<?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 Seun Runsewe on Medium]]></title>
        <description><![CDATA[Stories by Seun Runsewe on Medium]]></description>
        <link>https://medium.com/@runsewe-seun?source=rss-d0877f410e8b------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*lbgLVc1j3CvEd3jdgkwM2A@2x.jpeg</url>
            <title>Stories by Seun Runsewe on Medium</title>
            <link>https://medium.com/@runsewe-seun?source=rss-d0877f410e8b------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 14:04:49 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@runsewe-seun/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[The First Attack Surface in Your Medical Device Isn’t the Cloud]]></title>
            <link>https://runsewe-seun.medium.com/the-first-attack-surface-in-your-medical-device-isnt-the-cloud-2bd43fb4f657?source=rss-d0877f410e8b------2</link>
            <guid isPermaLink="false">https://medium.com/p/2bd43fb4f657</guid>
            <category><![CDATA[femtech]]></category>
            <category><![CDATA[medical-devices]]></category>
            <category><![CDATA[iot-security]]></category>
            <category><![CDATA[cybersecurity]]></category>
            <category><![CDATA[hardware-engineering]]></category>
            <dc:creator><![CDATA[Seun Runsewe]]></dc:creator>
            <pubDate>Sun, 15 Mar 2026 11:40:37 GMT</pubDate>
            <atom:updated>2026-03-15T11:40:37.879Z</atom:updated>
            <content:encoded><![CDATA[<p>Most hardware security frameworks start at the cloud. This one starts at the silicon, and works forward from there.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Gi7qzyCfkVW0QD6mJJ55hg.jpeg" /></figure><p>Most security reviews of connected health devices start at the API layer.</p><p>By the time you’re auditing your cloud endpoints, you’ve already walked past the real problem. You walked past it in the hardware lab. You walked past it on the factory floor. You probably shipped it to your first thousand users.</p><p>The first attack surface in a connected medical device is the PCB. The silicon.</p><p>Before the firmware, BLE stack or the cloud.</p><p>More specifically, it’s the debug interfaces your firmware engineers used to build the product : the JTAG ports, the UART headers, the SWD connections that made development faster and that nobody formally disabled before the device went into production. Disabling debug interfaces is a separate manufacturing step, it happens after the engineering work feels done, and under pressure it is the kind of step that gets skimmed over.</p><p>I know this because I did the threat modeling for my IoT security architecture for Biamo’s connected health devices. And when I built the framework from scratch, the question “<em>where does an attacker actually begin” </em>was the lamp to my feet.</p><p>Not a compliance checklist 👀.</p><h3>The £30 problem</h3><p>Here is the scenario that reorients the entire threat model.</p><p>A device gets returned under warranty. Or a unit gets purchased second-hand. Or a research engineer buys one off a resale platform.</p><p>They sit down with a Bus Pirate, a hardware tool that costs less than a hydrafacial in London, probe the exposed debug ports, and within minutes they are reading the raw firmware from the device’s flash memory.</p><p>Everything is in there. The application logic. The communication protocols. And, if the device was provisioned the way most early IoT products are provisioned: <em>the cryptographic keys</em>.</p><p>This is not a sophisticated attack. It only costs forty-five minutes, a cheap tool, and a firmware binary that was not designed to resist static analysis.</p><p>The question that follows is the one that determines whether this is a single-device problem or a catastrophic one:</p><blockquote><strong>Are those keys unique per device, or are they shared across the fleet?</strong></blockquote><p>In most early-stage connected hardware products, they are shared. One key, embedded in firmware, deployed to every unit. It seems efficient at the time. This same condition means that extracting credentials from a single returned device can expose the cryptographic identity of every device ever manufactured.</p><p>That is a fleet-level compromise waiting for a single careless moment. The fix is conceptually straightforward, though it requires real manufacturing discipline to implement.</p><p>Each device must carry its own cryptographic identity:</p><p>• <strong>Unique per-device keys</strong><br> • <strong>Keys generated at provisioning time</strong><br> • <strong>Injected during manufacturing</strong><br> • <strong>Stored in a hardware-backed secure element rather than addressable flash</strong></p><p>The security implication is simple. When a single device is compromised, the compromise stops there. The blast radius remains contained.</p><p>This is arguably the single highest-impact security decision a connected hardware company can make. Early-stage teams typically make the wrong call because no one in the room is drawing the blast radius on a whiteboard for all to see.</p><h3>When the attacker owns the firmware, they own the device</h3><p>Once an attacker has extracted the firmware binary, the second phase of the attack begins: reverse engineering.</p><p>Tools like Ghidra, Binwalk, and IDA Pro are free or low-cost, widely documented, and designed exactly for this purpose. A determined researcher (or adversary) can disassemble the binary, identify safety-critical routines, and understand exactly how the device behaves under conditions the manufacturer never intended to expose.</p><p>For a consumer IoT device, this is a privacy problem. For a medical device, it is a patient safety problem.</p><p>Consider what a connected breast pump actually controls: suction strength, pumping rhythm, session timing. The BLE command interface to that device is not a data pipe. It is a physical control API. Commands sent over Bluetooth do not only transmit information — they execute actions in the physical world, against a postpartum body, during a physiologically sensitive period in a woman’s life.</p><p>When you frame it that way, the security posture changes. You are no longer securing a data stream. You are securing a medical control interface. And the physiological expereince of a woman in this context.</p><p>Firmware should be encrypted where possible, and always cryptographically signed. A device must refuse to run code whose integrity it cannot verify. Code signing must protect the integrity of safety-critical routines — to prevent tampering and make tampering detectable.</p><p>And importantly: the firmware signing keys must never exist on the provisioning station itself.</p><h3><strong>The nuclear option: OTA</strong></h3><p>Over-the-air firmware updates are one of the most powerful capabilities a connected device can have. The ability to push fixes, improvements, and security patches to devices already in users’ hands without a recall, without a service centre, without asking anyone to do anything, transforms the economics of hardware.</p><p>It also creates the single largest attack surface in the entire system.</p><p>In the wrong hands, OTA becomes a remote control for every device in your fleet. If the update signing mechanism is compromised if an attacker can produce a firmware image that your device accepts as legitimate arbitrary code can be deployed simultaneously across thousands of devices with the full trust of your users.</p><p>The device will accept it. The user will never see it. And if that device interacts with the physical world, the consequences are no longer a data breach. They become a safety incident at scale.</p><p>The 2016 Mirai botnet demonstrated the consequences at scale. The targets were simple consumer devices. The result was internet infrastructure failure across large portions of the United States.</p><p>Medical and health devices raise the stakes further.</p><p>Securing OTA needs more discipline than exotic architecture:</p><p>• Firmware images must be <strong>cryptographically signed</strong><br> • Devices must <strong>verify signatures before applying updates</strong><br> • Signing keys must live inside a <strong>hardware security module (HSM)</strong><br> • Update servers must use <strong>mutual TLS</strong><br> • <strong>Rollback protection</strong> must prevent downgrade attacks</p><p>Fleet-wide update systems benefit from <strong>multi-person authorization</strong>, an engineering safeguard that prevents a single compromised credential from triggering a fleet-wide incident.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*T8kIHGSKjTHbe8ga9sUfpw.png" /></figure><h3><strong>The manufacturing moment most companies miss</strong></h3><p>There is an attack surface in this stack that receives almost no attention in security writing aimed at hardware startups. If it fails, every downstream control becomes irrelevant.</p><p>The manufacturing and provisioning pipeline.</p><p>Secure boot, unique device keys, and firmware signing all depend on a single assumption: that the keys establishing device trust were generated and injected in a secure manufacturing environment, under controlled conditions, with access restricted to authorized systems.</p><p>If that assumption fails, nothing else matters.</p><p>A provisioning station that stores firmware signing keys in software — accessible to any process running on the same machine — becomes a lone point of catastrophic compromise. An attacker who gains access to that station, whether through a supply chain breach, insider threat, or misconfigured remote access tool, can sign malicious firmware with your legitimate key. Devices produced afterward carry a trusted backdoor. Your users’ devices will accept adversarial updates because your key says they should.</p><p>The design pattern that prevents this is well established, though still underused in practice. The signing key never resides on the provisioning station. Instead, the station sends the firmware artifact to a hardware security module (HSM), which signs it and returns the signed binary. The raw key material never touches a networked system. Access to the HSM is logged, audited, and gated behind multi-person controls.</p><p>When I designed the provisioning architecture for Biamo, this was the layer that required the most careful attention. The technology is not particularly complex. The challenge is that these decisions must be made before manufacturing scale demands them.</p><p>Early teams often postpone this work until volumes to justify the effort.</p><blockquote>But infrastructure, at its core, is safety.</blockquote><p>The systems we design at the beginning eventually sit beneath real human experiences, shaping how people interact with their health, their bodies, and the technology they trust.</p><p>The right time to design for HSM-backed provisioning is before the first batch ships. Don’t wait for the first vulnerability disclosure.</p><h3><strong>End of life is a security problem too</strong></h3><p>Product teams tend to treat device retirement as a post-launch concern. In connected health systems, it is part of the security model.</p><p>A returned, resold, or discarded connected health device carries something most IoT hardware never does: a longitudinal record of intimate physiological behaviour.</p><p>Feeding patterns. Session duration. Device usage at 2 a.m. Biometric baselines from the postpartum period.</p><p>If flash memory is not securely wiped during factory reset — a step that requires explicit engineering design, not a simple “clear user data” command — the next owner of that device inherits the previous user’s health history.</p><p>In clinical environments where devices rotate between patients, that becomes a HIPAA breach. In consumer resale markets, it becomes a privacy failure no encryption key can fix, because the data sits unencrypted in local storage waiting to be extracted.</p><p>Device decommissioning is therefore a workflow to be centered too. It must be designed into the system from the beginning, including certificate revocation at the cloud layer so a factory-reset device cannot silently reconnect under its previous identity.</p><h3><strong>What this framework produces</strong></h3><p>Running the threat model from the PCB outward, rather than beginning in the cloud, changes more than the checklist. It changes the posture of the entire system.</p><p>The questions start to look different:</p><ul><li>What happens when a competitor buys one of our devices and takes it apart?</li><li>What is the blast radius if a single device key is extracted?</li><li>Can the provisioning pipeline be trusted once manufacturing reaches scale?</li></ul><p>These questions surface naturally when the threat model begins at the device itself. Hardware security forms the ground beneath the product — the layer upon which firmware, software, and user trust ultimately rest.</p><p>Firmware, update mechanisms, provisioning pipelines, and decommissioning workflows shape the experience every user eventually touches. These decisions determine whether a device ages into trust or into risk.</p><p>And like most foundational decisions in engineering, the timing matters.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=2bd43fb4f657" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Holding your phone at 3 AM, Mama? You’re Doing Better Than You Think.]]></title>
            <link>https://runsewe-seun.medium.com/holding-your-phone-at-3-am-mama-youre-doing-better-than-you-think-dbeaef23a007?source=rss-d0877f410e8b------2</link>
            <guid isPermaLink="false">https://medium.com/p/dbeaef23a007</guid>
            <category><![CDATA[maternal-health]]></category>
            <category><![CDATA[motherhood]]></category>
            <category><![CDATA[femtech]]></category>
            <category><![CDATA[postpartum]]></category>
            <category><![CDATA[mental-health]]></category>
            <dc:creator><![CDATA[Seun Runsewe]]></dc:creator>
            <pubDate>Mon, 19 May 2025 23:13:14 GMT</pubDate>
            <atom:updated>2025-05-19T23:55:19.615Z</atom:updated>
            <content:encoded><![CDATA[<p>This is one of the notes I tucked into the Biamo app.<br>For the moms scrolling in the dark, wondering if they’re doing enough.<br>You are.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*zWN5vkT5FXcwogpk1jzQsg.jpeg" /></figure><p>It’s usually this time of night when questions bubble up, and you rush to Google or mom forums. Or you opened this app because something felt hard tonight: worrying about your baby’s latch, dealing with the soreness, or the overwhelming feeling of needing a moment where you are seen, not just the baby.</p><p>So let me say this, gently and clearly:</p><p><strong>You’re doing better than you think.</strong></p><p>I know it doesn’t always feel like it. From the milk spills to the crying that feels overwhelming. Your body doesn’t quite feel like yours yet. Maybe your thoughts too.</p><p>But you’re here. Showing up and finding your way in the dark. That really counts. That <em>matters</em>.</p><p>Motherhood isn’t neat. It’s leaky, lonely and full of tiny decisions you didn’t know you’d have to make. And if you’re questioning everything, it’s because you care deeply. That’s how I know you’re doing amazing. That, mama, is love.</p><p>I built Biamo because I clearly remember this hour. I’d be sat on the kitchen floor at 3 AM, with engorged breasts, desperately searching for answers I was too embarrassed to voice. The discomfort of pumping with tools that felt clinical and detached from my reality, I longed for something, someone who just understood.</p><p>That’s why this app is here. Not just to help you track sessions and pump efficiently, but to be a quiet companion in these moments. To remind you that you’re not alone, and to offer tools designed with <em>you</em> in mind — intuitive tracking, AI-backed intelligence that eases your worries, and a hand to hold you as you navigate this life-giving window of time in your life.</p><p>You don’t need to read more if your eyes are heavy. This will be here when you need it. Or not.</p><p>But if tonight you needed to hear:</p><p>“You’re not crazy. You’re not failing. This is just incredibly hard.”</p><p>Then let this be that voice.</p><p>And when the morning comes, as it always does, you’ll still be standing. Maybe tired. But still you. Still here.</p><p>We see you.</p><p>— Biamo</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=dbeaef23a007" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Building a local development environment using Vagrant, Docker Compose, and Localstack (Part 1)]]></title>
            <link>https://runsewe-seun.medium.com/building-a-local-development-environment-using-vagrant-docker-compose-and-localstack-part-1-fff6e4e2e9c1?source=rss-d0877f410e8b------2</link>
            <guid isPermaLink="false">https://medium.com/p/fff6e4e2e9c1</guid>
            <category><![CDATA[staging]]></category>
            <category><![CDATA[devops]]></category>
            <category><![CDATA[cloud]]></category>
            <category><![CDATA[localstack]]></category>
            <category><![CDATA[docker]]></category>
            <dc:creator><![CDATA[Seun Runsewe]]></dc:creator>
            <pubDate>Wed, 13 Nov 2024 14:27:35 GMT</pubDate>
            <atom:updated>2024-11-14T19:22:19.723Z</atom:updated>
            <content:encoded><![CDATA[<p>“Borrowing staging for 20 mins.”</p><p>I read that a lot on Slack. As a developer, have you ever found yourself waiting in line to use the staging environment, to test your code for a few minutes? Yup, my team too.</p><p>Imagine five developers, each working on different feature branches, all needing to deploy and test their changes at the same time. The bottleneck isn’t just frustrating — it’s a productivity killer.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*W3Hx-3ejyh0RaDz2IZBCTA.png" /><figcaption>Some context from a Senior Engineer on my team.</figcaption></figure><p>We wanted a way for each developer to spin up their own full-stack environment that mirrors production as closely as possible, without stepping on each other’s toes or racking up hefty AWS bills.</p><p>Isolation, speed, and reliability on local machines. That was the goal.</p><p>So I turned to <strong>Vagrant</strong>, <strong>Docker Compose</strong>, and <strong>Localstack</strong>. With these tools, I built a local development environment that emulates our entire AWS-based stack — including services like SQS, S3, and IAM — right on our laptops. Now, every developer can deploy, test, and debug their feature branches independently, without waiting or blocking others.</p><p>In this article, I’ll walk you through how I set up this environment. We’ll start with configuring Vagrant and then dive into each Docker service, explaining how everything fits together. Whether you’re tired of staging bottlenecks or just want a reliable local environment that feels like production, read on —this should help.</p><h3>Prerequisites and Initial Setup</h3><p>This setup depends on Vagrant to handle our virtual environment, Docker and Docker Compose for orchestrating our services, and a few handy plugins to keep everything working nicely together.</p><p>Here’s the shopping list:</p><h4>Tools You’ll Need</h4><ol><li><strong>Vagrant</strong>: This is the star of the show. Vagrant is our go-to for creating reproducible development environments. <a href="https://www.vagrantup.com/">Download it here.</a></li><li><strong>Docker</strong> and <strong>Docker Compose</strong>: These handle the containers that actually run our services. <a href="https://www.docker.com/products/docker-desktop">You’ll want to have both installed</a> so that Vagrant can hand off the heavy lifting to Docker.</li><li><strong>Vagrant Docker Compose Plugin</strong>: Since we’re using Vagrant to manage Docker Compose, we’ll need this plugin to bridge the gap:</li></ol><pre>vagrant plugin install vagrant-docker-compose</pre><ul><li>Once you have these in place, it’s time to grab the code and set things in motion.</li></ul><h4>Cloning the Repository</h4><p>Our starting point is my Vagrant Docker Localstack repo. This is the playground we’ll use to bring all the pieces together, so let’s get it on your machine:</p><pre>git clone https://github.com/srunsewe12/Vagrant-Docker-Localstack.git<br>cd Vagrant-Docker-Localstack</pre><p>FYI: you’re standing on the edge of something cool. All that’s left to do is fire up Vagrant and let it work.</p><h4>Bringing Up the Environment</h4><p>With the repo ready to play around, we can kick off the environment using the following command:</p><pre>vagrant up --provider=docker</pre><p>This command tells Vagrant to spin up a Docker-based environment using the configurations we’ve defined in the Vagrantfile. Once you hit enter, sit back for a second (or make iced coffee like me) — Vagrant will pull in the Docker images and build up your local ecosystem.</p><h4>Stopping and Cleaning Up</h4><p>When you need to pause the environment, or reclaim some system resources, you can stop or completely shut down the environment:</p><ul><li><strong>Stop</strong> the environment (without tearing it all down):</li></ul><pre>vagrant halt</pre><ul><li><strong>Destroy</strong> the environment (when you’re ready for a fresh start next time):</li></ul><pre>vagrant destroy</pre><p>This way, you’re in full control. You can spin the environment up whenever you need it, halt it to free up memory, or nuke it entirely and start fresh. The flexibility here is honestly a breath of fresh air for anyone used to waiting in line for staging.</p><h3>Breaking Down the Vagrantfile</h3><p>Alright, our environment prepped. Next, let’s get to the heart of this setup: the <strong>Vagrantfile</strong>.</p><h4>What Exactly Is a Vagrantfile, and Why Should We Care?</h4><p>This unassuming file is the <em>belle</em> of your local development ball. This is where we define the “what” and “how” of your environment, seeing that everything plays nicely together. By using Vagrant, we’re able to automate the creation and configuration of our development environment. No more manual setups, no more “but it worked on my machine.” Everything is defined, consistent, and repeatable.</p><p>Using Vagrant also means we can leverage <strong>provider plugins</strong> — in our case, Docker — to run lightweight containers instead of full-blown virtual machines. This gives you a lean and efficient setup, which is great for a local environment that’s heavy on services.</p><h4>Peeking Under the Hood: Our Vagrantfile Line by Line</h4><p>Let’s dissect this Vagrantfile and see why it makes our environment work.</p><p>1. <strong>Selecting the Base Image</strong></p><pre>Vagrant.configure(&quot;2&quot;) do |config|<br>  config.vm.provider &quot;docker&quot; do |d|<br>    d.image = &quot;ubuntu:20.04&quot;<br>    d.has_ssh = false<br>  end</pre><p>Here, we’re telling Vagrant to use Docker as our provider and to pull the ubuntu:20.04 image as our base. Why Ubuntu 20.04? It&#39;s stable, widely used, and aligns closely with many production environments. Using a consistent OS between development and production helps avoid those sneaky OS-specific bugs.</p><p>Setting d.has_ssh = false keeps things simple—we don&#39;t need SSH access into our Docker container from Vagrant. Less complexity, fewer problems. This bit spun me a few times.</p><p>2. <strong>Provisioning Docker and Docker Compose</strong></p><pre>config.vm.provision &quot;docker&quot;<br>  config.vm.provision :docker_compose, yml: &quot;/vagrant/docker-compose.yml&quot;, run: &quot;always&quot;</pre><p>This is where the magic happens. We’re instructing Vagrant to:</p><ul><li><strong>Install Docker</strong>: This makes sure Docker is available in our environment.</li><li><strong>Run Docker Compose</strong>: This uses the docker-compose.yml file located in our project directory to spin up all our services.</li></ul><p>By setting run: &quot;always&quot;, we&#39;re making sure Docker Compose starts up our services every time we run vagrant up. It&#39;s the &quot;set it and forget it&quot; of environment setups.</p><p><em>But why use Docker Compose inside Vagrant?</em> You’re stacking two power tools together. Vagrant manages the overarching environment, ensuring consistency across different machines, while Docker Compose handles the orchestration of individual services. This combination gives you the best of both worlds.</p><p>3. <strong>Setting Environment Variables</strong></p><p>Now, we’re getting into the fine details. This shell script runs inside our Vagrant environment and adds a series of environment variables to /etc/environment.</p><pre>config.vm.provision &quot;shell&quot;, inline: &lt;&lt;-SHELL<br>    # Set environment variables<br>    echo &quot;export LOG_LEVEL=DEBUG&quot; &gt;&gt; /etc/environment<br>    echo &quot;export API_KEY=your_api_key_here&quot; &gt;&gt; /etc/environment<br>    echo &quot;export PLATFORM_ENDPOINT=https://your-platform-endpoint.com&quot; &gt;&gt; /etc/environment<br>    echo &quot;export INTEGRATOR_ENDPOINT=https://your-integrator-endpoint.com&quot; &gt;&gt; /etc/environment<br>    echo &quot;export DEFAULT_REGION=us-east-1&quot; &gt;&gt; /etc/environment<br>    echo &quot;export ENVIRONMENT=development&quot; &gt;&gt; /etc/environment<br>    echo &quot;export AWS_SQS_QUEUE_NAME=your_sqs_queue_name&quot; &gt;&gt; /etc/environment<br>    echo &quot;export AWS_SQS_EXPORTS_QUEUE_NAME=your_exports_queue_name&quot; &gt;&gt; /etc/environment<br>    echo &quot;export CALLBACK_API_KEY=your_callback_api_key&quot; &gt;&gt; /etc/environment<br>    echo &quot;export EXTERNAL_API_URL=https://your-external-api.com&quot; &gt;&gt; /etc/environment<br>    echo &quot;export STORAGE_BUCKET_NAME=your_bucket_name&quot; &gt;&gt; /etc/environment<br>    echo &quot;export INTERCOM_APP_ID=your_intercom_app_id&quot; &gt;&gt; /etc/environment<br>    echo &quot;export DEFAULT_LLM_MODEL=your_llm_model&quot; &gt;&gt; /etc/environment<br>    echo &quot;export LLM_API_KEY=your_llm_api_key&quot; &gt;&gt; /etc/environment<br>    echo &quot;export DEFAULT_LLM_URL=https://your-llm-url.com&quot; &gt;&gt; /etc/environment<br>  SHELL<br>end</pre><p>You need these variables—they configure your application and services, telling them where to find APIs, how to authenticate, and what settings to use.</p><p>By externalizing these settings into environment variables, you keep your codebase clean and make it easy to change configurations without touching the code. Plus, you avoid the cardinal sin of hardcoding sensitive information (<em>seriously, don’t do that.</em>)</p><p>4. <strong>Bringing It All Together</strong></p><p>With these configurations, our Vagrantfile is going to create a consistent, fully-featured development environment with a single command. When you run vagrant up, here&#39;s what happens:</p><ol><li><strong>Vagrant Reads the Vagrantfile</strong>: It knows to use Docker as the provider and to pull the specified Ubuntu image.</li><li><strong>Provisioning Starts</strong>: Docker and Docker Compose are installed and configured.</li><li><strong>Environment Variables Are Set</strong>: The shell script runs, injecting all your environment variables into the system.</li><li><strong>Docker Compose Spins Up Services</strong>: Using your docker-compose.yml, all the services (Django app, PostgreSQL, Redis, Kafka, Localstack, etc.) start up and connect as defined.</li></ol><h4>Why This Setup Works</h4><p>By using Vagrant in combination with Docker and Docker Compose, we’re achieving:</p><ul><li><strong>Consistency</strong>: Every developer gets the same environment. No more “but it works on my machine”.</li><li><strong>Isolation</strong>: Your services run in containers, avoiding conflicts with other software on your machine.</li><li><strong>Reproducibility</strong>: Need to onboard a new developer? They can clone the repo, run vagrant up, and be up and running quickly.</li><li><strong>Scalability</strong>: Want to add a new service? Just update the docker-compose.yml, and everyone gets it next time they bring up the environment.</li></ul><h4>A Quick Note on Environment Variables</h4><p>Before you start the environment, remember to replace the placeholder values in the shell script with your actual configurations. Ideally, use a .env file to manage these variables more securely and conveniently.</p><h3>Conclusion</h3><p>In this article, we have dissected the Vagrantfile and laid a great foundation for our local environment. Further, we have set up Vagrant to provision Docker and Docker Compose, configured our base image, and defined important environment variables. This groundwork affords you a consistent and reliable setup that mimics the real production environment every time you do a spin-up for testing.</p><p>Feel free to clone the repo and play around with it — I’m itching to learn how it works out for you and your team.</p><p>In <strong>Part 2</strong>, we’ll dive deeper into the docker-compose.yml file, where we&#39;ll define each of our services in detail. We&#39;ll explore how Localstack emulates AWS services, how our database and caching layers fit in, and how everything works in harmony. See you then!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=fff6e4e2e9c1" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Monitoring Celery: My Walk with Flower, Prometheus and Grafana]]></title>
            <link>https://runsewe-seun.medium.com/monitoring-celery-my-walk-with-flower-prometheus-and-grafana-4dcab785561b?source=rss-d0877f410e8b------2</link>
            <guid isPermaLink="false">https://medium.com/p/4dcab785561b</guid>
            <category><![CDATA[monitoring]]></category>
            <category><![CDATA[prometheus]]></category>
            <category><![CDATA[docker]]></category>
            <category><![CDATA[celery]]></category>
            <category><![CDATA[grafana]]></category>
            <dc:creator><![CDATA[Seun Runsewe]]></dc:creator>
            <pubDate>Wed, 30 Oct 2024 13:57:31 GMT</pubDate>
            <atom:updated>2024-10-30T13:57:31.412Z</atom:updated>
            <content:encoded><![CDATA[<p>First, what is Celery and why am I walking with it?</p><p>Celery is a background task manager that helps you run tasks asynchronously in your applications. Imagine you’re a busy CEO with many tasks that keep your work running — scheduling meetings, generating reports and processing large amounts of data. Celery, like a solid personal assistant, handles the important repetitive tasks while you do the more important ones.</p><p>However, having a distributed task queue system is one thing. Monitoring it to see what’s going on under the hood is just as important. I set up a team to monitor Celery here:</p><ol><li>Flower offers visibility into the distributed task processing system (Celery), which will feed into…</li><li>Prometheus, plays the role of metrics collector and record keeper.</li><li>And then there’s , Grafana. She’s the visual one — you will see all the data about Celery tasks in bite-size, easy-on-the-eye (and your cognitive load) formats to help you understand what’s going on and take action where necessary.</li></ol><p>In this article, I walk you through my process of setting these tools up to monitor Celery. Of course, I’m going to share most of my mistakes too so you learn from them but in the end, we get there.</p><p><strong>Here is what I set out to achieve:</strong></p><ul><li>Set up Celery with a Redis broker</li><li>Install and configure Flower for real-time task monitoring</li><li>Integrate Prometheus to collect metrics</li><li>Visualize everything with Grafana dashboards</li><li>Containerize our setup with Docker (because who doesn’t love containers?)</li></ul><p>Let’s dive in.</p><h3>Getting the Work Done: Setting Up and Monitoring Celery</h3><h4>1. Setting Up Celery with Redis as the Broker</h4><p>First, I set up Celery itself and set it in motion to do those background tasks. I used Redis for my setup like so:</p><ul><li><strong>Install Celery and Redis:</strong></li></ul><pre>pip install celery redis</pre><ul><li><strong>Configure Celery:</strong></li></ul><p>In my Django app, I configured Celery with Redis as the broker in a celery.py file.</p><pre>from celery import Celery  <br><br>app = Celery(&#39;my_app_name&#39;,            <br>             broker=&#39;redis://localhost:6379/0&#39;, <br>             backend=&#39;redis://localhost:6379/0&#39;,              <br>             include=[&#39;my_app_name.tasks&#39;])</pre><ul><li><strong>Create a </strong><strong>tasks.py file: </strong>Here, I defined the tasks I wanted and integrated Prometheus metrics to report to me on those tasks:</li></ul><pre>from celery import Celery<br>from prometheus_client import start_http_server, Counter, Gauge<br><br>app = Celery(&#39;tasks&#39;, broker=&#39;redis://localhost:6379/0&#39;)<br><br># Define Prometheus metrics<br>task_counter = Counter(&#39;celery_tasks_total&#39;, &#39;Total number of Celery tasks&#39;)<br>task_in_progress = Gauge(&#39;celery_tasks_in_progress&#39;, &#39;Number of Celery tasks in progress&#39;)<br><br>@app.task<br>def add(x, y):<br>    task_counter.inc()<br>    task_in_progress.inc()<br>    result = x + y<br>    task_in_progress.dec()<br>    return result<br><br># Start Prometheus HTTP server to expose metrics<br>if __name__ == &#39;__main__&#39;:<br>    start_http_server(8000)<br>    app.start()<br>    add.delay(4, 6)</pre><h4>2. Running Celery Workers</h4><ul><li><strong>Start the Celery worker:</strong></li></ul><pre>celery -A my_app_name worker --loglevel=info</pre><ul><li><strong>Make sure the workers are running: </strong>Check the status of your workers by running:</li></ul><pre>celery -A my_app_name status</pre><h4>3. Setting Up Flower for Monitoring Celery</h4><p>Next, we set up Flower for Celery. Flower is the tool that gives you a visual dashboard for Celery.</p><ul><li><strong>Install Flower:</strong></li></ul><pre>pip install flower</pre><ul><li><strong>Run Flower: </strong>when you’re sure the Celery workers are active:</li></ul><pre>celery -A my_app_name flower --port=5555</pre><ul><li><strong>Access Flower: </strong>navigate to http://localhost:5555 in your browser to monitor your Celery tasks.</li></ul><h4>4. Integrating Prometheus for Metrics Collection</h4><ul><li><strong>Install Prometheus Client:</strong></li></ul><pre>pip install prometheus_client</pre><ul><li><strong>Expose Metrics in Your App: </strong>I added a /metrics endpoint to the main Python file to request timings:</li></ul><pre>from prometheus_client import start_http_server, Summary<br>import time<br>import random<br><br>REQUEST_TIME = Summary(&#39;request_processing_seconds&#39;, &#39;Time spent processing request&#39;)<br><br>@REQUEST_TIME.time()<br>def process_request(t):<br>    time.sleep(t)<br><br>if __name__ == &#39;__main__&#39;:<br>    start_http_server(8000)<br>    while True:<br>        process_request(random.random())</pre><ul><li><strong>Prometheus Configuration: </strong>I then configured Prometheus to scrape metrics from the application in a prometheus.yml file (create a new file called prometheus.yml ):</li></ul><pre>global:<br>  scrape_interval: 15s<br>  evaluation_interval:15s<br><br>scrape_configs:<br>  - job_name: prometheus<br>    static_configs:<br>      - targets: [&#39;localhost:9090&#39;]<br>  - job_name: flower<br>    static_configs:<br>      - targets: [&#39;localhost:5555&#39;]</pre><ul><li><strong>Start Prometheus:</strong></li></ul><pre>prometheus --config.file=prometheus.yml</pre><h4>5. Setting Up Grafana for Visualisation</h4><p>Let’s make the data look pretty with Grafana (trust me its prettier than Flower)</p><ul><li><strong>Install Grafana: </strong>Download and install Grafana based on your OS instructions from the Grafana website.</li><li><strong>Add Prometheus as a Data Source: </strong>Navigate to Grafana’s web interface at <a href="http://localhost:3000.">http://localhost:3000.</a></li><li>Go to <strong>Configuration &gt; Data Sources</strong> and add Prometheus with the URL <a href="http://localhost:9090.">http://localhost:9090.</a></li><li><strong>Create Dashboards: </strong>Create custom dashboards using the metrics exposed by Prometheus.</li></ul><h4>6. Setting Up Sentry for Error Tracking</h4><p>Errors are all to important to understand what is going on when things are not working as they ought to. Let’s use Sentry for tracking this.</p><ul><li><strong>Install Sentry SDK:</strong></li></ul><pre>pip install sentry-sdk==2.11.0</pre><ul><li><strong>Integrate Sentry in Your App: </strong>Initialize Sentry in your app’s main entry point (e.g., settings.py for Django in my case):</li></ul><pre>import sentry_sdk<br>from sentry_sdk.integrations.django import DjangoIntegration<br><br>sentry_sdk.init(<br>    dsn=&quot;YOUR_SENTRY_DSN&quot;,<br>    integrations=[DjangoIntegration()],<br>    traces_sample_rate=1.0,<br>    send_default_pii=True<br>)</pre><h4>7. Running Everything Concurrently with Docker</h4><ul><li><strong>Dockerize the Setup: </strong>Make sure each service (Celery, Redis, Flower, Prometheus, Grafana) has its own container by creating a docker-compose.yml file to manage them together.</li></ul><pre>version: &#39;3&#39;<br>services:<br>  redis:<br>    image: &quot;redis:alpine&quot;<br>  web:<br>    build: .<br>    command: python manage.py runserver 0.0.0.0:8000<br>    volumes:<br>      - .:/code<br>    ports:<br>      - &quot;8000:8000&quot;<br>    depends_on:<br>      - redis<br>  celery:<br>    build: .<br>    command: celery -A my_app_name worker --loglevel=info<br>    volumes:<br>      - .:/code<br>    depends_on:<br>      - redis<br>  flower:<br>    build: .<br>    command: celery -A my_app_name flower --port=5555<br>    ports:<br>      - &quot;5555:5555&quot;<br>    depends_on:<br>      - redis</pre><ul><li><strong>Start the Environment: </strong>with this command</li></ul><pre>docker-compose up</pre><ul><li><strong>Monitor and Troubleshoot: </strong>Here are few ways to consider monitoring and troubleshooting:</li><li>Use multiple terminals to monitor logs, metrics, and services.</li><li>Regularly check Prometheus and Grafana dashboards for system health.</li><li>Monitor Celery tasks through Flower’s web interface.</li></ul><p>This setup covers the primary steps needed to integrate Celery with Flower, Prometheus, Grafana, and Sentry, to make for a robust and observable environment.</p><h3>Lessons Learned: My Troubleshooting Diary</h3><p>Alright, I have been itching to tell you the gist of it — I find that the bulk of my learning happened in the mistakes I made or the errors I ran into. Here are some of them:</p><h4><strong>Redis ghosted me</strong></h4><p>You know that moment when you’ve set up everything perfectly and you hit “run,” then nothing works? That was me— I’d spent ages configuring Celery and its components and Redis had stopped running.</p><p><em>What I learned:</em> Always, always check if Redis is actually running. If you’re using Docker (like I did), spin up a Redis container with:</p><pre>docker run -p 6379:6379 redis</pre><p>Pro tip: Make sure Redis is alive on port 6379 before you start pulling your hair out.</p><h4><strong>Grafana and Prometheus were not getting along…</strong></h4><p>The way integrations work is that all the tools should depend on each other to give you the value you’re looking for. Prometheus typically feeds Grafana with data and it just wasn’t working. I may or may not have screamed at my three screens a few times — I’d set everything up correctly! (Or not).</p><p><em>What I learned:</em></p><p><strong>a. Verify Prometheus is Running:</strong> First, check if Prometheus is active athttp://localhost:9090. If it’s not, that’s your starting point.</p><p>b. <strong>Add Prometheus as a Data Source:</strong> In Grafana, go to <strong>Configuration</strong> &gt; <strong>Data Sources</strong> &gt; <strong>Add Prometheus</strong> and ensure the URL is set to <a href="http://localhost:9090/">http://localhost:9090</a>.</p><p>c. <strong>Check Network Connectivity:</strong> If Grafana still can’t connect, there might be network issues or firewall rules blocking communication. Verify that both tools are on the same network and that no firewalls are interfering.</p><p>d. <strong>Restart and Recheck:</strong> Sometimes, a simple restart of the Docker containers for both Prometheus and Grafana can resolve connection issues. If you still have issues, inspect logs for any errors.</p><p><em>Pro Tip</em>: Be almost paranoid about your configurations pointing to the right URLs and ports. A small typo can sometimes be the culprit.</p><h4><strong>Docker was hugging my CPU</strong></h4><p>Docker is usually a lifesaver, but not when it makes your laptop sound like a jet engine. I was in the middle of my environment setup when my CPU was maxed out, my fan was blasting, and my machine was overheating. Not fun.</p><p><em>Here’s what I did:</em></p><p>a. <strong>Stop Unnecessary Containers</strong>: First, check what’s running. Use docker ps to list all containers and docker stop &lt;container_id&gt; for any you don’t need. This freed up some resources right away.</p><p>b. <strong>Manage Container Restarts</strong>: If Docker keeps restarting, it could be due to auto-restart settings. Check Docker’s settings to ensure only necessary containers have restart policies.</p><p>c. <strong>Manually Stop Docker</strong>: Sometimes, you need to take more drastic measures. I found that manually stopping Docker Desktop from the application itself helped calm things down.</p><p><em>Pro Tip: </em>Keep an eye on Docker’s resource usage (the activity monitor on your Mac is your friend). If your CPU starts spiking, it’s usually because of a runaway container. Docker is powerful, but it can easily overrun your system if left unchecked.</p><h4><strong>VSCode couldn’t find my Python Packages</strong></h4><p>I had just installed a Python package and VS Code insisted it couldn’t find it when I tried to deploy my configuration. Wahala!</p><p>Here’s what was really happening:</p><p><strong>a. Virtual Environment Confusion:</strong> VS Code was using a different Python interpreter than the one in my active virtual environment.</p><p><strong>b. Cached Python Interpreter:</strong> VS Code sometimes caches the Python interpreter path, which might not reflect recent changes to your environment. Despite all my changes, it refused to acknowledge the new packages.</p><p><strong>c. Workspace Settings:</strong> My VS Code workspace settings pointed to an outdated Python interpreter, probably from an old project I’d forgotten about. Pointing to the wrong Python interpreter can happen if you have conflicting configurations.</p><p><em>The solution:</em> Make sure your virtual environment is not just created but activated. When in doubt:</p><pre>pip install -r requirements.txt</pre><p>And if VS Code is still stubborn, try reopening your project in a containerized environment. It will give your code editor a fresh start.</p><p><em>Pro Tip</em>: Always double-check which Python interpreter VS Code is using. It’s in the bottom left corner of the VS Code window. If it’s not pointing to your virtual environment, that’s your culprit!</p><h4><strong>Flower Power… or Lack Thereof</strong></h4><p>I had finished deploying Celery and was ready for the tasks to be monitored by Flower but it just wasn’t working.</p><p><em>What worked for me:</em></p><p>a. Check your Celery Workers to make sure they’re up and running by running this command:</p><pre>celery -A &lt;my_app_name&gt; status</pre><p>Replace &lt;my_app_name&gt; with the name of your Celery application. This command will list all active workers and their status, showing whether they are connected and processing tasks.</p><p>If the workers are running, you’ll see their names and details like the number of tasks each has handled. If you see an error or no workers listed, that means they aren’t running, and you may need to start them with:</p><pre>celery -A &lt;my_app_name&gt; worker --loglevel=info</pre><p>This command will start the worker processes, allowing them to handle tasks.</p><p>b. Then, with the confidence of someone who definitely hasn’t made this mistake before, run:</p><pre>celery -A tasks flower --port=5555</pre><p>c. Check the Flower URL after starting Flower. Make sure you’re accessing the correct URL — it should be something like <a href="http://localhost:5555.">http://localhost:5555.</a></p><h4><strong>The Ghost of Sentry Past</strong></h4><p>While setting up Sentry for error monitoring, I encountered persistent errors about a missing sentry_sdk.</p><p><em>What I did</em>: After some investigation, I realized that the sentry_sdk package wasn&#39;t listed in my requirements.txt file. Since it wasn&#39;t installed, my app couldn&#39;t use Sentry, leading to the errors. To resolve this, I added sentry-sdk==2.11.0 to requirements.txt and ran:</p><pre>pip install -r requirements.txt</pre><p>This installed the SDK and cleared up the errors immediately.</p><p><em>Pro Tip:</em> Always double-check that all required packages are listed in requirements.txt to avoid these kinds of issues.</p><h4><strong>Docker Network Conflicts</strong></h4><p>During the setup, I ran into networking issues where ports were already in use (this happened because I’d played around with the same ports from another task), and containers couldn’t communicate as expected. These conflicts can be tricky to diagnose and disrupt the entire workflow, especially when trying to connect multiple services like Redis, Celery, and Flower.</p><p><em>Untangling the mess:</em></p><ul><li>Before you docker run, make sure the coast is clear. Use docker ps to see what&#39;s running.</li><li>If a port is taken, don’t be afraid to use a different one. For example, consider switching from -p 8000:8000 to -p 8080:8000 if 8000 is busy with some other tool you’ve played around with.</li></ul><p><em>Pro Tip:</em> Regularly monitor Docker’s running containers to avoid network conflicts, especially when reusing ports.</p><h4><strong>The Multi-Terminal Juggling Act</strong></h4><p>As the complexity of the setup grew, so did the number of terminals I needed open at once. Managing multiple processes like Celery workers, Flower, Redis, and Grafana required keeping track of different commands in separate terminal windows, which can quickly become overwhelming.</p><p><em>What I learned:</em><br>1. Use Multiple Terminals: Open separate terminals for each service (e.g., Celery, Flower, Redis).</p><p>2. Stay Organized: Keep track of what each terminal is doing to avoid confusion.</p><p>This one made me feel like a proper hacker from a 90s movie — remember the one where Angelina Jolie had that red leather top on with spiky hair?</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4dcab785561b" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Playing Around with ArgoCD: From Setup to Production]]></title>
            <link>https://runsewe-seun.medium.com/playing-around-with-argocd-from-setup-to-production-e2403676a898?source=rss-d0877f410e8b------2</link>
            <guid isPermaLink="false">https://medium.com/p/e2403676a898</guid>
            <category><![CDATA[github]]></category>
            <category><![CDATA[ci-cd-pipeline]]></category>
            <category><![CDATA[nginx]]></category>
            <category><![CDATA[kubernetes]]></category>
            <category><![CDATA[argo-cd]]></category>
            <dc:creator><![CDATA[Seun Runsewe]]></dc:creator>
            <pubDate>Thu, 01 Aug 2024 20:01:07 GMT</pubDate>
            <atom:updated>2024-08-01T20:01:07.159Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*4ev5XGagRA8qyi9ZcHioBQ.png" /></figure><p>I recently joined <a href="https://www.credrails.com/">Credrails</a>, an incredible company working to build the infrastructure for transaction reconciliation in Africa as a Cloud Engineer and was confident in my CI/CD skills…</p><p>Until I jumped on a call about some errors we were running into on a deployment with ArgoCD. I had been flirting so much with Jenkins that I forgot about ArgoCD.</p><p>If I was going to add value to my new team, I needed to get up to speed with this tool — and fast.</p><p>So I got my hands dirty and I am here to share my experience with anyone looking to pick this up quickly. In this article, we cover the scope of the task I gave myself, my approach to getting it done and some mistakes I made along the way (so you don’t have to).</p><p><strong>Here’s the task I completed to get up to speed with ArgoCD.</strong></p><p>On my MacBook Pro, I wanted to create a local Kubernetes environment, configure ArgoCD, and manage a Kubernetes application.</p><p>My goal was to gain hands-on experience with ArgoCD by:</p><ul><li>Setting up a local Kubernetes environment on macOS</li><li>Installing and configuring ArgoCD</li><li>Organizing a <a href="https://github.com/srunsewe12/argocd-playaround.git">GitHub repository</a> with Kubernetes manifests</li><li>Creating and managing an application using ArgoCD</li></ul><h4><strong>The Journey</strong></h4><p><strong>1. Setting Up the Kubernetes Environment</strong></p><p>I began by creating a local Kubernetes cluster using Docker Desktop for Mac, which includes Kubernetes:</p><ul><li><strong>Install Docker Desktop for Mac (if not already installed):</strong></li></ul><pre>brew install --cask docker</pre><ul><li><strong>Enable Kubernetes in Docker Desktop:</strong></li></ul><ol><li>Open Docker Desktop</li><li>Go to Preferences &gt; Kubernetes</li><li>Check “Enable Kubernetes”</li><li>Click “Apply &amp; Restart”</li></ol><ul><li><strong>Verify the Kubernetes setup:</strong></li></ul><pre>kubectl get nodes</pre><p>This step provided a local Kubernetes environment for my ArgoCD playaround.</p><p><strong>2. Installing ArgoCD</strong></p><p>With Kubernetes running, I installed ArgoCD:</p><pre>kubectl create namespace argocd<br>kubectl apply -n argocd -f https://raw.githubusercontent.com/argoproj/argo-cd/stable/manifests/install.yaml</pre><p>To access the ArgoCD UI locally, I set up port forwarding:</p><pre>kubectl port-forward svc/argocd-server -n argocd 8080:443</pre><p><strong>3. Accessing and Configuring ArgoCD</strong></p><p>To access ArgoCD for the first time, I retrieved the initial admin password:</p><pre>kubectl -n argocd get secret argocd-initial-admin-secret -o jsonpath=&quot;{.data.password}&quot; | base64 -d</pre><p>When I encountered login issues, I reset the admin password:</p><pre>argocd account update-password</pre><p><strong>Note:</strong> This command requires the ArgoCD CLI, which can be installed via Homebrew:</p><pre>brew install argocd</pre><p><strong>4. Organizing the Repository Structure</strong></p><p>I structured my GitHub repository to manage Kubernetes manifests:</p><pre>mkdir nginx-app<br>mv nginx-deployment.yaml nginx-app/</pre><p>Then, I created a kustomization.yaml file:</p><pre>cat &lt;&lt;EOF &gt; nginx-app/kustomization.yaml<br>resources:<br>- nginx-deployment.yaml<br>- service.yaml<br>EOF</pre><p><strong>5. Creating and Managing the ArgoCD Application</strong></p><p>I created an ArgoCD application by first writing an application.yaml:</p><pre>apiVersion: argoproj.io/v1alpha1<br>kind: Application<br>metadata:<br>  name: nginx-app<br>  namespace: argocd<br>spec:<br>  project: default<br>  source:<br>    repoURL: &#39;https://github.com/srunsewe12/argocd-playaround.git&#39;<br>    targetRevision: HEAD<br>    path: &#39;nginx-app&#39;<br>  destination:<br>    server: &#39;https://kubernetes.default.svc&#39;<br>    namespace: default<br>  syncPolicy:<br>    automated:<br>      prune: true<br>      selfHeal: true</pre><p>Then, I applied the application specification:</p><pre>kubectl apply -f application.yaml</pre><p><strong>6. Troubleshooting and Verification</strong></p><p>To ensure everything worked correctly, I performed several verification steps:</p><pre>kubectl get pods -n default<br>kubectl get svc -n default</pre><p>For troubleshooting, I checked the ArgoCD application controller logs:</p><pre>kubectl logs -n argocd -l app.kubernetes.io/name=argocd-application-controller</pre><h4>The Shege I saw (The Roadblocks I faced) along the way…</h4><ol><li><strong>ArgoCD Login Issues</strong> — I couldn’t log into the ArgoCD UI with default credentials. To troubleshoot (fancy word for figuring out what the problem is and fixing it), I tried to reset the admin password using the ArgoCD CLI. Here is the sequence of commands that fixed this:</li></ol><p><strong>i. Retrieve the Existing Password</strong>: If you haven’t reset the password and want to retrieve the default password, you can use this command:</p><pre>kubectl -n argocd get secret argocd-initial-admin-secret -o jsonpath=&quot;{.data.password}&quot; | base64 -d</pre><p><strong>ii. Generate a New Bcrypt Hash for the Password</strong>: Instead of generating a new bcrypt hash yourself, you can use a pre-provided hash to reset the admin password. The hash I used corresponds to the plaintext password “password” Here’s the command I used:</p><pre>kubectl -n argocd patch secret argocd-secret \<br>-p &#39;{&quot;stringData&quot;: {<br>  &quot;admin.password&quot;: &quot;$2a$10$7EQytL4kUwPPL1BqCbuNWe9eUb5KgKt/zWfa8qVcAohdHTZd1g6Ni&quot;,<br>  &quot;admin.passwordMtime&quot;: &quot;&#39;$(date +%FT%T%Z)&#39;&quot;<br>}}&#39;</pre><p><strong>iii. Verify the ArgoCD Server is Running:</strong> After patching the secret, you need to make sure the ArgoCD server is running and ready to accept the new credentials.</p><pre>kubectl get pods -n argocd</pre><p><strong>iv. Port Forward to Access the ArgoCD UI:</strong> If not already done, set up port forwarding to access the ArgoCD UI locally.</p><pre>kubectl port-forward svc/argocd-server -n argocd 8080:443</pre><p><strong>v. Log into ArgoCD UI:</strong> Open a web browser and navigate to https://localhost:8080. Use the new password “password” you set to log in as the admin user.</p><p><strong>vi. Confirm Login:</strong> Once logged in, verify that you can access the ArgoCD dashboard and manage applications.</p><p>Even though it took me 2 days, by following these steps, I was able to reset the admin password and successfully log into the ArgoCD UI, allowing me to proceed with my tasks. Follow this approach to gain access even if the original credentials are lost or forgotten.</p><p><strong>Quick Fix:</strong> If you’re in a hurry, just run steps 2 and 4, then log in with the new password.</p><p><strong>Lesson Learned:</strong> Always have a plan to reset credentials, and don’t panic if you can’t log in initially!</p><p>2. <strong>Repository Structure Issues:</strong> Fixing Manifest Generation Errors — My initial repository structure didn’t support Kustomize, which led to manifest generation errors. To resolve this, I needed to reorganize my repository to align with Kustomize’s requirements. Here’s how I addressed this problem:</p><p>i. <strong>Create a Directory for the Application:</strong> I created a new directory within my GitHub repository to organize the Kubernetes manifests for my application.</p><pre>mkdir nginx-app</pre><p>ii. <strong>Move the Deployment YAML to the New Directory:</strong> Next, I moved the existing nginx-deployment.yaml file into the newly created nginx-app directory. This step ensured that all related manifests were in one place.</p><pre>mv nginx-deployment.yaml nginx-app/</pre><p>iii. <strong>Create a kustomization.yaml File:</strong> I then created a kustomization.yaml file within the nginx-app directory. This file specified which resources Kustomize should manage.</p><pre>cat &lt;&lt;EOF &gt; nginx-app/kustomization.yaml<br>resources:<br>- nginx-deployment.yaml<br>EOF</pre><p>iv. <strong>Address Missing Service File:</strong> To avoid sync errors, I created a missing service.yaml file and included it in thekustomization.yaml file:</p><pre>cat &lt;&lt;EOF &gt; nginx-app/service.yaml<br>apiVersion: v1<br>kind: Service<br>metadata:<br>  name: nginx-service<br>spec:<br>  selector:<br>    app: nginx<br>  ports:<br>  - protocol: TCP<br>    port: 80<br>    targetPort: 80<br>EOF</pre><p>Then, I updated the kustomization.yaml to reference the new service file:</p><pre>echo &quot;- service.yaml&quot; &gt;&gt; nginx-app/kustomization.yaml</pre><p>v. <strong>Verify the New Structure:</strong> Finally, I made sure the new structure was correct by syncing the application in ArgoCD and checking the resources created in the Kubernetes cluster.</p><pre>kubectl get pods -n default<br>kubectl get svc -n default</pre><p><strong>Lesson Learned:</strong> Properly structuring your repository is crucial when working with tools like Kustomize. Ensuring that all referenced manifests are present and correctly organized can prevent errors and streamline your deployment process.</p><p>This version accurately reflects the steps you took to resolve the issue and reorganize your repository structure.</p><p>3. <strong>Port Forwarding Issues</strong> — After troubleshooting all these issues (including resetting the admin password), I restarted the ArgoCD server pod to ensure all changes were applied correctly. However, this caused the port forwarding session to stop, denying my access to the ArgoCD UI (…again arrgghhh!!!). Here’s how I resolved it:</p><p>i. <strong>Verify the ArgoCD Server Pod Status:</strong></p><ul><li>First, I checked the status of the ArgoCD server pod to ensure it was running properly. This step was essential to confirm that the pod was up and ready to accept connections.</li></ul><pre>kubectl get pods -n argocd</pre><ul><li>I verified that the ArgoCD server pod was in a Running state, which indicated that the server was ready for access.</li></ul><p>ii. <strong>Re-establish Port Forwarding:</strong></p><ul><li>Since the port forwarding session had been terminated due to the pod restart, I needed to set up port forwarding again to regain access to the ArgoCD UI.</li></ul><pre>kubectl port-forward svc/argocd-server -n argocd 8080:443</pre><ul><li>By running this command, I re-established the connection to the ArgoCD server, making the UI accessible through <a href="https://localhost:8080.">https://localhost:8080.</a></li></ul><p>iii. <strong>Log into the ArgoCD UI:</strong></p><ul><li>After re-establishing port forwarding, I opened a web browser and navigated to https://localhost:8080 to access the ArgoCD UI. With the port forwarding in place, I could now login using the admin credentials.</li></ul><p><strong>Lesson Learned:</strong> When working with Kubernetes and tools like ArgoCD, remember that port forwarding sessions need to be re-established after any pod restarts or system reboots. It’s a simple step but crucial for maintaining access to your tools.</p><h4><strong>Key Takeaways and Lessons Learned</strong></h4><ul><li><strong>Local Development Convenience</strong>: Docker Desktop makes it easy to run Kubernetes locally, providing a great environment for experimentation.</li><li><strong>Declarative Management with ArgoCD</strong>: Using ArgoCD highlighted the power of declarative application management and GitOps practices.</li><li><strong>Attention to Detail</strong>: Small configuration errors can cause significant issues, emphasizing the need for precision.</li><li><strong>Troubleshooting Skills</strong>: Developing a systematic approach to troubleshooting is crucial for resolving issues efficiently.</li><li><strong>Documentation</strong>: Keeping detailed notes of commands and configurations is invaluable for both troubleshooting and future reference.</li><li><strong>Continuous Learning</strong>: Every challenge encountered was an opportunity to deepen my understanding of Kubernetes and ArgoCD.</li></ul><p>This hands-on exploration of ArgoCD on my local machine offered valuable insights into continuous delivery practices and the benefits of GitOps. I hope this helps anyone looking to play around with ArgoCD in a local environment.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e2403676a898" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Learn Linux, Learn the Cloud.]]></title>
            <link>https://runsewe-seun.medium.com/learn-linux-learn-the-cloud-c0907352a410?source=rss-d0877f410e8b------2</link>
            <guid isPermaLink="false">https://medium.com/p/c0907352a410</guid>
            <category><![CDATA[cybersecurity]]></category>
            <category><![CDATA[cloud-computing]]></category>
            <category><![CDATA[devsecops]]></category>
            <category><![CDATA[career-advice]]></category>
            <category><![CDATA[devops]]></category>
            <dc:creator><![CDATA[Seun Runsewe]]></dc:creator>
            <pubDate>Mon, 18 Mar 2024 11:50:21 GMT</pubDate>
            <atom:updated>2024-03-18T11:50:21.932Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*QWHRdUv97NQ2TSWGZ7-eig.jpeg" /><figcaption>I have a feeling this image is not helping to make Linux more affable. That’s the job of the article so I see the dissonance here :) Trust me. Keep scrolling.</figcaption></figure><p>I missed my LPI Linux Essentials exam a few weeks ago.</p><p>Yes, I was slightly ill but that never stopped me before. In truth, I was scared I’d fail because I’d failed 4 mock exams after studying hard every day for 2 months.</p><p>Let’s talk about why I chose to put myself through this rigour right after a Master&#39;s Degree in Cybersecurity and the AWS Cloud Practitioner qualification.</p><p>In addition to Linux being a necessary component of my dissertation on Cloud Engineering, my favourite Cloud Engineering creator on TikTok, <a href="https://www.tiktok.com/@ephemeralbytes_?_t=8kmNw4jKwpK&amp;_r=1">The0x</a> (he’s awesome, go follow him), said “Learn Linux, Learn the Cloud” when I was scrolling mindlessly one day.</p><p>He’s right — it is a simple but important step in building a vibrant career in Information Security and Cloud Engineering. Think of Linux as the foundation on which the giant houses in the cloud domain are built.</p><p>So I decided. I’m getting certified.</p><h4>How did Linux become so critical to Cloud Computing Environments?</h4><p>As I mentioned before, Linux is omnipresent in the cloud. It is the backbone of many cloud infrastructures because it is flexible and compatible across different environments and cloud providers. Further, it is open to customization and scalability, making space for the different use cases you’d see in cloud services (because it is open-source).</p><p>My favourite perk to learning Linux is the cloud provider-agnostic nature. Linux allows for transferable skills across various cloud platforms and complementary tools (more on this further down), unlike MacOS and Windows. This universality makes Linux expertise incredibly valuable for professionals who want to remain versatile and adaptable in a multi-cloud ecosystem. In my case, I feel confident about navigating Azure and GCP even though I’ve only trained and certified with AWS.</p><h4>Linux Can Future-proof an IT Career</h4><p>Cloud adoption is spreading like wildfire across the globe. In the future of IT, Linux, cloud computing, and cybersecurity will intersect. A solid understanding of Linux commands and systems is critical for securing cloud environments where Linux-based servers are prevalent — most of the EC2 instances I’ve deployed on AWS are Linux servers. I would need to use tools such as iptables, fail2ban, and SELinux (I totally worked the last 3 words in to flex, nothing to be afraid of, just keep scrolling if you’re not technical), all integral to Linux. At the core, I would be managing vulnerabilities and securing cloud infrastructure.</p><h4>Cloud Engineering Career Paths Enabled by Linux Proficiency</h4><p>Linux can open some really cool doors for you in the Cloud. Here are a few ways you can shine when you have this skill:</p><ol><li>Cloud Systems Administrator: Imagine being the person who keeps things running smoothly on the cloud. In this role, you’d be the go-to person for managing cloud environments and fixing things when they go wrong. Everyone will depend on you to make their cloud environment work.</li><li>DevOps Engineer: Think of DevOps as the heartbeat of tech projects, where speed and efficiency are key — essentially, your software products don’t get to the hands of your customers without DevOps. You will need to understand Linux environments to implement CI/CD pipelines, do automation, and work with configuration management tools like Jenkins, Ansible, and Docker, which are fundamental to DevOps methodologies</li><li>Cloud Security Specialist: With cyber threats evolving alongside cloud adoption, Linux is a solid tool for cloud security specialists. In this role, you will be responsible for ensuring the safety of data from hackers and making sure that both your company and its customers can enjoy a secure digital environment.</li><li>Site Reliability Engineer (SRE): Imagine ensuring that software systems aren’t just good but outstandingly reliable. SREs have Linux in their toolkits to help create and optimise systems that stand strong, making sure that services are always there when people need them, and that cloud services meet their reliability targets.</li></ol><h4>Complementing Cloud Certifications with Linux Expertise</h4><p>Besides adding another credential to my portfolio, the LPI Linux certification complements cloud certifications, by providing a deeper understanding of the underlying systems that power cloud services.</p><p>This combination of certifications signals to recruiters and hiring managers that you can navigate multiple cloud environments and cloud engineering tools, making you more versatile and competitive in the job market.</p><h4>Will I make another attempt at certifying?</h4><p>Oh, you bet I will.</p><p>To prepare for a 12-week DevOps BootCamp, I studied Linux for 8 weeks. I’m a few weeks in and it has helped me navigate the lessons and assessments so far, but I still need more hands-on experience to feel confident about the exam. The certification would be an asset to this journey. Lifelong learning is integral to staying relevant and competitive in the ever-evolving field of cloud computing and cybersecurity.</p><p>Wish me luck? But, also, consider trying it out too?</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c0907352a410" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[IAM can be more affable and shape the future of Information Security.]]></title>
            <link>https://runsewe-seun.medium.com/identity-access-management-can-be-more-affable-and-shape-the-future-of-information-security-71ab1d418466?source=rss-d0877f410e8b------2</link>
            <guid isPermaLink="false">https://medium.com/p/71ab1d418466</guid>
            <dc:creator><![CDATA[Seun Runsewe]]></dc:creator>
            <pubDate>Thu, 24 Aug 2023 17:16:56 GMT</pubDate>
            <atom:updated>2023-12-08T16:25:55.329Z</atom:updated>
            <content:encoded><![CDATA[<p>Cybersecurity is intimidating.</p><p>People look at me like I have special powers when I say I’m rounding up my Master’s degree in the field.</p><p>As part of my coursework, I dove into a literature review on the topic, and I’m publishing it to boost the ‘affability score’ of IAM (Identity and Access Management).</p><p>Historically, its depth and complexity were reserved for scholars and professionals sifting through dense layers of information. My hope is this piece will bring IAM closer to everyone.</p><p><strong><em>Abstract — Identity Access Management (IAM) is critical for protecting organizational assets, enabling secure access to systems and data, and ensuring compliance with regulatory frameworks. However, the fast-evolving nature of IAM technologies and the increasing adoption of cloud-based solutions have led to new challenges and research areas. Future research should focus on improving the usability and effectiveness of IAM systems, addressing emerging threats such as insider attacks, and examining the impact of IAM on organizational productivity and security.</em></strong></p><p><strong><em>Keywords — Authentication, Access, Identity, Attacks</em></strong></p><p>I. INTRODUCTION</p><p>Identity access management (IAM) is a critical research area in cybersecurity that focuses on the process of identifying, authenticating, and authorising individuals or systems to access specific resources or information (Nguyen &amp; Smith, 2021). According to Arora and Dhillon (2019), IAM is essential for ensuring the confidentiality, integrity, and availability of sensitive data and systems and protecting against unauthorised access, data breaches, and cyber threats.</p><p>One of the main challenges in IAM is the need to balance security and convenience and the growing complexity of identity management in the digital world, where individuals and organisations often have multiple online identities and access various systems and networks (Ren &amp; Zhu, 2019). IAM solutions must also be able to adapt to evolving threats and changing user needs and support a wide range of authentication methods, such as passwords, biometrics, and multi-factor authentication (Ahmad et al., 2022).</p><p>To a large extent, IAM is important for an M.Sc in cybersecurity, as it is a foundational topic in cybersecurity and is relevant to many areas of the field, including network security, cloud security, and cyber risk management. A strong understanding of IAM is essential for developing and implementing effective security strategies and technologies to protect organisations and individuals from cyber attacks (Arora &amp; Dhillon, 2019). Furthermore, IAM is a key component of data privacy and compliance, as it helps to ensure that only authorised individuals have access to sensitive or personal data. It is also a critical aspect of business continuity and disaster recovery planning, as it helps to ensure that essential systems and resources can be accessed by authorised personnel in the event of an emergency or outage.</p><p>IAM is a vital research area in cybersecurity that requires a comprehensive and interdisciplinary approach to ensure the security and integrity of identity and access management systems. M.Sc in cybersecurity students who specialise in IAM will be well positioned to pursue careers in cybersecurity and help organisations and individuals protect against cyber threats and data breaches. Based on the article “Identity Management Systems: A Comparative Analysis” by Vikas Kumar and Aashish Bhardwaj and other scholarly articles on Identity access management as a research area, this paper summarises the current state of practice and research in identity access management. Beyond a summary of current research, it also considers future research areas.</p><p>II. LITERATURE REVIEW</p><p>Identity Management Systems: A Comparative Analysis, by Vikas Kumar and Aashish Bhardwaj, is a thorough review of the various existing identity management systems. The authors begin by discussing the importance of identity management in today’s digital age, highlighting the need for secure and efficient systems to manage user identities and access to resources (Kumar &amp; Bhardwaj, 2020). They then go on to provide a comparative analysis of the different types of identity management systems, including single sign-on (SSO), federation, and attribute-based access control (ABAC).</p><p>One of the key strengths of this article is its detailed and comprehensive coverage of the different identity management systems. The authors provide a clear, concise explanation of the various systems, highlighting their strengths, weaknesses, and suitability for different types of organisations. They also provide a helpful comparison of the various systems in terms of their functionality, cost, and complexity, which is helpful for those looking to choose an identity management system (Kumar &amp; Bhardwaj, 2020).</p><p>Arora and Dhillon&#39;s review of identity and access management in cloud computing is another vital contribution to the literature on identity and access management. This article focuses on the challenges and considerations of managing identity and access in cloud computing environments. The authors argue that the increasing adoption of cloud computing has led to new challenges in identity and access management, including the need to support a diverse range of users, devices, and applications, as well as the need to ensure security and compliance in a highly dynamic and distributed environment (Arora &amp; Dhillon, 2019).</p><p>One of the key insights provided by Arora and Dhillon is the importance of adopting a holistic approach to identity and access management in cloud computing. They argue that more than traditional approaches to identity management, such as single sign-on, may be needed in the cloud and that organisations need to adopt more comprehensive and flexible approaches that can adapt to the changing needs of cloud users. They also highlight the importance of leveraging new technologies, such as identity and access management as a service (IAMaaS), to support these more flexible and scalable approaches (Arora &amp; Dhillon, 2019).</p><p>Overall, both articles provide valuable insights into the challenges and considerations involved in identity and access management in the digital age. Both authors highlight the importance of adopting a flexible and comprehensive approach to identity management and provide useful insights into the systems and technologies available to support this. These articles are valuable resources for those looking to understand the landscape of identity management. They can help organisations make informed decisions about the best approaches to managing identity and access in their environments.</p><p>III. CURRENT PRACTICES IN IDENTITY ACCESS MANAGEMENT</p><p>Identity Access Management (IAM) is a crucial aspect of modern-day society, as it plays a key role in determining who has access to what resources and information. In the context of the current practices as a research area, IAM is an active and evolving field, with various approaches and technologies being used to ensure that individuals and organisations can identify themselves in a secure and reliable manner.</p><p>According to Thomas &amp; Chandrasekaran (2021), one of the main challenges facing IAM is the need to balance security and convenience. On the one hand, there is a clear need to ensure that only authorised individuals are able to access certain resources and information. At the same time, however, there is also a desire to make the process of identifying oneself as easy and convenient as possible to minimise disruption to users and organisations. This can be particularly challenging in the digital age, where the use of passwords and other traditional forms of authentication may be seen as cumbersome or inconvenient.</p><p>To address these challenges, many organisations have turned to alternative forms of identity verification, such as biometric authentication or single sign-on (SSO) systems (Kumar &amp; Bhardwaj, 2020). Biometric authentication involves using physical characteristics such as fingerprints, facial recognition, or voice recognition to verify an individual’s identity. A single login credential allows users to access multiple applications simultaneously with SSO. These approaches can offer a more convenient and secure alternative to traditional passwords, but they also come with their own set of challenges and limitations (Hamza et al., 2020).</p><p>One of the main concerns around biometric authentication is the potential for data breaches and identity theft. If biometric data is not properly protected, it can be used to impersonate an individual and gain access to sensitive resources. Additionally, biometric authentication can also be vulnerable to spoofing attacks, where an attacker uses a replica of someone’s biometric data to gain access to a system (Indu et al., 2018). To address these concerns, many organisations have implemented additional security measures, such as multi-factor authentication, which requires users to provide additional forms of verification in addition to their biometric data.</p><p>Similarly, SSO systems can also be vulnerable to security risks if not properly implemented and maintained. If an attacker can gain access to a user’s single set of login credentials, they can potentially gain access to multiple applications and systems. To mitigate these risks, organisations should ensure that they have robust security protocols in place to protect against unauthorised access and regularly update and monitor their SSO systems to identify and resolve potential vulnerabilities (Kumar &amp; Bhardwaj, 2020).</p><p>In addition to these technical challenges, social and cultural considerations need to be taken into account when implementing IAM systems. For example, the use of biometric authentication may be seen as intrusive or even discriminatory in certain communities, and organisations need to be mindful of these concerns and take steps to address them (Ren &amp; Zhu, 2019). Similarly, the use of SSO systems may raise privacy concerns if users are required to share their login credentials with multiple applications, and organisations need to be transparent about how this data is used and protected.</p><p>Overall, the current state of IAM is one of constant evolution as organisations seek to find the right balance between security and convenience in the face of constantly evolving threats and technologies (Hamza et al., 2020). While there are certainly challenges to be addressed, there is also a range of promising approaches and technologies that can help organisations and individuals identify themselves in a secure and reliable manner. Future research in this area will likely lead to even more innovative and effective solutions.</p><p>IV. THE FUTURE OF IDENTITY ACCESS MANAGEMENT</p><p>Identity access management (IAM) remains a crucial aspect of online security, as it determines who has access to what digital resources and information. As technology continues to advance, the future of IAM will likely involve significant changes in the way that users interact with digital services and information. This section explores ways in which future approaches to IAM can positively affect user experience, including biometric authentication, the adoption of decentralised identity systems, and the integration of artificial intelligence (AI) into IAM systems.</p><p>One of the key trends in the future of IAM is the use of biometric authentication methods. Biometric authentication refers to the use of biological characteristics, such as fingerprints, facial recognition, or iris scans, to verify the identity of a user. Biometric authentication has several advantages over traditional methods, such as passwords or security tokens (Eludiora &amp; Kehinde, 2021). For one, biometric authentication is more secure, as it is difficult for someone to steal or replicate a person’s biometric data. Additionally, biometric authentication is more convenient for users, as it does not require them to remember complex passwords or carry around physical tokens (Ren &amp; Zhu, 2019).</p><p>Several studies have explored the potential benefits of biometric authentication for IAM. A study published in the Journal of Cybersecurity found that biometric authentication methods, such as facial recognition and fingerprint scanners, are more effective at preventing identity fraud than traditional methods (Mohammed, 2019). Another study published in the journal Human-Computer Interaction found that users had a more positive experience with biometric authentication than with traditional methods, citing increased convenience and security as the primary reasons (Nguyen &amp; Smith, 2021)</p><p>In addition to biometric authentication, the future of IAM may also involve the adoption of decentralised identity systems. Decentralised identity systems are not controlled by a central authority but are distributed across a network of nodes (Indu et al., 2018). This adoption has several potential benefits for users, as decentralised identity systems are more resilient to attacks and less vulnerable to data breaches. Additionally, decentralised identity systems can give users more control over their personal data, as they can choose what information to share and with whom (Mohammed, 2019).</p><p>One example of a decentralised identity system is blockchain-based identity systems, which use distributed ledger technology to store and manage identity data. A study published in the Journal of Information Security and Applications explored the potential benefits of using blockchain for IAM, finding that it could increase the security and privacy of identity data and improve the efficiency of identity verification processes (Dragoş &amp; Mangiuc, 2019). Another study published in the journal Future Internet found that decentralised identity systems have the potential to enable more inclusive and equitable access to digital services and information, particularly in developing countries where traditional identity systems are often lacking (Cremer et al., 2022)</p><p>Finally, the integration of AI into IAM systems could also have a positive impact on user experience in the future. AI-powered IAM systems can use machine learning algorithms to analyse user behaviour and identify potential security threats in real time (Kumar &amp; Bhardwaj, 2020). This can help prevent identity fraud and other types of cyber-attacks and improve the accuracy of identity verification processes.</p><p>A study published in the journal Information and Computer Security explored the potential benefits of using AI for IAM, finding that it can improve the security and efficiency of identity verification processes and reduce the workload of security personnel (Zhou et al., 2019). Another study published in the journal IEEE Access found that AI-powered IAM systems can enhance the user experience by providing personalised recommendations and services based on the user’s identity and behaviour (Nguyen &amp; Smith, 2021).</p><p>In conclusion, the future of Identity Access Management is likely to involve significant changes in how identity and access management (IAM) systems are designed and implemented. In the coming years, there is likely to be a greater emphasis on using biometric and behavioural data in IAM systems and developing more advanced machine learning algorithms that can accurately identify and authenticate users. (Yahuza &amp; Khan, 2020). In addition to these technological trends, Nguyen and Smith (2021) suggested that the future of IAM is likely to be influenced by several regulatory and compliance factors. As data privacy concerns continue to grow, organisations will be under increasing pressure to implement robust IAM systems that protect sensitive data from unauthorised access. This will require the development of new policies and procedures to ensure compliance with these regulations and standards.</p><p>Overall, the future of identity and access management will likely involve significant changes and innovations in how these systems are designed and implemented. With the growing importance of data privacy and the increasing reliance on biometric and machine learning technologies, organisations must adapt and evolve their IAM systems to remain secure and compliant in an increasingly complex and dynamic digital landscape (Thomas &amp; Chandrasekaran, 2021).</p><p>V. CONCLUSION</p><p>Identity access management (IAM) has significantly influenced cybersecurity as a research area in recent years. IAM refers to the processes, policies, and technologies used to manage digital identities and the access rights of individuals to various resources and systems (Cremer et al., 2022). It is an essential component of cybersecurity as it helps to ensure that only authorised individuals have access to sensitive information and systems, protecting organisations from potential threats and breaches.</p><p>Over the years, IAM has become increasingly complex due to the proliferation of digital devices, the rise of remote work, and the increasing use of cloud-based systems. As a result, researchers have focused on developing new and innovative approaches to IAM that can effectively manage and secure access to these resources (Jolliffe &amp; Cadima, 2016). This has included the development of new authentication methods, such as biometric and multi-factor authentication, and the use of identity and access management platforms that can automate and streamline the IAM process.</p><p>In addition, IAM has also influenced the development of new technologies, such as single sign-on (SSO) and identity as a service (IDaaS), which have made it easier for organisations to manage user access and reduce the risk of unauthorised access. SSO allows users to access multiple systems and resources with a single set of credentials, while IDaaS allows organisations to outsource their IAM functions to a third-party provider (Eludiora &amp; Kehinde, 2021).</p><p>Overall, the research in IAM has significantly impacted cybersecurity as it has helped improve access management processes’ security and efficiency. It has also enabled organisations to better protect themselves from potential threats and breaches, ensuring their systems and resources’ confidentiality, integrity, and availability. As the digital landscape continues to evolve, IAM will continue to be a critical research area for cybersecurity as organisations look for ways to manage and secure access to their systems and resources effectively (Cremer et al., 2022).</p><p>VI. REFERENCES</p><p>Ahmad, S., Mehfuz, S., Mebarek-Oudina, F., &amp; Beg, J. (2022). RSM analysis based cloud access security broker: a systematic literature review. <em>Cluster Computing</em>. <a href="https://doi.org/10.1007/s10586-022-03598-z">https://doi.org/10.1007/s10586-022-03598-z</a></p><p>Arora, J., &amp; Dhillon, G. (2019). <em>A review of identity and access management in cloud computing. International Journal of Cloud Computing,.</em> 8(1), 37–57.</p><p>Choudhary, A., &amp; Dubey, A. (2020). <em>A review of identity and access management in cloud computing. International Journal of Cloud Computing</em>. 8(1), 58–69.</p><p>Cremer, F., Sheehan, B., Fortmann, M., Kia, A. N., Mullins, M., Murphy, F., &amp; Materne, S. (2022). Cyber risk and cybersecurity: a systematic review of data availability. <em>The Geneva Papers on Risk and Insurance — Issues and Practice</em>, <em>47</em>. <a href="https://doi.org/10.1057/s41288-022-00266-6">https://doi.org/10.1057/s41288-022-00266-6</a></p><p>Dragoş, M., &amp; Mangiuc. (2019). CLOUD IDENTITY AND ACCESS MANAGEMENT — A MODEL PROPOSAL. <em>Accounting and Management Information Systems</em>, <em>11</em>(3), 484–500. <a href="http://online-cig.ase.ro/RePEc/ami/articles/11_3_8.pdf">http://online-cig.ase.ro/RePEc/ami/articles/11_3_8.pdf</a></p><p>Eludiora, S., &amp; Kehinde, L. (2021). A User Identity Management Protocol for Cloud Computing Paradigm. <em>International Journal of Communications, Network and System Sciences</em>, <em>04</em>(03), 152–163. <a href="https://doi.org/10.4236/ijcns.2011.43019">https://doi.org/10.4236/ijcns.2011.43019</a></p><p>Hamza, M., Abubakar, H., &amp; Danlami, Y. (2020). Identity and Access Management System: a Web-Based Approach for an Enterprise. <em>Path of Science</em>, <em>4</em>(11), 2001–2011. <a href="https://doi.org/10.22178/pos.40-1">https://doi.org/10.22178/pos.40-1</a></p><p>Indu, I., Anand, R., &amp; Bhaskar, V. (2018). Identity and access management in cloud environment: Mechanisms and challenges. <em>Engineering Science and Technology, an International Journal</em>, <em>21</em>(4), 574–588. Sciencedirect. <a href="https://doi.org/10.1016/j.jestch.2018.05.010">https://doi.org/10.1016/j.jestch.2018.05.010</a></p><p>Jolliffe, I., &amp; Cadima, J. (2016). Principal component analysis: a review and recent developments. <em>Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences</em>, <em>374</em>(2065), 20150202. <a href="https://doi.org/10.1098/rsta.2015.0202">https://doi.org/10.1098/rsta.2015.0202</a></p><p>Kumar, V., &amp; Bhardwaj, A. (2020). Identity Management Systems. <em>International Journal of Strategic Decision Sciences</em>, <em>9</em>(1), 63–78. <a href="https://doi.org/10.4018/ijsds.2018010105">https://doi.org/10.4018/ijsds.2018010105</a></p><p>Mohammed, I. A. (2019). CLOUD IDENTITY AND ACCESS MANAGEMENT — A MODEL PROPOSAL. <em>International Journal of Innovations in Engineering Research and Technology</em>, <em>6</em>(10), 1–8. <a href="https://repo.ijiert.org/index.php/ijiert/article/view/2781">https://repo.ijiert.org/index.php/ijiert/article/view/2781</a></p><p>Nguyen, K., &amp; Smith, A. (2021). <em>Identity and Access Management in Financial Services: A Review.”</em>.</p><p>Ren, Y., &amp; Zhu, F. (2019). Identity Management and Access Control Based on Blockchain under Edge Computing for the Industrial Internet of Things. <em>Applied Sciences</em>, <em>9</em>(10), 2058. <a href="https://doi.org/10.3390/app9102058">https://doi.org/10.3390/app9102058</a></p><p>Thomas, M. V., &amp; Chandrasekaran, K. (2021). <em>Identity and Access Management in the Cloud Computing Environments</em>. Identity Theft: Breakthroughs in Research and Practice. <a href="https://www.igi-global.com/chapter/identity-and-access-management-in-the-cloud-computing-environments/167219">https://www.igi-global.com/chapter/identity-and-access-management-in-the-cloud-computing-environments/167219</a></p><p>Yahuza, M., &amp; Khan, S. (2020). Systematic Review on Security and Privacy Requirements in Edge Computing: State of the Art and Future Research Opportunities. <em>IEEE Access</em>, <em>8</em>, 76541. <a href="https://www.academia.edu/51172195/Systematic_Review_on_Security_and_Privacy_Requirements_in_Edge_Computing_State_of_the_Art_and_Future_Research_Opportunities">https://www.academia.edu/51172195/Systematic_Review_on_Security_and_Privacy_Requirements_in_Edge_Computing_State_of_the_Art_and_Future_Research_Opportunities</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=71ab1d418466" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Inclusion is Product Leadership]]></title>
            <link>https://runsewe-seun.medium.com/inclusion-is-product-leadership-f069e602613b?source=rss-d0877f410e8b------2</link>
            <guid isPermaLink="false">https://medium.com/p/f069e602613b</guid>
            <category><![CDATA[value-creation]]></category>
            <category><![CDATA[product-management]]></category>
            <category><![CDATA[stakeholder-management]]></category>
            <dc:creator><![CDATA[Seun Runsewe]]></dc:creator>
            <pubDate>Wed, 04 May 2022 11:51:22 GMT</pubDate>
            <atom:updated>2022-05-04T12:26:00.231Z</atom:updated>
            <content:encoded><![CDATA[<p>From the outside, it looks like Product people have these grand ideas and boom, a thriving tech business is born. The big idea around Product Management is working with your team to create value for a set of users i.e the market.</p><p>But let’s talk about value first.</p><p>Value is multi-dimensional because there are many components to it — there are many things that can make something perceived as valuable to multiple people. Broadly, value can be:</p><ol><li>Functional</li><li>social</li><li>emotional or</li><li>contextual (specific to a situation).</li></ol><p>The products we build at <a href="https://chippercash.com/">Chipper Cash</a> bring a mix of these values to the 4m+ users across 8 countries who use the product. For example, some users in Uganda can raise their social currency in their circles because they now have dollar investments, diaspora users in the UK and US would derive the emotional value of staying connected to home because Chipper has given them the tool to send money home in Africa to their loved ones at the very best rates.</p><p>I, however, have a bias toward emotional value so I’m going to paint a picture of the value we’ve created from an impact perspective. Here’s a success story I love — Edwina is a middle-aged Ugandan woman who lives in Kampala, Uganda who listens to Afrobeats and downloaded the Chipper Cash app thanks to Burna Boy’s giveaway in 2020. She’s gone on to be an active transactor for our crypto and global micro-investment stocks — she essentially can buy global shares like Tesla and Apple for as little as $1 (UGX 3,524). Before Chipper, she’d need to fork out about $3,000 to buy a share in Amazon, making it harder for her to make the right money decisions to improve her financial life. Further, she would not have had access to the tools to buy and sell cryptocurrency on a small scale.</p><p>In the process of adopting these use cases on the application, Chipper is:</p><ol><li>Educating her on these investment options — fractional stocks and cryptocurrencies</li><li>Making the tools available to her to execute those transactions</li><li>Helping her build the muscle of making these transactions a part of her daily life by reducing the minimum amount to what she would consider bite size.</li></ol><p>Giving her these things will ultimately give Edwina financial control and agency so she can nurse ideas like “I’d like my kids to go to better schools” or “it’s time to expand my business”.</p><p>That’s the value. We’re building for Edwina.</p><p>But how do we do it?</p><p><strong>The Inclusion Dance</strong></p><p>There are distinct phases in the product development cycle and for each stage, there are key collaborators. Typically, the start of the cycle is <em>ideation</em> where product people talk to customers, study the market and coin a Product Requirements <em>Document </em>(PRD) to define the problem and solution to get the ball rolling on <em>building</em> the product. At this point, the engineering team is looped in to build and then everything else is figured out after.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*bswbpjXqMa8rByA5BHNbcw.png" /><figcaption>High-level product development cycle :)</figcaption></figure><p>I’d advocate for doing things a little differently.</p><p>Drive inclusion.</p><p>When you ideate, include data analysts and UX researchers. When you document, the Risk, Engineering, Design, Operations and Finance teams should be of an open, continued conversation. Further, in the time leading up to launch (as the product is being built), try to pull in Partnerships/ Commercial, Marketing/ Growth, and Customer Support teams into the dance party. An elegant way to keep a cross-functional conversation going is having a Slack channel where you host these conversations in public, outside of team meetings.</p><p>Think of it as you’re co-creating your PRD with these teams internally so you can <em>get buy-in</em> quickly, <em>reduce friction</em> and <em>ship</em> out a valuable product in as little time as possible. As you co-create, this is a great opportunity to advocate for the user and paint the picture of what value means to the user to your team.</p><p>I have a laundry list of boxes to be ticked when I’m rolling out a feature or product:</p><ol><li>Provide go-live acceptance criteria — Engineering Team</li><li>Legal team — to make sure I’m not about to run the business into the ground or get us in too much trouble :)</li><li>Business process changes — Administrative/ Operations Team</li><li>Communication Activities (Content &amp; Messaging Plan) — Marketing Team</li><li>Technical Support Process — Technical Support Team/Engineering Team</li><li>Customer Support (Training and Test-Run) — Customer Success Team</li><li>Support channel set-up and SLA with escalation matrix from partners — Partnerships Team</li><li>Finance and Risk approval (settlement and reconciliation, portfolio management process with partners) — Finance Operations and Risk Teams</li><li>Data infrastructure to monitor metrics (Admin Dashboard) — Data Team</li><li>Schedule for go-live — Product, Mobile Dev and Operations and Support Teams</li><li>Set up feedback collection mechanism — Customer Success and Data Teams</li></ol><p>Not every organisation is set up the same so feel free to tweak this list to suit the realities in your organisation.</p><p><strong>Delivering value through inclusion</strong></p><p>Hosting this inclusion dance is a neat way to build credibility within your team. As you engage your team early on in your process, you also teach the team to be product thinkers which is immensely valuable to you and the business long term. Further, you’re better able to play shepherd and be a part of decisions made in the later stages of the product life cycle because you’re in better sync with the cross-functional team.</p><p>And finally, my favourite one, you build muscle around being a charmer :). It is a great opportunity to build and exercise the soft skills product management demands. You’d juggle relationships and expectations better.</p><p>In the end, Edwina is better off for it — she’s getting value.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f069e602613b" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[What My Intern from Harvard Taught Me about Leadership]]></title>
            <link>https://runsewe-seun.medium.com/what-my-intern-from-harvard-taught-me-about-leadership-734f3aed0594?source=rss-d0877f410e8b------2</link>
            <guid isPermaLink="false">https://medium.com/p/734f3aed0594</guid>
            <category><![CDATA[intern]]></category>
            <category><![CDATA[mentoring]]></category>
            <category><![CDATA[leadership]]></category>
            <dc:creator><![CDATA[Seun Runsewe]]></dc:creator>
            <pubDate>Fri, 11 Dec 2020 08:08:51 GMT</pubDate>
            <atom:updated>2020-12-11T09:08:12.924Z</atom:updated>
            <content:encoded><![CDATA[<p>A few months ago, a student from the Harvard MBA program reached out to chat about fintech in Nigeria, and the possibility of getting some experience in the field for her summer internship.</p><p>According to her, Nigeria is thought to have the sexiest fintech ecosystem. I’d take that.</p><p><a href="https://gh.linkedin.com/in/abena-anima-nyantekyi-owusu-1a64185b">Abena</a> and I spent two hours discussing her background in healthcare, her keen interest in fintech, and what it was like to lead a fintech business in Nigeria. She went on to spend three months working with me remotely from Ghana to flesh out processes for Research as a standalone function of the Product Team at OPay.</p><p>On a personal note, I was grateful for a fresh mind on the team with whom I could discuss self and team audits. At the end of her internship, we had a 360-review conversation about my leadership style and I found her feedback interesting. I love getting feedback and I will share some of it, as well as what I learned from mentoring Abena.</p><h3>Feedback/Learning Points</h3><ul><li><strong>I clearly have an affinity for women:</strong> I was so proud of this one! I cheer a little louder when the women on the team make great contributions, I push for them to speak in public, and I amplify the perception of their value. I love that! I made a mental note to be more deliberate about curating a work environment where the men on my team feel psychologically safe enough to bring their best selves to the table.</li><li><strong>I pass the mic, my team must be seen and heard:</strong> The Product team was Seun, <a href="https://www.linkedin.com/in/gbotosho-afeez8b6877146/">Afeez</a>, <a href="https://ng.linkedin.com/in/omojonegedu">Omojo</a>, <a href="https://ng.linkedin.com/in/adebukolami-agboola-76ba6120">Bukolami</a>, <a href="https://ng.linkedin.com/in/kamaldeen-product-manager">Kamal</a>, <a href="https://ng.linkedin.com/in/abiodun-bada-51431a62">Bada</a> and <a href="https://ng.linkedin.com/in/olayinka-odetola">Yinka</a> — not allowing anyone address us as Seun and everybody else. I ensure they’re acknowledged by name, shine the spotlight on what they’re great at, make sure they own what is theirs, and appreciate them for the work that they do. The gold stars are theirs, not mine.</li><li><strong>When it comes to communication, I connect better in one-on-one conversations: </strong>Naturally, I leaned into that. While this helped me better influence each member of the team, the collective conversations suffered. More frequent group conversations would have served the team better because people would receive the same information at the same time, improving the team’s concerted effort. It’s super important to lead the collective, not just individuals. It promotes connectedness in the team.</li><li><strong>I love teaching:</strong> I didn’t even realise how much I loved it until I’d gathered experience and was in the position to lead a team. Now, I even teach outside of work (at universities) and speak at conferences. I feel responsible for what my people know and I take advantage of teachable moments — not in a threatening or condescending manner but in a way that sells the idea that I’m invested in my team’s learning.</li><li><strong>Short feedback loops</strong>: Our feedback system would benefit from shorter feedback loops, instead of focusing only on hitting our goals — or the standard half-year reviews. I’d take the end of a specific project like a new feature being shipped, as an opportunity to give feedback. The long-term benefits of an improved system cannot be overemphasised, and we get 1% better every time we use our muscles.</li><li><strong>A charming personality helps a ton!:</strong> The team was very happy to run in whatever direction I articulate because they liked the leader. I should say here that it’s important to strike a balance between loved boss and a tough boss. You’re not Ice Cream.</li><li><strong>I own the energy in the room:</strong> Afeez on the team asked me once, “Boss, why are you always happy?” People feed off my energy so I have a responsibility to curate the enabling environment for them to do their best work and feel safe to talk about grievances within and outside the team.</li><li><strong>I self-developed in front of them:</strong> I am not ashamed of not knowing something and publicly asking to be educated on the more technical details of our work — <a href="https://medium.com/swlh/what-ive-learned-diving-into-api-specifications-e7d4074e4b0c">APIs</a> for example. It helped the team to lean more on each other’s strengths and this improves our know-how as a collective. Our team was a safe place where people could ask for help while fixing their mistakes and strengthening their weak areas.</li></ul><p>On a final note, if you’re going to lead, you must be willing to:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*R5-Z9Yg1N5wv8wOFTuN2dw.png" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=734f3aed0594" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[What I’ve Learned Diving Into API Specifications]]></title>
            <link>https://medium.com/swlh/what-ive-learned-diving-into-api-specifications-e7d4074e4b0c?source=rss-d0877f410e8b------2</link>
            <guid isPermaLink="false">https://medium.com/p/e7d4074e4b0c</guid>
            <category><![CDATA[engineering]]></category>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[api-specifications]]></category>
            <category><![CDATA[product-management]]></category>
            <category><![CDATA[json-web-token]]></category>
            <dc:creator><![CDATA[Seun Runsewe]]></dc:creator>
            <pubDate>Mon, 10 Aug 2020 13:11:02 GMT</pubDate>
            <atom:updated>2020-08-26T13:39:38.890Z</atom:updated>
            <content:encoded><![CDATA[<p>The ultimate goal of product management is to build a product that incrementally meets the needs of users. If you can successfully create a loved product that meets the needs of users, you’re creating value and will ultimately capture value.</p><p>It is widely accepted that Product lies in between Business, Tech and Design.</p><p>I lead a product organisation at OPay that is structured around:</p><ol><li>Research</li><li>Core Product (merchant and internal tools, mobile and technical integrations)</li><li>Design (UI/UX)</li></ol><p>My background and experience have given me breadth but I didn’t have as much depth on the tech side of things. Like I did at Switch, I would give high level thought leadership on desired functionality and API requirements, then someone else would flesh it out with all the tech speak.</p><p>However, in the last few weeks, I decided to dive into API specifications and data manipulation (hello <a href="https://twitter.com/thesqlbabe">SQL Babe</a>) to build my muscle there.</p><p>I’m going to share a bit about my API specs experience so far. First, I’d send a shout out to the best teacher <a href="https://www.linkedin.com/in/adebukolami-agboola-76ba6120/">Bukolami Agboola</a>. I feel fortunate that she’s on my team and a fantastic teacher.</p><h3>What are APIs in the First Place?</h3><p>API is an acronym, and it stands for “Application Programming Interface.” Essentially, products run on APIs. APIs are what make products work behind the scenes — they lend function to your product.</p><p>There are multiple types of APIs- internal APIs, public APIs and partner APIs. I’m going to share more about partner APIs in this article because the use case we will explore involves a 3rd party betting partner for OPay.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/1*oWOi1UGonSa4fPMSMawMcw.png" /></figure><p>On OPay today, users can fund their betting wallets to place bets on sporting events from the OPay app. If a user wanted to fund their Ice Cream Bet account, they’d have to provide their User ID and specify the amount they want to fund Ice Cream Bet.</p><p>For this transaction to successfully take place, OPay’s Systems and Ice Cream Bet’s APIs have to successfully interact.</p><p>Think of it like a plug (OPay) and socket (Ice Cream Bet). They need to interact for electricity to flow (Ice Cream Bet getting funded from OPay).</p><p>Does that make sense?</p><h3>What do Product Managers have to do with API Specs?</h3><p>Valuable APIs provide valuable functions to the user. Using an understanding of what’s valuable to the user, the product manager will help the engineering team know what to build when it comes to function (for internal and external integration use cases).</p><p>An API Specification Document explains what an API is meant to be able to do and how it is meant to behave.</p><p>At a high level, the API Specification will have:</p><ol><li>The Use Case</li><li>The Detailed Flow Description</li><li>The Parameters</li><li>The Environments</li><li>The Error Structure</li></ol><h3>The Use Case</h3><p>The Use Case describes the usage scenario or the value we are trying to create for the user. Let’s use the fund your Ice Cream Bet account on OPay use case (the example from before). In that user interaction, the user is meant to enter their ID and Amount to complete the transaction, it looks pretty simple, yes?</p><p>No.</p><p>On the backend, you need to understand what APIs (think of them as actors) that need to interact for this transaction to happen:</p><ul><li><strong>The OPay System</strong>— to send information from the OPay Android or iOS app where the user is entering information to Ice Cream Bet’s API.</li><li><strong>The Ice Cream Bet User ID Lookup API</strong> — to validate the user ID provided by the user on Ice Cream Bet systems. We can’t have users funding accounts that don’t exist in the Ice Cream Bet’s business.</li><li><strong>The Ice Cream Bet Deposit API</strong> — to allow the funds to move from the user’s OPay account to his validated betting account. Here we can also decide what the user would need to provide to authenticate the transaction.</li><li><strong>Transaction Status API</strong> — to check the status of the transaction for both parties to confirm that a transaction has successfully happened.</li></ul><h3>The Detailed Sequence of Events</h3><p>This is just a fancy way of describing how the actors in the use case would interact to provide the value to the user. These are the technical events that will happen between OPay and the Ice Cream Bet API to allow a user to fund his betting account.</p><p>I’ve come to learn this bit is just a sequence of calls and responses.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/938/0*xNkeylo_F4XyHstm.png" /></figure><p>For the APIs of two businesses to interact, there are a couple of things we need to get out of the way:</p><ul><li>the first is an authentication Protocol to grant access for the APIs to safely exchange information — we use the <a href="https://oauth.net/2/">OAuth 2.0</a> for this</li><li>the second thing to consider is a way to authenticate and authorize API calls and we use <a href="https://jwt.io/introduction/">JWT</a> for this.</li></ul><p>Back to our sequence of events, this interaction is initiated by <strong>The OPay System</strong> by passing an access token + the User ID entered on the app by the user to <strong>The Ice Cream Bet User ID</strong> <strong>Lookup API</strong></p><p>then,</p><p>The <strong>Ice Cream Bet User ID Lookup API</strong> sends a response to confirm or deny the User ID <strong>The OPay System </strong>passed</p><p>then,</p><p><strong>The OPay System </strong>calls <strong>The Ice Cream Bet Deposit API</strong> to place an order for a deposit transaction. With the access token validating our access rights, OPay sends the validated User ID, Transaction Amount and OPay generated transaction reference</p><p>and</p><p><strong>The Ice Cream Bet Deposit API </strong>sends a response to <strong>The OPay System </strong>to inform us of a successful deposit (or otherwise).</p><p>If the status of a transaction changes, we use <a href="https://documentation.operapay.com/docs/api_guide/api_guide_webhooks/">webhooks</a> to communicate this between OPay and the Ice Cream Bet’s APIs.</p><h3>The Parameters</h3><p>Parameters are simply the pieces of information that need to be exchanged for the transaction to happen. In the flow I’ve described above, I mentioned User ID, transaction amount, transaction reference. The parameters you define influence the response you will get from the API you’re calling.</p><p>For example, if T<strong>he OPay System </strong>sends a <em>User ID</em> to <strong>The Ice Cream Bet User ID lookup API, </strong>the expected response is to confirm whether or not the User ID exists on the Ice Cream Bet’s business.</p><p>Think of them as the building blocks for a transaction. For each step of a transaction, it helps to define what parameters should be shared.</p><h3>The Environments</h3><p>Integrations between two external APIs happen in <em>sandbox</em> environments. Let’s think of them as playgrounds for make-believe transactions. When the integration is complete and tested, we move the integration to the <em>live</em> environment where real-life transactions can happen. Let’s think of this migration as the debutante ball.</p><p>Product and Engineering teams use these two environments to make sure APIs function as they should when implemented before they are used on the product.</p><h3>Error Structures</h3><p>An error is an event that happens when something isn’t quite right. If I pass a User ID that doesn’t exist, an error message should communicate that to me.</p><p>Here is an example:</p><p>“User does not exist”</p><p>So you know to check the User ID you’re sending.</p><p>Error codes and messages tell where you’re wrong and point you in the right direction. There are standard <a href="https://httpstatuses.com/">HTTP status codes</a> but there are more comprehensive error codes that account for whatever could go wrong and communicates that to help both parties resolve any problems with integration.</p><h3>Final thoughts</h3><p>This is pretty much what I have learned so far but it’s is only scratching the surface. I’m going to spend the next few weeks digging deeper and playing around with a lot more use cases.</p><p>The main task for the product manager or leader is to understand how to communicate what is valuable to the user, to the engineering teams in API specification documents.</p><p>I’m excited about building this muscle because APIs allow for collaboration with partners to create new value, drive customer acquisition and of course, bring revenue to the business.</p><p>I should add that I’d get brownie points with the Engineering team.</p><p>Listen, I dig it!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e7d4074e4b0c" width="1" height="1" alt=""><hr><p><a href="https://medium.com/swlh/what-ive-learned-diving-into-api-specifications-e7d4074e4b0c">What I’ve Learned Diving Into API Specifications</a> was originally published in <a href="https://medium.com/swlh">The Startup</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>