guess why) tech startup based in Lagos, the DevOps team had been working tirelessly to move their monolithic app to a containerized microservices architecture. With Kubernetes orchestrating their containers, they felt confident that this was the right direction. No more manual scaling. No more deployment nightmares. Everything had been running smoothly in the staging environment for weeks, and today was the day they’d roll it out to production.</p><p>Sade, the lead DevOps engineer, sat at her desk sipping Zobo from her favorite mug as she watched the deployment logs scroll by. “This is it. We’re finally here,” she thought. The pipeline had triggered automatically, deploying the latest version of the app across their Kubernetes clusters. Pods were coming online, services were being registered with the load balancers, and the health checks were green.</p><p>It felt almost too easy.</p><p>But as the last few pods spun up, things started to go sideways.</p><p>A few minutes after the deployment was completed, users began reporting issues. At first, it was sporadic, with just one or two complaints about a feature being down. Then, it escalated.</p><p>“Sade, this app don crash!” shouted Chuka, the senior developer, from across the room. “The user dashboard no gree load!”</p><p>Sade frowned, quickly opening the Kubernetes dashboard on her laptop. Pods for the frontend service were all green, but the logs told a different story. Error messages piled up faster than she could read them. The service was crashing repeatedly, but only intermittently. Some requests would work, while others would time out.</p><p>“Wetin dey happen?” she muttered, scrolling through the logs. Everything looked fine with the containers themselves. The resource limits were within spec. There was no sign of a memory leak or any overutilization.</p><p>She quickly fired up the metrics dashboard. There, her heart dropped: the service was being throttled by Kubernetes’ Horizontal Pod Autoscaler. The application was receiving bursts of traffic, but it wasn’t scaling fast enough to handle the load.</p><p>“Okay, so we need to adjust the autoscaler,” Sade said, her voice steady, but her mind racing. She went to modify the settings for the deployment, increasing the replica count. But then something even stranger happened: the newly deployed pods didn’t come up. Kubernetes couldn’t schedule them onto the nodes.</p><p>“Sade, are we sure the node pool is up to spec?” Chuka asked, walking over to her.</p><p>“I thought so,” Sade said, switching tabs to the cloud provider’s dashboard. “We’ve got enough nodes, we’ve got capacity — ”</p><p>And that’s when she saw it.</p><p>The node pool was maxed out. Despite scaling rules being set for auto-scaling, the cloud provider had hit a limit on available instances due to unexpected high demand across the region. The node pool couldn’t handle the additional pods.</p><p>“Ah, Omo, we don enter soup now,” she muttered. “Not only is the app struggling with scaling, but now we’ve hit a cloud provider limit.”</p><p>Chuka leaned in, offering a solution. “We could try a rolling restart on the app. Maybe that would flush out any underlying issues.”</p><p>Sade frowned. The problem wasn’t just a deployment glitch, it was a lack of resources. But she had to keep things moving. If they couldn’t scale up, the service would continue to degrade for users.</p><p>With a quick decision, Sade initiated a manual scaling process, increasing the node pool limits. At the same time, she manually triggered a rolling restart of the affected service pods. As the pods came online one by one, the application started to stabilize. The rolling restart ensured that only a few pods were restarted at a time, preventing a complete outage.</p><p>“Okay, we’re getting somewhere,” Sade said, breathing a little easier. But they weren’t out of the woods yet.</p><p>Sade and Chuka dove into the Kubernetes events, quickly tracing the issue to a combination of factors: traffic spikes from an unexpected marketing campaign, a misconfigured autoscaler, and an oversubscription of cloud resources. Kubernetes had done its best to handle it all, but the perfect storm of conditions had led to chaos.</p><p>The team scrambled. They adjusted the Kubernetes scaling policies, ensuring that pods would be spread more evenly across nodes. They also tweaked the load balancing configuration to better distribute the requests. Finally, they reached out to the cloud provider support to request an increase in their resource limits.</p><p>As the day went on, things began to improve. The application stabilized, and the user complaints slowed to a trickle. But Sade knew this wasn’t over. She could hear the echoes of the chaos in her mind.</p><p>They’d learned valuable lessons in the heat of battle: The importance of monitoring, the need for better load forecasting, and the inevitability of cloud provider limitations. But most importantly, they had learned that with Kubernetes, even the best-laid plans can go awry.</p><p>Sade sat back in her chair, staring at the dashboard as everything slowly returned to normal. “We’ve survived this one,” she said to Chuka. “But next time, we’re going to be ready for this chaos.”</p><p>Chuka nodded. “Let’s make sure our auto-scaling policy doesn’t crash and burn again.”</p><p>Sade couldn’t help but smile. “Agreed. Abeg, make I order Amala and Gbegiri make I relax. It’s almost time for break”</p><p>As the team celebrated their hard-earned victory, they knew they’d face even more challenges ahead. But for now, they’d lived through the chaos of container orchestration, and they were a little bit wiser because of it.</p><h3><strong>Key Takeaways</strong></h3><ol><li><strong>Infrastructure Limits Matter</strong><br>Kubernetes can automate scaling, but your cloud provider has hard limits. If you hit them, autoscaling won’t save you. Monitor resource caps and plan ahead.</li><li><strong>Autoscaling Needs Fine-Tuning</strong><br>The Horizontal Pod Autoscaler (HPA) doesn’t always react fast enough to traffic spikes. Poor configurations can make scaling inefficient. Tweak your policies to handle sudden surges.</li><li><strong>Expect the Unexpected with Traffic Spikes</strong><br>External events, like a marketing campaign , can flood your system with requests. Load forecasting and real-time alerting help you stay ahead of surprises.</li><li><strong>Rolling Restarts Help, But Aren’t a Fix</strong><br>Restarting pods can stabilize performance temporarily, but it won’t solve deeper resource issues. Use it as a quick patch, not a permanent solution.</li><li><strong>Comprehensive Monitoring is Essential</strong><br>Without real-time metrics on resource usage and application health, troubleshooting becomes guesswork. Invest in robust monitoring and logging.</li><li><strong>Cloud Resources Don’t Scale Instantly</strong><br>Auto-scaling policies won’t help if your cloud provider can’t allocate more resources. Ensure your provider limits won’t bottleneck your system.</li><li><strong>Even the Best Systems Can Fail</strong><br>Kubernetes, well-tuned scaling, and redundancy won’t prevent every failure. Prepare for worst-case scenarios and build resilience into your infrastructure.</li><li><strong>Post-Incident Reviews Drive Growth</strong><br>Every outage is a learning opportunity. Analyzing what went wrong (and why) helps you refine your system and prevent future chaos.</li></ol><h3>Important Terms</h3><p><strong>Kubernetes: </strong>An open-source container orchestration platform that automates the deployment, scaling, and management of containerized applications.</p><p><strong>Pod: </strong>The smallest and simplest Kubernetes object, consisting of one or more containers that share storage, networking, and configuration.</p><p><strong>Horizontal Pod Autoscaler (HPA):</strong> A Kubernetes component that automatically adjusts the number of pod replicas based on observed metrics, such as CPU or memory usage.</p><p><strong>Node Pool: </strong>A set of nodes (virtual machines) in a Kubernetes cluster that are used to run containers. A node pool typically contains identical nodes in terms of size and configuration.</p><p><strong>Rolling Restart: </strong>A deployment strategy in which Kubernetes restarts pods one by one, ensuring that the application remains available without downtime during the process.</p><p><strong>Cloud Provider Limit: </strong>The maximum capacity (e.g., number of virtual machines, amount of storage) that a cloud provider allows a user to provision within a region or account, often limiting scaling operations when reached.</p><p><strong>Auto-Scaling: </strong>The process of automatically adjusting the number of instances (pods or nodes) in a Kubernetes cluster based on traffic demand, helping to balance load without manual intervention.</p><p><strong>Load Balancer:</strong> A system that distributes incoming network traffic across multiple instances (pods or nodes) to ensure no single instance is overwhelmed, helping to improve availability and reliability.</p><p>In DevOps, chaos is inevitable, but with the right strategies, you can survive it. <br>Thanks for reading, I hope you enjoyed the read. Feel free to leave some claps ❤️</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8ca5ac89ecc5" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[A short piece on AWS IAM roles.]]></title>
            <link>https://medium.com/@Robinson-uche/a-short-piece-on-aws-iam-roles-22b8db4d4120?source=rss-d19a4cd86f3d------2</link>
            <guid isPermaLink="false">https://medium.com/p/22b8db4d4120</guid>
            <dc:creator><![CDATA[Robinson uche]]></dc:creator>
            <pubDate>Tue, 14 Jan 2025 13:19:02 GMT</pubDate>
            <atom:updated>2025-01-14T13:19:02.299Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/241/1*42d6BSGA__A_MHsJWV3zYA.png" /></figure><p>An IAM role is an AWS identity that defines what actions can be taken on AWS resources, but it differs from an IAM user. While an IAM user is associated with a single person or service, an IAM role is intended to be assumed temporarily by principals (users, applications, or AWS services).</p><p>IAM roles are useful when dealing with an unknown or large number of principals. For example, if a mobile application has millions of users, using IAM roles makes sense because AWS imposes a hard limit of 5,000 IAM users per account. By using IAM roles, you can grant temporary access to resources without exceeding this limit or hardcoding credentials.</p><p>IAM roles are designed for temporary access. They are “assumed,” just like putting on a hat. When you assume a role, you gain the specific permissions defined in the role’s policies. Importantly, an IAM role does not represent a particular user; instead, it represents a set of permissions or a level of access in an AWS account.</p><p><strong>IAM roles have two main types of policies:</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/700/1*2YNyeji7gMabPT2k77sKiA.png" /></figure><p><strong>Trust policy: </strong>This contains the identities allowed to assume that role. It references identities in the same account and other AWS accounts too.</p><p><strong>Permission policy: </strong>These define what actions and resources the role gives access to when it’s assumed.</p><p><strong>What happens when an identity assumes a role?</strong><br>When a principal (a user, app, or service) listed in the trust policy assumes a role, AWS generates temporary security credentials through its <strong>Security Token Service (STS)</strong>. These credentials include:<br>- An access key ID<br>- A secret access key<br>- A session token</p><p>The best part? These credentials expire after a short time. This makes roles safer than creating IAM users with long-term credentials because it eliminates the need to store access keys in your code (no more worrying about accidentally pushing keys to GitHub!) and every time a role is assumed, new credentials are generated, improving security.</p><p><strong>I hope you learned a thing or two. Thanks for reading! </strong>😊</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=22b8db4d4120" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Understanding high availability, fault tolerance, and disaster recovery.]]></title>
            <link>https://medium.com/@Robinson-uche/understanding-high-availability-fault-tolerance-and-disaster-recovery-f55e51291df0?source=rss-d19a4cd86f3d------2</link>
            <guid isPermaLink="false">https://medium.com/p/f55e51291df0</guid>
            <dc:creator><![CDATA[Robinson uche]]></dc:creator>
            <pubDate>Mon, 06 Jan 2025 23:23:01 GMT</pubDate>
            <atom:updated>2025-01-06T23:23:01.200Z</atom:updated>
            <content:encoded><![CDATA[<p>Think about the last time your favorite app crashed, or your online banking went down right when you needed to make a payment. Frustrating, right? In our increasingly connected world, we expect our digital services to work flawlessly 24/7. But behind the scenes, keeping these systems running smoothly is like juggling while walking a tightrope — it takes careful planning, the right tools, and sometimes a bit of redundancy.</p><p>That’s where high availability, fault tolerance, and disaster recovery come in. While these terms might sound like technical jargon, they’re actually different strategies for answering one simple question: “What happens when things go wrong?” And trust me, things do go wrong — from simple server crashes to natural disasters that can take down entire data centers.</p><p>Whether you’re running a small business website or managing critical hospital systems, understanding these concepts isn’t just about preventing headaches — it’s about keeping your digital world spinning even when the unexpected happens. Let’s break down what each of these terms really means and why they matter in today’s always-on world.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ygzqHYO7fKj6iqrNapLaRQ.jpeg" /><figcaption>Image from <a href="https://mycloudwiki.com/cloud/ha-vs-ft-vs-dr/">mycloudwiki</a></figcaption></figure><p>A <strong>highly available (HA)</strong> system is one designed to maximize uptime. It’s designed in such a way that when there is a fault in one or more components, it can be fixed or replaced as soon as possible, which often involves the use of automation. This, in some cases, requires having a standby server or component that could easily replace the faulty one. With very little downtime.</p><p>A <strong>fault-tolerant (FT) </strong>system, on the other hand, is the property of a system that keeps operating even during a fault or failure of one or more components. This is more expensive than a system designed to be highly available only. This can be implemented in industries with high accuracy and no tolerance for a discrepancy. Think about some systems in hospitals where a little error could be catastrophic. Hence this requires a good level of redundancy and certain other things, although expensive and complex to implement. One should know when a system needs this or just needs to be designed as a highly available system.</p><p><strong>Disaster recovery (DR) </strong>talks about the set of policies, tools, and procedures that are in place to enable the continuation or recovery of vital technology infrastructure and systems following a natural or human-induced disaster. It’s not just about what happens when the disaster occurs, but what happens before it occurs.</p><p>That’s why regular backups are emphasized. Backups that shouldn’t be stored in the same location. Well-documented plans for the recovery process, which also include copies of processes, login passwords, and other important details should be made. DR testing should be done frequently too. The main aim of DR is to keep the crucial and non-replaceable parts of your systems safe. “</p><p>In summary</p><p>- High availability minimizes any downtimes.</p><p>- A fault-tolerant system operates through faults.</p><p>- Disaster recovery is used when HA and FT don’t work in the event of a disaster.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f55e51291df0" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Data Science, ML & AI: Where Math Meets Tech]]></title>
            <link>https://medium.com/@Robinson-uche/data-science-ml-ai-where-math-meets-tech-41e9d8f2042f?source=rss-d19a4cd86f3d------2</link>
            <guid isPermaLink="false">https://medium.com/p/41e9d8f2042f</guid>
            <dc:creator><![CDATA[Robinson uche]]></dc:creator>
            <pubDate>Sun, 01 Sep 2024 23:05:09 GMT</pubDate>
            <atom:updated>2024-09-01T23:05:09.438Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="A robot standing and thinking in front of a blackboard that has calculations all over it" src="https://cdn-images-1.medium.com/max/1000/0*wXBnojZOYRsev3JQ.jpg" /></figure><p>These days, especially on LinkedIn and some other platforms, I’ve come across many bios of individuals with the word “enthusiast” in them.</p><p>Are we really enthusiastic about these things we say we are, or do we just claim to be while trying to manifest it into reality? After all, it’s a career we’ve pursued for many years, or at least it’s a career path we’re trying to pave for ourselves. The least we could do is be enthusiastic about it.</p><p>I never really felt like I’d met an enthusiastic person until I became friends with an AI developer who specializes in Natural Language Processing and some other fancy words. Just after the first few conversations about data science, ML &amp; AI, I fell in love — with those topics, of course. You can’t blame me because why does it seem equivalent to having a superpower? Being able to create intelligent systems that can analyze vast amounts of data, learn patterns, make predictions, automate complex tasks, and solve problems across diverse fields. This blows my mind steadily.</p><p>Knowing how much this involves mathematics, I got even more excited. Finally, an area in tech where my love-hate relationship with mathematics can be applied. I decided to fuel this enthusiasm that developed with every conversation I had with this friend. Casual conversations always led to us talking about this absolutely cool thing he built or this model that can solve this or that, or some of the many different problems that could be solved with AI.</p><p>Now, when asked, “What superpower would you like to have?”, I have a definitive answer: to be an AI developer.</p><h4>The Journey</h4><p>Luckily, this isn’t unattainable like being invincible or having the power to teleport. With numerous courses and YouTube videos on these topics, I’ve chosen my guide for this journey, a course on Udemy titled:</p><blockquote>Machine Learning A-Z: AI, Python &amp; R + ChatGPT Prize [2024] created by Kirill Eremenko, Hadelin de Ponteves, SuperDataScience Team, Ligency Team.</blockquote><h4>What I’ve learned so far</h4><p>To make this short, I’d like to share the things I learned so you can learn a few things about this exciting topic too.</p><p>First, we know one of the most consistent words used in this field is “Data”, because it’s the foundation of AI. To use this data, it needs to undergo some pre-processing for it to be used to build and train a model that would, in turn, make predictions.</p><p><strong>Data pre-processing</strong> steps involve importing the data, cleaning the data, and splitting the data into the training and testing sets.</p><p>One fascinating aspect is the data splitting process. Typically, you divide your dataset into two parts: 80% for training and 20% for testing. The model learns from the training set, then you evaluate its performance using the test set. By comparing the model’s predictions on the test data with the known actual values, you can assess the model’s accuracy and efficiency. This helps ensure the model can generalize well to new, unseen data.</p><p>Then I got into<strong> feature scaling</strong>: Imagine you’ve got a table with the salaries and ages of three different people, and you’re trying to figure out who the person in the middle is most similar to. Here’s the thing — you can’t just compare the numbers in the salary column or the age column directly because they don’t represent the same thing, so the numbers have different weights. That’s where feature scaling comes into play. It transforms every feature into a similar scale so they can be compared fairly. There are two ways to do this: the standardization method or the normalization method.</p><blockquote><strong>The standardization method </strong>is like telling everyone to stand in line and measuring how far they are from the average height. It makes the average of your data 0 and spreads everything out based on how different it is from that average. The equation looks like this: (x — mean) / standard deviation.</blockquote><blockquote><strong>The normalization method</strong>, on the other hand, is like squishing all your data between 0 and 1. It’s great when you want everything neatly packed in a specific range. The formula for this one is: (x — min) / (max — min).</blockquote><p>Both of these methods help so you can compare apples to apples (or in this case, salaries to ages).</p><p>Just as Peter Parker was bitten by a radioactive spider and gained his superpowers, my chance encounter with an AI developer marks the beginning of my journey into the world of unlimited possibilities through AI, I could call that my personal “spider-bite moment”. I hope you enjoyed this read and it inspires you too.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=41e9d8f2042f" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Why Choose Linux?]]></title>
            <link>https://medium.com/@Robinson-uche/linux-for-beginners-d34f2052d8c6?source=rss-d19a4cd86f3d------2</link>
            <guid isPermaLink="false">https://medium.com/p/d34f2052d8c6</guid>
            <category><![CDATA[linux-for-beginners]]></category>
            <dc:creator><![CDATA[Robinson uche]]></dc:creator>
            <pubDate>Sat, 29 Oct 2022 15:42:12 GMT</pubDate>
            <atom:updated>2023-04-12T22:16:01.662Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*yFw3ypLQIrbflmy5nGj4GQ.jpeg" /><figcaption>Image from <a href="https://blog.knoldus.com/creating-custom-daemons-in-linux/">Knoldus Blogs</a></figcaption></figure><p>If you’re only familiar with the Windows operating system, it’s time to be introduced to something more interesting:<strong> Linux.</strong></p><h4>A brief history</h4><p>Linus Torvalds, a student at the University of Helsinki, started a personal project to create an open-source operating system similar to MINIX, a UNIX operating system in the early 1990s.</p><blockquote>Open source: a software with code that can be inspected, modified, and enhanced by anyone, which is in most cases, free.</blockquote><p>The Linux kernel and GNU utilities equal a complete, open-source UNIX-like operating system.</p><h4><strong>Some Linux principles;</strong></h4><ul><li><strong>Everything is a file:</strong> In Linux, files, directories, and external devices like your mouse, USB, etc. are all handled as files since doing so makes writing programs easier.</li><li><strong>Small and simple programs:</strong> There are rarely bulky programs in Linux that’s why there are so many dependencies because one program does one job and does it well.</li><li><strong>Configuration data is stored in text files: </strong>Such files are usually with a .conf extension and text editors like VI, VIM, NANO, etc are used when editing the files.</li><li><strong>Chaining of programs: </strong>Programs can be chained together to perform complex operations.</li></ul><h4>Why choose Linux?</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3ybqqoBFDvp2RzLH5HlBrQ.jpeg" /><figcaption>Image from <a href="https://www.pexels.com/photo/female-software-engineer-coding-on-computer-3861972/">Pexels.com</a></figcaption></figure><blockquote>All the best people in life seem to like LINUX.</blockquote><blockquote>-Steve Wozniak</blockquote><p>These days, Linux is used in many devices, from smart TVs to supercomputers, android phones, and even most websites you visit. It also offers a wide range of other uses.</p><p>Here are a few reasons why you should choose Linux:</p><p><strong>Linux is open source</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/1*9epWMh1H89oy5j1Lvp-mKA.jpeg" /><figcaption>Image from <a href="https://www.xlike.org/what-is-open-source-strengths-and-examples-of-software/">Xlike.org</a></figcaption></figure><p>Because of the way the operating system was designed, the source code can be inspected, modified, and enhanced; it is also available for users to download and use for free. By being open source since its inception, Linux has a community of developers and individuals who have the hearts and brains to advance the technology.</p><p><strong>Linux supports a wide variety of hardware</strong></p><p>This comprises a variety of hardware such as routers, automation controls, smart home appliances, video gaming consoles, televisions, cars, and spacecraft.</p><p><strong>Most servers run on Linux</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*wF3sLseYQCa9CvC4N6BvEg.png" /><figcaption>Image from Plesk.com</figcaption></figure><p>Due to its stability, security, and flexibility, Linux servers are among the most commonly utilized in today&#39;s world.</p><p><strong>Security</strong></p><p>For developers, Linux is the best solution because of its security features. It isn’t fully secure, but it is less exposed than in other places. The admin user must approve each application. A virus is only activated when the admin user types in the password. So no antivirus software is needed on Linux systems.</p><h4>Architecture of Linux</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/560/1*sXE3DwgMHnscxIix6E3YZw.jpeg" /><figcaption>Image from <a href="https://www.tutorialspoint.com/operating_system/os_linux.htm">Tutorialspoint.com</a></figcaption></figure><p>The Linux operating system’s architecture mainly contains some of these components:</p><p><strong>Hardware: </strong>This consists of several peripheral devices such as the RAM, CPU, and HDD.</p><p><strong>Kernel: </strong>The Linux kernel reads and understands the hardware while passing a signal from the hardware to the shell.</p><p><strong>Utilities</strong>: Lots of tools are built around the Linux kernel.</p><h4>Linux distros</h4><p>A Linux distro is an operating system made from a software collection that includes the Linux kernel and, often, a package management system. Linux users usually obtain their operating system by downloading one of the Linux distributions, which are available for a wide variety of systems. You can download any of these distros to get started with Linux on your computer.</p><p>Some popular Linux distros;</p><ul><li>Ubuntu Linux</li><li>Linux mint</li><li>Arch Linux</li><li>Fedora Linux</li><li>Debian</li><li>OpenSuse</li></ul><h4>Most used Linux distros according to their packaging methods</h4><p><strong>RedHat package manager (RPM) based</strong>: This is considered a stable and secure operating system. Examples include; RHEL, Centos, Oracle Linux</p><p><strong>Debian-based:</strong> This is easy to install and focuses more on user-friendliness. Examples include; the Ubuntu server, Kali Linux, etc.</p><p>With all this being said, I encourage you to choose Linux as your operating system; here’s a sign to start now.</p><h4>Feel free to drop a few claps and leave a comment on some questions you may have or what you’d like me to improve on, I’m open to suggestions. Thank you.</h4><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d34f2052d8c6" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Virtualization and the Role it Plays in DevOps.]]></title>
            <link>https://medium.com/@Robinson-uche/virtualization-and-the-role-it-plays-in-devops-328f8b6b2de6?source=rss-d19a4cd86f3d------2</link>
            <guid isPermaLink="false">https://medium.com/p/328f8b6b2de6</guid>
            <dc:creator><![CDATA[Robinson uche]]></dc:creator>
            <pubDate>Wed, 26 Oct 2022 21:46:48 GMT</pubDate>
            <atom:updated>2022-10-27T15:43:15.764Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*F_4jWhLz7NvY21CUaq25CQ.jpeg" /><figcaption>Image from <a href="https://circleci.com/blog/top-6-benefits-of-virtualization/">circleci.com</a></figcaption></figure><p>Let me briefly describe life before virtualization to give you an understanding of what it was created for, before delving more into the role of virtualization.</p><p>Before Virtualization, we needed servers which were physical computers to run services, and mind you each main service had to have its own server for the purpose of isolation. For instance, it would be disastrous to operate a web server and a database server on the same system. Also, servers were always over-provisioned and under-utilized because, if a service needs a RAM size of 4Gb you would have to go for 8Gb RAM and most of the time a lot of storage goes to waste when they end up being under-utilized. At the end of the day, the cost of maintenance and procurement is alarming when these problems can be easily solved at a cheaper price by the introduction of “<strong>virtualization</strong>”.</p><p>Without virtualization;</p><ul><li>To run applications/services one needs a physical computer (server) in a data center or in your home.</li><li>For the purpose of isolation, one service — one server.</li><li>Servers are always over-provisioned.</li><li>Most server resources are underutilized.</li><li>It’s capital-intensive and maintenance is high.</li></ul><p><strong>How does virtualization work?</strong></p><blockquote>One computer does the job of multiple computers by running multiple operating systems on one physical machine simultaneously.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/920/1*oVCcs8iLTx7FIqHe96_dZg.png" /><figcaption>Image from <a href="https://linuxiac.com/virtualization-on-pc/">linuxiac.com</a></figcaption></figure><p>Instead of running multiple main services together on one computer, we run multiple operating systems and then run the main services on each operating system. This addresses the need for <strong>isolation</strong>. Your physical resource is partitioned into virtual resources through virtualization. Asides from OS virtualization we also have server virtualization, network virtualization, storage virtualization, etc.</p><p>We can further say with virtualization;</p><ul><li>One computer can run multiple operating systems</li><li>Partition of physical resources into virtual resources</li><li>Server virtualization is the most common type of virtualization.</li></ul><h4><strong>Some terms used in virtualization</strong></h4><p><strong>Host OS</strong>: The operating system of the physical machine (hardware).</p><p><strong>Guest OS</strong>: The operating system of the virtual machines which are sometimes referred to as guest machine.</p><p><strong>VM</strong>: Virtual machine.</p><p><strong>Snapshot</strong>: This is a way of taking a backup of the virtual machine.</p><p><strong>Hypervisor</strong>: This is a tool/software that is used to create virtual machines and enables virtualization. An example is the Oracle VM VirtualBox.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Tyo7bGOXtUW7e6WrXcRZEw.jpeg" /><figcaption>Image from <a href="http://nakivo.com/">nakivo.com</a></figcaption></figure><p>There are two types of hypervisors: Type 1 and Type 2. The primary distinction between type 1 and type 2 hypervisors is that type 1 hypervisors run directly on the physical computer as an OS and are primarily used for production, whereas type 2 hypervisors run as software that can be installed on the computer and are primarily used for learning and testing.</p><p>In conclusion, it can be said that <strong>virtualization</strong> significantly contributes to DevOps by enhancing research and development whilst also enabling development and testing in simulated environments while working on live projects.</p><p>Thank you for reading!❤️</p><p>Please leave a clap👏 and/or a comment if you have any thoughts about this topic.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=328f8b6b2de6" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>