<?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 Afkham Azeez on Medium]]></title>
        <description><![CDATA[Stories by Afkham Azeez on Medium]]></description>
        <link>https://medium.com/@afkham-azeez?source=rss-d32ea2ba6958------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*zU1QfLdZngvaLcJnACzq_Q.jpeg</url>
            <title>Stories by Afkham Azeez on Medium</title>
            <link>https://medium.com/@afkham-azeez?source=rss-d32ea2ba6958------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sat, 21 Nov 2026 14:09:52 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@afkham-azeez/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[Vibe coding Cockpit-FT891]]></title>
            <link>https://medium.com/azeezs-notes/vibe-coding-cockpit-ft891-654e948a5551?source=rss-d32ea2ba6958------2</link>
            <guid isPermaLink="false">https://medium.com/p/654e948a5551</guid>
            <category><![CDATA[vibe-coding]]></category>
            <dc:creator><![CDATA[Afkham Azeez]]></dc:creator>
            <pubDate>Fri, 17 Jul 2026 09:18:01 GMT</pubDate>
            <atom:updated>2026-07-17T12:49:51.608Z</atom:updated>
            <content:encoded><![CDATA[<p>I’m writing this post, cruising at 40,000 feet using a free Starlink Internet connection! How cool is that! Unimaginable just a few years ago. Anyway, back to the topic…</p><p>When I first heard the term vibe coding a few months ago, just like any other programmer, I was skeptical and dismissed it as some sort of joke. Never would I have believed if someone said that I’d become a vibe coding practitioner and advocate.</p><p>So what’s all the fuss about? When it comes to software architecture and development, you would first start with a problem and flesh out the requirements. Typically the next steps would be building a prototype or architectural thin slice, to try out the hypothesis, course correct and iterate, before building an MVP, then the first GA release and so on. What if we could start by describing the problem in natural language and coding AI tools can then create the code for you? Now we are in vibe coding territory. Actually modern vibe coding tools such as Claude code can do much more than just generate code from natural language. You could have at your command an expert team consisting of researchers, business analysts, architects, developers, reviewers, programming language specialists, QA, SRE, cloud specialists and more. In a typical project, you would assemble a team, and take your time articulating the problem and co-developing the software requirements specification (SRS), developing sprint plans, all while vibing with the AI tools.</p><p>As Head of SRE at WSO2, my immediate attention turned toward building automation to make our proceses more efficient. Eliminating the boring aspects of incident management and problem management were at the top of my priorities. I have built a toolkit called RIA (Reliability Intelligence Agents) This post is not about that, that’s for another time. This post is about ham radio, one of my many hobbies.</p><p>One lazy Sunday morning, as I lay in bed just thinking, I was wondering how I could improve the usability of my Yaesu FT-891 transceiver. This is a compact portable radio with many compromises to reduce the form factor. Inevitably this meant having fewer physical buttons and the functionality is hidden behind several layers of menus which results in some of the worst UX. You will often see old hams grumbling about it. Larger base stations come with better displays and physical buttons for all the major features.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/764/1*0m_acpiJc2MQOkF9XfNMZA.jpeg" /><figcaption>FT 891</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/588/1*EOtrI8DTd0DhBBAuiHFeIg.jpeg" /><figcaption>FTDX-1200</figcaption></figure><p>Less buttons and more touch screen menus doesn’t mean better… Chinese EVs… wink wink</p><p>What if we can have the best of both worlds — small portable radio with a computer driven rich UI? This thought resulted in my app which I ended up naming Cockpit FT-891.</p><p>I opened Claude code and started writing the problem statement. One important thing we learned from Sanjiva is that before jumping into the solution, try describing the problem in as much detail as possible. It reduces lots of problems down the line. So I spent time articulating the problem. The good thing with vibe coding is that you can describe the problem using many media including hand drawn images and screenshots, not just text. I remember giving Claude an image of an FTDX-10 and saying that I want to build a UI that mimicks that radio, but works with the FT 891. I spawned off parallel agents to do deep research on how to communicate and control the FT-891. These agents went and found the user manual and advanced manual, and even went through user group and forum posts. When I was happy that we had enough information to write down the SRS for the first version, I got Claude to do that. This was an interactive session where I asked Claude to clarify things with me without simply assuming everything. Next using the PM skills, we developed a sprint plan. I left the technology choice to Claude and he chose to develop this as an Elecron app. I went ahead with this choice and in record time, we had something working and I was able.to control some of the core features on my radio!</p><p>Over the next few weeks the app evolved, adding more features to provide a one stop shop for the FT-891, including digital modes, logbook, customizable UI themes and so on. But this lead to a different problem. The Electron app became quite heavy and bloated. The distribution size was 450+ MB and it was one of the apps consuming most of the resources on my computer. It is at that moment that a bulb lit up in my mind! What if we completely rewrote the entire thing from scratch in C++. I’m not expert in C++ but I have done some programming with it in University as well as when I was developing IoT stuff many years ago. My C++ skills were rusty at best. But I didn’t have to worry. I had all the experts at my service! Let’s vibe boys! Since we already had a fully functioning app and SRS as reference, the big guns were deployed and the porting begun. The result — 58mb distribution and using a fraction of the resources of the Electron app, and better responsive app overall. My respect for C++ has immensely increased.</p><p>Anyway that’s the story of one of my vibe coding experiences and I’m still learning this art. Enjoy the screenshots from my app.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*8j4V6Ia10sHbdSofJqw_6w.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*6eydBeJRDR4jhovYO4-g9w.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ApeRqyy7v3gKde7ivOzbkw.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*BoQw676fV_RATTLNHrbHNg.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/718/1*fVIn-JBZj1V_8ar6U0qHog.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/374/1*_hM__0zmNn6jT_IrVCgFmw.jpeg" /></figure><p>Vibe coding is real and here to stay. If you are a skeptic like me, just try it once and experience it for yourself.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=654e948a5551" width="1" height="1" alt=""><hr><p><a href="https://medium.com/azeezs-notes/vibe-coding-cockpit-ft891-654e948a5551">Vibe coding Cockpit-FT891</a> was originally published in <a href="https://medium.com/azeezs-notes">Azeez’s Notes</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Thank you Sanjiva!]]></title>
            <link>https://afkham-azeez.medium.com/thank-you-sanjiva-5c0c950c5fce?source=rss-d32ea2ba6958------2</link>
            <guid isPermaLink="false">https://medium.com/p/5c0c950c5fce</guid>
            <category><![CDATA[sanjiva]]></category>
            <category><![CDATA[wso2]]></category>
            <dc:creator><![CDATA[Afkham Azeez]]></dc:creator>
            <pubDate>Mon, 01 Jun 2026 06:16:24 GMT</pubDate>
            <atom:updated>2026-06-02T04:18:52.754Z</atom:updated>
            <content:encoded><![CDATA[<p>Today is a sad day since yesterday was officially the last day at WSO2 for Sanjiva. Everybody knows about his technical vision and brilliance, and people who’ve worked with him know the great human qualities in him as well as those qualities he instilled in anyone who had worked with him.</p><p>Needless to say, the lives of many of us have been changed in ways we couldn’t have imagined through Sanjiva deciding to come back to Sri Lanka, leaving his comfortable life in the US and starting WSO2. Readers of my blog may have already read about it in <a href="https://afkham-azeez.medium.com/the-journey-of-a-builder-b8fbadc18daf">Chronicles of a Builder</a> and <a href="https://medium.com/@afkham-azeez/a-tale-of-2-decades-8d05f0849ea8">A tale of two decades</a>.</p><p>I am at a loss for words to thank and appreciate Sanjiva, so I’ll post a story from almost 2 decades ago that only a few people know, and post a few random pics that I clicked over the years.</p><p>First the story… many of you at WSO2 may not know the <strong>origins of the WSO2 Club chat room</strong>. Almost 20 years ago, a bunch of us started a Yahoo chat group called “Ape Fun Talks” (AFT) — අපේ Fun Talks, not ape :D .This group had only a selected few “boys” and soon it turned into an office gossip chat room. One thing led to another, and it soon degraded into very ugly gossip, targeting individuals at times. Sanjiva got to know about this after a while and some of us who were still working at the company were called in for an important meeting where we got an earful from him and got taught a life lesson. It took us a while to realize how immature we were and what we did was contrary to WSO2 and human values. Many of us were expecting to get fired, but instead he made it an opportunity to teach us a life lesson. The AFT Yahoo chat was shutdown, and having realized the need to have a channel for informal or fun conversations, he suggested starting an official chat room, which ended up being WSO2 club and continues to this date, comprising all employees and interns, without any discrimination. This was a valuable lesson and we’ve not repeated such mistakes.</p><p>Now for the photos in no particular order,</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*5iLYshQ6S59mefmxopxdUg.png" /><figcaption>2006 — WSO2 Tungsten 1.0 launch.</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*fZaViAiCJAfgsD8c3VvFVw.png" /><figcaption>2014 — Planning an RFID + Raspberry Pi based solution for WSO2Con attendee registration and session attendance</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*VSb5SNpwx1garL0el-kIvw.png" /><figcaption>2022 — Planning meeting at Shangri La, Hambantota</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*GrJsC77DIEmEoZjOTc9ygQ.png" /><figcaption>2022 — Dubai office opening</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*1ppdvpnSB-xCs7_7DAQJWQ.png" /><figcaption>2023 — Prof. Frank Leymann inaugrated as the first WSO2 Fellow</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*BX9aJtqLy_wujMaHP7EVQg.png" /><figcaption>2024 — January Architecture meeting</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Ok69yYR3URGMtaDGFH3k1g.png" /><figcaption>2024 — WSO2Con US. Reunion of some of the earliest employees of WSO2 including Employee # 1, Ruchith</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_49jyAkesB5mpmEnHwtUlA.png" /><figcaption>2025 — January architecture meeting. Sanjiva trying to peep at Mifan who was behind the whiteboard :D</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*lL8d2Y8edlIB_JuHupJ25w.png" /><figcaption>2024 — I saw this magazine while at the Bank</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*itL2Ry9fXGxUrfx6jxcgHA.png" /><figcaption>2021 — Awards night 2020, conducted with a small crowd amidst COVID.</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*trVLrgzJ__uWPFXQ" /><figcaption>2026 — Awards night 2025 where I completed 20 years. Never imagined that Sanjiva would leave before me.</figcaption></figure><p>I can’t imagine a WSO2 without Sanjiva, but the journey has to continue, and the tribute to him would be to carry the flag with the company, leadership and values he built and instilled.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5c0c950c5fce" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[New Operational Sympathy publication]]></title>
            <link>https://afkham-azeez.medium.com/new-operational-sympathy-publication-b65702932ade?source=rss-d32ea2ba6958------2</link>
            <guid isPermaLink="false">https://medium.com/p/b65702932ade</guid>
            <category><![CDATA[software-architecture]]></category>
            <category><![CDATA[sre]]></category>
            <category><![CDATA[operational-sympathy]]></category>
            <dc:creator><![CDATA[Afkham Azeez]]></dc:creator>
            <pubDate>Wed, 18 Feb 2026 09:19:48 GMT</pubDate>
            <atom:updated>2026-02-18T09:19:48.391Z</atom:updated>
            <content:encoded><![CDATA[<p>A group of us have started a new Medium publication on Operational Sympathy <a href="https://medium.com/operational-sympathy/welcome-1e3bec6e7787">https://medium.com/operational-sympathy/welcome-1e3bec6e7787</a>. If you are an architect, developer, SRE or devops person or are simply interested in this area, please subscribe to this publication. We have a interesting stream of articles lined up. Please also share your thoughts and we are open to discussion on this emerging area.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b65702932ade" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Welcome!]]></title>
            <link>https://medium.com/operational-sympathy/welcome-1e3bec6e7787?source=rss-d32ea2ba6958------2</link>
            <guid isPermaLink="false">https://medium.com/p/1e3bec6e7787</guid>
            <category><![CDATA[sre]]></category>
            <category><![CDATA[software-architecture]]></category>
            <category><![CDATA[operational-sympathy]]></category>
            <dc:creator><![CDATA[Afkham Azeez]]></dc:creator>
            <pubDate>Wed, 18 Feb 2026 08:12:20 GMT</pubDate>
            <atom:updated>2026-02-18T08:12:20.740Z</atom:updated>
            <content:encoded><![CDATA[<p>Welcome to the Operational Sympathy publication. We are a team of writers who are passionate about the subject and will share our thoughts and concepts. The blog on <a href="https://afkham-azeez.medium.com/operational-sympathy-8a9c5dc26b5a">Operational Sympathy</a> introduced the concept and you can read more about it there.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*63NAr0SjHMG5SzeILeurlA.png" /></figure><h3>Definition</h3><blockquote>Operational Sympathy is a mindset and practice where software architects, designers, and developers intentionally design how their systems behave in production under load, during failures, and under security threats by planning failure modes, graceful degradation, built-in observability, and clear operational run books that enable early detection and fast recovery.</blockquote><p>The thinking would be, if you the developer were the person to be woken at 2 AM by an alert, what would you do to make your life easy so that the incident could be resolved with the least effort and shortest possible time.</p><p>If you are a software architect, developer, SRE or devops person, please subscribe to this publication to get notified about articles on this area.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1e3bec6e7787" width="1" height="1" alt=""><hr><p><a href="https://medium.com/operational-sympathy/welcome-1e3bec6e7787">Welcome!</a> was originally published in <a href="https://medium.com/operational-sympathy">operational-sympathy</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Operational Sympathy]]></title>
            <link>https://afkham-azeez.medium.com/operational-sympathy-8a9c5dc26b5a?source=rss-d32ea2ba6958------2</link>
            <guid isPermaLink="false">https://medium.com/p/8a9c5dc26b5a</guid>
            <category><![CDATA[formula-1]]></category>
            <category><![CDATA[sre]]></category>
            <category><![CDATA[wso2]]></category>
            <category><![CDATA[architec]]></category>
            <category><![CDATA[mechanical-sympathy]]></category>
            <dc:creator><![CDATA[Afkham Azeez]]></dc:creator>
            <pubDate>Wed, 28 Jan 2026 10:01:59 GMT</pubDate>
            <atom:updated>2026-01-28T10:01:59.581Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*iZDB1Pdzuzt64GO8Os6Mmg.png" /></figure><p>At the recently concluded <a href="https://www.youtube.com/@WSO2official"><strong>WSO2</strong></a><strong> Tech Conference</strong>, I gave a talk titled <strong>“</strong><a href="https://www.youtube.com/watch?v=hfADSpkhM90"><strong>Code for Cloud</strong></a><strong>”</strong>, built around a simple but uncomfortable truth: <strong>the cloud punishes complacency</strong>. Many of the assumptions that held true in traditional systems quietly collapse under cloud scale. Latency is no longer predictable, failures are not exceptional events, and complexity compounds faster than intuition. In the talk, I challenged a mindset that still persists across software teams: that <em>Non-Functional Requirements</em> can be deferred, optimized later, or delegated entirely to infrastructure and teams managing it. In reality, cost, resiliency, security, scalability, and operability are not “nice to have” attributes in cloud-native systems; they are hard design constraints that shape architecture, design decisions, and every feature that is developed</p><p>For a significant part of my own career, working in software architecture, design, and development, I was not immune to this way of thinking. My primary focus was on delivering functional correctness. i.e. making sure features worked as intended, while assuming that running production-grade systems was someone else’s responsibility. Today, standing on the other side of that fence as part of the WSO2 SRE team, where my teams and I are accountable for keeping large-scale production systems alive, stable, and cost-effective, I see the consequences of those earlier assumptions clearly. Ignoring NFRs does not merely accumulate technical debt; it manifests as outages, runaway cloud bills, fragile systems, and operational firefighting.</p><p>This article builds on that talk to argue for a mindset shift toward <strong>operational sympathy</strong> — designing and writing software with an explicit awareness of how it will behave, fail, and be operated in the real world. I haven’t found many references to the term operational sympathy but after my talk, I thought that this is a concept that should be popularized like mechanical sympathy.</p><h3>Mechanical sympathy</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*LF3dUvOAtskIsEIK_ez8wQ.jpeg" /></figure><p><a href="https://github.com/bugzmanov/mechanical_sympathy">This term</a> as applicable to software engineering, is attributed to <a href="https://www.blogger.com/profile/15893849163924476586">Martin Thompson</a>, who popularized it in the software world, drawing inspiration from <a href="https://en.wikipedia.org/wiki/Jackie_Stewart">Jackie Stewart</a>, the Formula One driver who originally used the phrase to describe understanding and respecting how a machine behaves to get the best performance without breaking it. You can’t win races by constantly pushing your race car beyonds its design parameters.</p><h3>What is operational sympathy?</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*VVQazqtNszwbOZLFTihzXg.png" /></figure><p>Operational Sympathy is a mindset and practice where software architects, designers, and developers intentionally design how their systems behave in production under load, during failures, and under security threats by planning failure modes, graceful degradation, built-in observability, and clear operational run books that enable early detection and fast recovery.</p><h3>Key elements of operational sympathy</h3><h4>1. Production-Aware Design</h4><p>Systems are designed with a clear understanding of how they will run in real production environments, including deployment models, scaling behavior, and day-to-day operational workflows. This is the most important aspect.</p><h4>2. Load and Scale Consciousness</h4><p>Software is built with explicit consideration for how it behaves under normal load, peak traffic, and unexpected spikes, avoiding assumptions that performance and scaling will be “handled by the cloud.”</p><h4>3. Failure-Aware Architecture</h4><p>Failure is treated as a normal operating condition. Teams proactively identify failure modes, design for graceful degradation, resilience, and ensure partial failures do not cascade into full outages.</p><h4>4. Built-In Observability</h4><p>Observability is designed into the system from the start through meaningful metrics, logs, and traces that allow failures and degradation to be detected early before users feel the impact.</p><h4>5. Operability and Recovery Simplicity</h4><p>Systems are designed to be easy to operate during incidents, with clear levers for mitigation, safe rollbacks, feature toggles, and predictable recovery paths.</p><h4>6. Security as a Runtime Concern</h4><p>Security is considered not just at design time, but as a continuous operational reality, covering threat detection, blast-radius reduction, and incident response. Do not rely primarily on perimeter security. Follow a zero-trust approach as much as possible.</p><h4>7. Cost Awareness by Design</h4><p>Architectural and implementation choices account for how resource usage translates into real cloud costs, preventing runaway bills caused by inefficient scaling or poor defaults. Also consider using native cloud services vs cloud-agnostic solutions. Carry out a cost approximation exercise before finalizing the architecture.</p><h4>8. Runbook-Driven Thinking</h4><p>Teams document known failure scenarios, diagnostic steps, and recovery actions as runbooks, ensuring that operational teams can respond quickly and confidently under pressure. Imagine yourself as the person handling an incident at 2AM, related to the feature you are developing.</p><h4>9. Shared Ownership of Production Outcomes</h4><p>Developers, architects, and operators share responsibility for system behavior in production, closing the gap between those who build systems and those who keep them running. Get out of the “it’s not my problem” mindset.</p><h3>Operational empathy checklist</h3><p>The following checklist provides a framework for architects, designer and developers to evaluate their designs and implementations for operational sympathy. You can make a copy of this spreadsheet and use it. <a href="https://docs.google.com/spreadsheets/d/1jryXy-aNQDoDgjMC8T2D5grdgP5bxQr-DwKJkB2hNfE/edit?gid=0#gid=0">https://docs.google.com/spreadsheets/d/1jryXy-aNQDoDgjMC8T2D5grdgP5bxQr-DwKJkB2hNfE/edit?gid=0#gid=0</a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ALP9b2of1nNXdXkGBV10qA.png" /></figure><p>The operations sympathy score will give an indication of how operationally sympathetic your design is.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/856/1*aitQcpPAPr31LVeGuQk_bg.png" /></figure><h3>Conclusion</h3><p>Non-functional requirements should never be an afterthought or responsibility of the operations teams. They are aspects to be considered from the first whiteboard sketch. Operational sympathy is the mindset that closes the gap between building software and running it, forcing architects and developers to confront how their decisions behave under load, failure, security threats, and real operational stress. When non-functional requirements are treated as design constraints rather than afterthoughts, systems become easier to operate, cheaper to run, faster to recover, and kinder to the people responsible for keeping them running.</p><p>The key takeaway is simple but uncompromising: <strong>if you don’t consciously design for production, production will expose every shortcut you took.</strong></p><p>PS: My talk at the WSO2 Tech Conference which resulted in this article.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FhfADSpkhM90%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DhfADSpkhM90&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FhfADSpkhM90%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/dbe31a09a2d9794a7f29a5965a617da0/href">https://medium.com/media/dbe31a09a2d9794a7f29a5965a617da0/href</a></iframe><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8a9c5dc26b5a" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Fixing an auto-electrical issue using ChatGPT]]></title>
            <link>https://afkham-azeez.medium.com/fixing-an-auto-electrical-issue-using-chatgpt-5512a438acb1?source=rss-d32ea2ba6958------2</link>
            <guid isPermaLink="false">https://medium.com/p/5512a438acb1</guid>
            <category><![CDATA[sre]]></category>
            <category><![CDATA[auto-electrical]]></category>
            <dc:creator><![CDATA[Afkham Azeez]]></dc:creator>
            <pubDate>Wed, 07 Jan 2026 05:07:22 GMT</pubDate>
            <atom:updated>2026-01-07T06:37:32.223Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*zlSFCljXR0sJZHsHYBe3tA.png" /></figure><p>This is my story about how I used AI to fix a long running, annoying starting and charging problem in my Nissan Patrol.</p><p>For a few months, I’ve been having an annoying issue where my batteries weren’t getting properly charged, and also the starter motor dragging before the vehicle eventually starts. I had taken the vehicle to my regular auto-electrician, and they even replaced the alternator saying that it seems weak, which results in the weak charging system. Needless to say, this didn’t fix my issue, which is when I decided to take things into my own hands and do some research, and turned to my trusty companion — ChatGPT.</p><p>Firstly, I bought a <a href="https://www.temu.com/goods.html?_bg_fs=1&amp;goods_id=601103580531461&amp;parent_order_sn=PO-187-09226657715833261&amp;_oak_order_sn=187-09226720630393261&amp;_oak_goods_num=1&amp;sku_id=17610532324064&amp;_x_sessn_id=596dugqrxv&amp;refer_page_name=bgt_order_detail&amp;refer_page_id=10045_1767760268883_ehtuq8c6n8&amp;refer_page_sn=10045">battery health tester</a> on Temu, and uploaded the results to ChatGPT to analyze. I wanted to verify myself that my batteries were healthy, even though battery shops ensured that it was the case. After some bitter experiences with local garages and vendors, I have grown to be skeptical about local businesses.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/0*tVWRYOADdV-UQc5g" /></figure><p>I ran the tests, and uploaded the screen shots to ChatGPT.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*e2eqHzWYDDRhqrCu8hatSQ.png" /></figure><p>This is the summary of the <a href="https://chatgpt.com/share/695de2ac-4464-800c-b595-8e22091997f9">analysis from ChatGPT</a>:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1020/1*4eBU1xM1opv-TwFUf_09Qg.png" /></figure><p>So having established that the batteries are healthy, I continued the analysis.</p><p>This is the prompt I gave ChatGPT:</p><blockquote>My Nissan Patrol Y61 (GU) is having an issue of poor starting and charging. The battery health (SOH) is 100%, the starter motor and alternator are new. What are the possible problems that could be causing the poor starting and charging issue?</blockquote><p>This is the response. ChatGPT asked me to take a look at something that the auto-electrician and I had never thought about — the battery to engine ground strap. It listed that as the top suspect.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*uOV47lWoSUGQqoKN9LXEsw.png" /></figure><p>Voila! The grounding wire was frayed, and corroded, with almost half the strands broken!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*gSMILf9rxsSBqKtlDBL91Q.png" /></figure><p><a href="https://chatgpt.com/share/695de7fb-fbf4-800c-8ebe-3d72c3aa8c83">AI seems to have properly diagnosed the issue</a>. To be sure, I uploaded this pic to ChatGPT to verify.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*6p0Xpu2oGSIUkDUZ8VWQGQ.png" /></figure><p>So I focused on replacing this wire. Since I had left over high gauge cable, I decided to fabricate new cables with redundancy (ha ha, my SRE mindset came in handy).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*92g1lpgtUXlpM3TGr4XIkA.png" /><figcaption>Double redundant thick gauge wire for the battery negative to engine grounding</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/670/1*DK5qiaKvxKGIAKx0hxx7yw.png" /><figcaption>Engine grounding and body grounding custom fabricated straps</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*TmyZLiB-XmEpxbEZcfa3wQ.png" /><figcaption>Big-ass soldering iron used for soldering the terminals — over-engineering for fault tolerance is the way to go</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/696/1*k3xVDgXu-RKPoA0hDLsB6A.png" /><figcaption>Finished product with extra insulation. Again, a bit of over-engineering for fault tolerance.</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*BjXV9Js9Jhkuj_whSXXzVw.png" /><figcaption>Engine strap fixed, and problem fixed!</figcaption></figure><p>No more starting or charging problems. The battery health tester also confirmed that the battery is properly charging after this fix. Finally, a problem that experienced auto-electricians couldn’t fix was diagnosed by AI and I implemented the solution myself, and the problem was fixed using left over wire and components at zero cost. ChatGPT saved me money and gave me peace of mind!</p><p>Next time you your garage asks you to do costly repairs, just consult an AI tool of your choice to get an opinion before spending your valuable money.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5512a438acb1" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The coming of age of WSO2 SRE]]></title>
            <link>https://afkham-azeez.medium.com/the-coming-of-age-of-wso2-sre-f2654fd76630?source=rss-d32ea2ba6958------2</link>
            <guid isPermaLink="false">https://medium.com/p/f2654fd76630</guid>
            <category><![CDATA[wso2]]></category>
            <category><![CDATA[sre]]></category>
            <dc:creator><![CDATA[Afkham Azeez]]></dc:creator>
            <pubDate>Sun, 28 Dec 2025 08:54:28 GMT</pubDate>
            <atom:updated>2025-12-28T08:54:28.069Z</atom:updated>
            <content:encoded><![CDATA[<p>The roots of today’s WSO2 SRE practice go back to the COVID work from home period in late 2020. At that point, WSO2 was running an increasing number of cloud hosted systems across Choreo, Asgardeo, private clouds, and managed customer cloud environments, which were being operated by different teams, each with their own toolkits, processes, and operational instincts. It worked but when a company grows, you have to think about measure which make operations scalable, repeatable and reduce overheads. This brings to my mind the inception of WSO2 Carbon, which was a framework born out of necessity when we started building more products by different teams.</p><p>During that time, Sanjiva suggested creating a dedicated cloud team responsible for everything WSO2 operated in the cloud, under an integrated set of operating practices and compliance expectations. That idea bloomed into what is now WSO2 SRE. It was really less about ‘create a new team and more create a shared way of operating’</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*OXcNr8UXxw_NxczgG0HDDw.png" /><figcaption>An early photo from one of our WSO2 4x4 trips which shows some of the early SRE leaders including Sanjaya Ratnaweera, Manula Waidyanatha and Muhammed Shariq</figcaption></figure><p>The early SRE group brought together people from systems engineering, devops, and platform engineering backgrounds. What made those early days work was not structure, because there was very little of it, but intent. A shared belief that running production systems deserved first class engineering attention. That incidents should lead to learning, not blame. That reliability should be designed, not improvised.</p><p>This was the vision and objectives from the slide deck I put together for the inauguration:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*VT3fxp3X9gOFwHwzWSCjDg.png" /><figcaption>From the vision slide deck created in 2021</figcaption></figure><p>We approached this across several dimensions, with <strong>people, processes, tools, compliance, and skill development</strong> all evolving together. When different teams have been following different approaches there is always the challenge of getting everyone to buy in to a common vision where they have to unlearn some things they have been following for years, and learning and adopting new approaches.</p><h4>People and teams</h4><p>New stuff cannot be forced onto people and you need to be patient and educate them about how things would be done in a better and more efficient way with less cognitive load. This was the biggest challenge which took time to ingrain in the people. The message I used to constantly deliver was that people should be working smart which will make life easier for them, and also deliver more value for our customers. When operations teams are used to certain manual ways of doing things, adopting repeatable automation could be a challenge since they feel that the manual way is working and may not understand why it has to be changed. This is where leaders have to step up and carefully guide their teams. This could be tiring for leaders but patience during this phase is of utmost importance. I was fortunate that the SRE team had many good leaders who had the respect of their team and we were able to slowly transform the team into adopting newer or different and more efficient ways of getting things done.</p><h4>Processes</h4><p>When WSO2 was a startup, we felt that processes were an unnecessary overhead and I was very much anti-process. We had the “just do it” mentality and approach. That may be the way to go for a startup because taking time to define processes could be a waste of time, since startups have to try many things and fails fast, and processes also would have to be frequently changed. However, once you grow into an established business serving a globally distributed large customer base, you need to have consistency, stability, repeatability and compliance. Practices that lived in people’s heads had to be written down. Expectations had to be made explicit. This naturally led to one of the most defining milestones in the SRE journey - the birth of the WSO2 Cloud Standard Operating Procedures (v 1.0) in early 2024, which was developed with inputs from all teams concerned. This was the first time Integration, IAM, and security operations aligned around a single operational playbook. Incident handling, severity definitions, escalation paths, post incident reviews, and ownership models were no longer implicit or team specific. They became shared and repeatable.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qnXm-_8xhuYZfP5C-ryiuQ.png" /><figcaption>Cloud SOP — Incident management process</figcaption></figure><p>Once formalized, we conducted internal trainings to ensure that the teams were aware of these procedures. We ensured that the “<strong>why</strong>” factor was explained because people should understand the reasoning behind <strong>how</strong> certain things are done, or else they would just think that it is technical debt or process overhead.</p><h4><strong>Tools</strong></h4><p>Unification of the tools used by the teams is essential for many factors including building familiarity, operational excellence, governance and cost savings. Our toolkit is built around ServiceNow which has excellent functionality to support most of our SOPs and allows great levels of customization. We have also standardized on monitoring and alerting tools, compliance validation tool etc.</p><h4>Compliance</h4><p>Ensuring compliance with relevant legal, statutory, regulatory, and contractual obligations, industry standards, and organizational policies is paramount in maintaining the integrity, security, and trustworthiness of WSO2 cloud deployments. The WSO2 security and compliance team, under the guidance of the WSO2 CISO directly works with the SRE team, and the Security Operations Center (SOC) to ensure that the processes are compliant with standards including SOC2 and ISO/IEC 27001.</p><h4>Skills development</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1008/1*O2VxO8r06jtCi3Oi9xat3Q.jpeg" /></figure><p>From the beginning we understood that training our team members on technologies, methodologies, tools and practices was paramount to up lifiting the SRE team and providing the best possible service to our customers. Hence we came up with plan to train and get team members certified on cloud technologies and funded training, which is fully or partially funded by the company. We always had the risk of people who built up their skills levels leaving and joining other companies or migrating, but we kept the pipeline going with shadowing people so that others were ready to step up.</p><h3>2023 — SRE becomes part of WSO2 Customer Success</h3><p>A major milestone in the journey was integrating the SRE team into WSO2 CS, which turned out to be a natural fit where SRE has found its permanent home. Under the leadership of Chief Customer Officer, Ruwan Linton, we have been able to shape not only the SRE team, but part of the CS team was reshaped into a Customer Reliability Engineering (CRE) team. Also, the synergy that resulted in this merger resulted in SOPs defined for other aspects of CS including onboarding and support. We also recently launched the WSO2 CS portal — the one stop shop for customer success at WSO2.</p><h3>2025 — The coming of age</h3><p>Over the past year, a new generation of leaders has emerged from within the SRE team itself. These leaders include engineers who were raised within the ranks of the practice itself. They experienced incidents, SOP implementation, on-calling, and post-incident reviews. They grasp the purpose or intent behind all these because they were the ones who influenced the development thereof.</p><p>This means a lot to me. WSO2 SRE is no longer reliant on a few early adopters to build SRE for the company. The burden is no longer the responsibility of a few. Decision making is no longer the responsibility of a few. It is no longer centralized to a few.</p><p>The impact of younger blood in the leadership ranks is also reflected in a difference in the overall energy of the group. It is a place of curiosity and challenge and a willingness to take a look at what have all along been assumed to be true. But at the same time, there is a respect for what has been built upon for so long. This is no small thing and is a hallmark of real maturity.</p><p>Chathura Ekanayake, and old timer at WSO2 taking up the role of <strong>database performance specialist and FinOps</strong> lead was another great addition this year. We have been able to rectify some pressing performance issues, identify areas of cost optimization and build dashboards that provide much needed cost insights. There is a lot more to be done in this area, but we are off to a good start.</p><p>Another significant event during this year is the integration of <a href="https://wso2.com/acquisitions/moesif/">Moesif, which WSO2 acquired this year</a>, into the SRE team. It introduced new ways of thinking about scale, observability, and customer facing reliability. More importantly, it reinforced the idea that SRE leadership is not static. It must continuously evolve as the organization grows and as new capabilities are added.</p><p>Moesif becoming part of WSO2 SRE was not about absorbing a team. It was about integration. Aligning practices. Sharing expectations. Learning from each other. This kind of integration is only possible when a practice is confident in its identity and open to change at the same time.</p><p>Now when I consider these two situations together, the scenario that unravels itself is clear. Leaders arising internally and the capacity to absorb leaders and experience from outside are signs of having outgrown the adolescence of the organization.</p><p>This is why 2025 is special.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2hdoQAeEPoNHJnZ-daHnNQ.png" /><figcaption>Part of the SRE leadership team. From L-R: Akila Bandara, Suranga Nisiwasala, Athiththan Kathirgamasegaran, Deependra Ariyadewa, Afkham Azeez, Eshana Paranagama, Chalitha Kulathunga, Migara Pramod, Sachith Kasthuriarachchi, Vimukthi Perera</figcaption></figure><h3>The road ahead</h3><p>As we step into 2026, WSO2 SRE feels grounded, confident, and ready. The foundations are strong, the practice is mature, and the team has grown into a thoughtful blend of experience and fresh energy. We now have the clarity of well defined processes, the support of solid tooling, and most importantly, a team who care deeply about the purpose and goals.</p><p>The work ahead is no longer about proving ourselves. It is about refining what we have built, mentoring the next generation, and scaling reliability in ways that remain sustainable. With strong leadership, a shared sense of purpose, and a culture that values learning over blame, WSO2 SRE is well positioned for the road ahead.</p><p>Onwards and upwards!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f2654fd76630" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[A tale of 2 decades]]></title>
            <link>https://afkham-azeez.medium.com/a-tale-of-2-decades-8d05f0849ea8?source=rss-d32ea2ba6958------2</link>
            <guid isPermaLink="false">https://medium.com/p/8d05f0849ea8</guid>
            <category><![CDATA[wso2]]></category>
            <dc:creator><![CDATA[Afkham Azeez]]></dc:creator>
            <pubDate>Mon, 01 Dec 2025 18:59:12 GMT</pubDate>
            <atom:updated>2026-09-13T10:50:58.873Z</atom:updated>
            <content:encoded><![CDATA[<p>There’s an interesting story from my undergraduate days that still makes me smile. One of my favourite lecturers at university once advised us that we should <em>never</em> stay at the same software company for more than two years. “You stop learning or they stop valuing you, so you should move on.” he said</p><p>That idea deeply etched itself into my young mind and I took it to heart. During the first two years of my career, I switched jobs three times! Not because I disliked the places I worked, but because I believed that I had to move in order to grow. Two years was the limit. I was the one in my university batch who was continuously on probation :)</p><p>But then in December 2005, WSO2 happened! Gradually, that rule which I once held as unquestionable truth began to fall apart. Here’s my WSO2 story.</p><h3>How it started</h3><p>In August 2005, I had just begun working at a new company when I came across an eye catching and unique job advertisement from a small startup called WSO2. They were looking for software engineers and architects and the ad said that their architects were their best developers and that everyone was given a laptop computer to work on, which was rare those days. This was the only open source company at that time, and they were contributors on high profile projects at the Apache Software Foundation, and I genuinely wanted to apply, but I hesitated because I had only just joined my current workplace. It felt too early to make a move. I felt a sense of dissapointment too because I really wanted to join this company. Earlier during my internship, I came across an open source project called <a href="https://svn.apache.org/repos/asf/webservices/archive/soap/trunk/java/docs/intro.html">Apache SOAP</a>, and while browsing the source code, was thrilled to find that one of the authors by the name of <a href="https://www.linkedin.com/in/sanjivaweerawarana/"><strong>Sanjiva Weerawarana</strong></a> which was a Sri Lankan, Sinhala name and that was a time when we hadn’t heard about any Sri Lankan open source contributors. WSO2 happened to be a company founded by this guy, so why wouldn’t anyone jump at the opportunity to join it?</p><p>Around that time, my good friend and University batch mate <a href="https://www.linkedin.com/in/eran-withana/"><strong>Eran Chinthaka</strong></a> encouraged me to apply. He thought I would be a good fit and ended up recommending me. Not long after, I received a call inviting me for an interview.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ViUJ7Gi37RHrmV5ZpgagIw.jpeg" /></figure><p>I remember riding my Bajaj Pulsar 150 (HT-4418), which has a history of its own, to the WSO2 office during the lunch break. The company was very small back then, just a handful of people.A bit nervous and certainly expecting a grilling, but to my surprise, the interview turned out to be the easiest I’d ever faced. Sanjiva asked if I knew about SOAP or WebServices; I replied “no.” He took about 20 minutes to sketch out what SOAP/WebServices/WS-* meant, then asked if I knew XML. I said yes thinking XML was just like HTML with some weird angle bracket syntax. It was only after joining WSO2 that I realized how little I really knew about XML. We spoke briefly, and at the end he simply asked, “When can you start?” I wasn’t expecting that, and it caught me off guard.</p><p>After I went back to my current office, Sanjiva actually called the Director of Engineering there who happened to be his friend to check whether it would be alright for me to join WSO2 so soon after being hired. His room was adjoining my work desk and when the phone rang, I knew it was Sanjiva because he had told me that I would be hired only if the current company agreed to release me. That phone call made the whole situation a bit uncomfortable for me, but in the end everything worked out, and that’s how it all began.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*G-kLbRXETxOlezgQuEKmCw.png" /><figcaption>The email that changed my life</figcaption></figure><p>Stock options in the parent US company? What does that even mean? I didn’t really care because I was thrilled at the opportunity to work in this place. It was years later that I realized the real value of those stock options!</p><h3>December 1st, 2005</h3><p>This was a turning point in my life and I started working at WSO2, and never imagined that I would still be here 20 years later and just be one of the two who complete 2 decades of service. The other person completing 20 years is <strong>Flora Ariyapala</strong> from our administration team. She was the one who made sure everyone was taken care of and that our morning and evening refreshments arrived on time, without fail. Thanks to her, those small routines became part of our daily rhythm. The tradition of gathering for tea, standing in a circle, joking and chatting, continues even today.</p><h3>Very first assignment</h3><p>Those days there was no induction or orientation. You directly get to work.</p><p>I was teamed up with <a href="https://www.linkedin.com/in/ruchith/"><strong>Ruchith Fernando</strong></a> to take over a <strong>DO</strong>M + AXI<strong>OM</strong> object model, code named DOOM that was being developed so that we could utilize existing DOM libraries as well as the newly developed AXIOM to work in the same Java runtime in harmony. This is when I realized that I had to learn everything about XML from scratch. Object models, parsers, StAX (Streaming API for XML) for efficient parsing and so on. Initially I was a bit overwhelmed because I had to get my head around the recursive Java method calls to really understand what was going on, in addition to understanding the specifications and technologies. Late December is usually when I take road trips with my family. When I mentioned this to Ruchith, he reminded me that the DOOM implementation needed to be finished before Sanjiva returned from his overseas trip. So I ended up taking my laptop along, reading documents and specs, and writing code while riding as a passenger in the van. By the end of December, we had the first working version ready. I went on to work on other spec implementations like SAAJ (SOAP with Attachments for Java), JAX-WS, JAX-RS and so on. Reading a Web Services (WS) spec and implementing it became second nature for most of us at WSO2 those days.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*TrhsUMB5lbt_U3wsVNNckg.jpeg" /><figcaption>The entire company including James Clark, the angel investor of WSO2, in the middle</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*D1AC_C640ASk0ssqq2la3g.jpeg" /><figcaption>Eran Chinthaka, Ajith Ranabahu, Ruchith Fernando and Deepal Jayasinghe — pioneers at WSO2 who all went on to get their PhDs in the US.</figcaption></figure><h3>Axis2 contributions and Apache committership</h3><p>One of the defining parts of my early years at WSO2 was working on <a href="https://axis.apache.org/axis2/java/core/">Apache Axis2</a>. It was both a learning experience and a chance to contribute to a major open-source project that shaped web services at the time. I started by working on smaller fixes and improvements, gradually taking on more complex areas of the codebase, eventually receiving my <a href="https://azeez78.blogspot.com/2006/02/i-received-my-apache-committership_01.html?view=sidebar">committership</a> on the 1st of February 2006. I consider <strong>Dims</strong> (<a href="https://www.linkedin.com/in/davanum/">Davanum Srinivas</a>) to be my first mentor at WSO2 who taught me the ropes of open source software development and contribution. I recall receiving a couple of strongly worded emails from him after I messed up some files - I think some Maven files :) This ensured that I was much more careful and checked my work thoroughly before committing any changes. Those days where builds would take ages was a fun time and we would play indoor cricket in office until the build completes. Many builds ended up in test failures and after fixing the failures, we would resume the cricket match.</p><p>I learned the Apache Way, Community over Code, meritocracy and many other values from the Apache Software Foundation and these have become engrained in me as well as WSO2.</p><p>One of the highlights in my career was developing the clustering framework for Axis2. It was first based on Tomcat Tribes and later I wrote a Hazelcast based implementation.</p><p>Eventually I went on to earn my PMC membership and <a href="https://azeez78.blogspot.com/2008/12/apache-membersazeez.html">Apache Membership in 2008 </a>in recognition of my sustained contributions.</p><h3>First release — Tungsten 1.0-alpha — 2006</h3><p>The first product release was a very memorable one where the team led by Eran Chinthaka worked for over 35 hours in a coding marathon to build the first MVP of Tungsten which was to be demoed to a prospective client. I wrote the entire persistence layer using Hibernate during this hackathon and remember having to deal with intermittent persistence issues due to race conditions, which was very stressful to troubleshoot under time pressure, but in the end we managed to create a working release complete with samples and basic documentation.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*OUQmAHkyi_eqYNgCmHGtCQ.jpeg" /><figcaption>Seated from L-R: Saminda Abeyruwan, Afkham Azeez, Deepal Jayasinghe, Ruchith Fernando. Standing from L-R: Chamil Thanthirimudalige, Eran Chinthaka, Ajith Ranabahu, Chamikara Jayalath, Deependra Ariyadewa</figcaption></figure><h3>First company trip — Habarana Lodge — 2006</h3><p>We had the tradition that started in 2006 of going on a company trip every April just before the April holiday break. The first one was to Habarana Lodge and the entire WSO2 staff including a few families could sit at a single long table. WSO2 was not a known name at that time and I recall the table reservation board said “Chamikara party”, I guess because <a href="https://www.linkedin.com/in/chamikaramj/"><strong>Chamikara</strong></a> may have made the hotel booking. Only a few us were married and it was only <a href="https://www.linkedin.com/in/asankha/"><strong>Asankha Perera</strong></a>, Sanjiva and I who had kids.</p><p>The entire team fit into one bus &amp; I drove with my wife and daughter in my trusty old Lancer Station Wagon (19–1720), which was my first car.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*m6bDxYs79y-0geGnaQ7nZA.jpeg" /><figcaption>A tank (wewa) opposite Habarana Lodge</figcaption></figure><h3>First official product release — Tungsten 1.0–2006</h3><p>We had a dinner party to celebrate the Tungsten 1.0 launch at the Gallery Cafe, Colombo. This was almost the entire company. The gentleman in the white shirt on the left in the photo below is <a href="https://www.linkedin.com/in/ptagare"><strong>Pradeep Tagare</strong></a> from Intel Capital which invested in WSO2 providing much needed capital in our startup.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_H6wVZEEikbphL1_OsyBVg.jpeg" /><figcaption>Almost the entire WSO2 team</figcaption></figure><p>This product was later renamed to WSO2 Web Services Application Server (WSAS) and I was the first product manager.</p><h3>Cloud research — 2007</h3><p>WSO2 was an early adopter of cloud technologies. We recognized the potential of the cloud even when many others remained skeptical. Encouraged by Sanjiva, I began exploring AWS in late 2007, back when EC2 and S3 were essentially the only services available. This work ultimately became the foundation for my Master’s thesis, <em>“</em><a href="https://dl.lib.uom.lk/items/63b37b76-9014-48b4-b144-8a62a54963c3"><em>Autoscaling Axis2 Web Service on Amazon EC2</em></a><em>”. </em>Autoscaling wasn’t a well known term at that time. I used Apache Synapse and Axis2 as the base technologies.</p><p>Some of my research work became the basis of WSO2 Stratoslive and included the Elastic Load Balancer (ELB) and the Axis2 clustering implementation.</p><p>I was also the first recipient of WSO2’s <strong>graduate studies sponsorship program</strong>, a program that continues to support employees’ education to this day.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*RALtHCZokSe34uxpOmScng.png" /><figcaption>An image from one of my slides where the traffic is represented by a growing or shrinking token queue size, which is used for scaling decisions by the load balancer which I called the elastic load balancer.</figcaption></figure><h3>WSO2 treats, parties and weddings</h3><p>Treats, parties, weddings and celebrations are part and parcel of WSO2 culture. In the early days, for any and every reason we had to give a treat, be it the first overseas trip, committership, promotion or for no reason at all. We could afford to do that because the team was small.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*LZG9ZrVF1MesXWmZ.jpg" /><figcaption>All dressed up for <a href="https://www.linkedin.com/in/hasminabdulcader/">Hasmin</a>’s wedding. From L-R: Lahiru Sandakith, Nandika Jayawardena, Afkham Azeez, Saminda Abeyruwan, Asankha Perera, Kaushalya Kapuruge, Ajith Ranabahu, Eran Chinthaka, Deepal Jayasinghe, Ruchith Fernando, Chamikara Jayalath, Sanjaya Ratnaweera, Sahan Gamage, Amila Suriarachchi, Sanka Samaranayake, Samisa Abeysinghe, Damitha Kumarage</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Ujb6DGmjChVgu-Uiy-rsTw.jpeg" /><figcaption>At Milinda Pathirage(Jambuwa)’s wedding</figcaption></figure><h3>Games@WSO2</h3><p>From the very beginning, WSO2 always had a table tennis room, thanks to Chinthaka convincing Sanjiva that we needed one, I think. Even today, almost every major WSO2 office has a TT table. Back then, every batch of interns somehow finished their internship as pretty good table tennis players.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*1DYdHGjguZVBPuU7zbkThA.jpeg" /></figure><p>During the early days, in the evenings, the WSO2 car park transformed into a basketball court. Any vehicle not moved before the games begin were physically carried and moved aside to make room for the game :) What we played at WSO2 was a very unconventional form of basketball where almost anything goes, much to the frustration of seasoned players like Asanka. And when <strong>Udeshika</strong> was on the opposing team, our priority was usually protecting certain other “assets,” with actually holding on to the basketball becoming a secondary concern.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/1*HlYJWx1_DBj0Wq0JmuIj5A.jpeg" /></figure><p>On certain days, the basketball court also used to transform into an indoor cricket arena.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FFLdiNvAE9JI&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DFLdiNvAE9JI&amp;image=http%3A%2F%2Fi.ytimg.com%2Fvi%2FFLdiNvAE9JI%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/aaf372d44b9b8651aa861bdf1344c683/href">https://medium.com/media/aaf372d44b9b8651aa861bdf1344c683/href</a></iframe><h3>The birth of Carbon — 2007</h3><p><em>The era of middleware a la carte</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*WuxpcHPzICtO7k92.png" /></figure><p>After building two products and beginning work on a third, it became obvious that many features were being reinvented across teams. We created a few shared libraries, and I even put together a simple user interface framework called the Management Framework UI, but it was clear that we needed a more standard and scalable approach to accelerate product development. At first, several of us explored the idea of creating our own in house framework. Around that time, <a href="https://www.linkedin.com/in/paulfremantle/"><strong>Paul Fremantle</strong></a>, our co founder and CTO, suggested that we adopt OSGi, a new way of organizing and managing Java libraries by packaging them into modules.</p><p>The idea was not well received at first. Many of us in the technical leadership were hesitant because OSGi was complex and not widely adopted. We wanted the freedom to build our own framework, and the discussions eventually reached a stalemate. That was when Sanjiva stepped in, made the call to adopt OSGi, and set a deadline for us to implement our features on top of it and ship our products. Looking back, that decision could not have come at a better time. Despite the steep learning curve, OSGi allowed us to build and release products much faster.</p><p>The team struggled at the beginning because there was very little documentation or community support around OSGi. Engineers like <a href="https://www.linkedin.com/in/samindaa/"><strong>Saminda Abeyruwan</strong></a>, and later <a href="https://www.linkedin.com/in/sameerajayasoma/"><strong>Sameera Jayasoma</strong></a>, spent countless late nights digging into the details and figuring out how it all worked. A major turning point came when we brought down <strong>Gunnar Ekolin</strong> from Knopflerfish to conduct a three day workshop. That training helped the entire team finally understand OSGi in depth and gave us the confidence to fully adopt it across our product stack. Gunnar’s session was the boost we needed and to manage the dependency loading order problem, <a href="https://www.linkedin.com/in/ruwanlinton/"><strong>Ruwan Linton</strong></a> even quickly wrote a dependency manager.</p><p>Powered by OSGi, we went on to build 25 products! At the heart of the framework was the Carbon Kernel, which provided all the utilities required to build products. I got the opportunity to lead this team for a while, with highly talented team members like Sameera Jayasoma, <a href="https://www.linkedin.com/in/kishanthan/"><strong>Kishanthan Thangarajah</strong></a>, <a href="https://www.linkedin.com/in/pradeepfn/"><strong>Pradeep Fernando</strong></a> and the likes.</p><h3>First Quick Start Program (QSP) — 2008</h3><p>In mid 2008, we got an interesting assignment. To build a solution for Computer Associates (CA). They had to propagate changes to their system through a multicast eventing mechanism, basically a solution based on event driven architecture (EDA). We put together a team lead by Paul Fremantle and included <a href="https://www.linkedin.com/in/asankaabeysinghe/"><strong>Asanka Abeysinghe</strong></a>, <a href="https://www.linkedin.com/in/chathuraekanayake/"><strong>Chathura Ekanayake</strong></a>, Ruwan Linton, <a href="https://www.linkedin.com/in/nandanamihindu/"><strong>Nandana Mihindukulasooriya</strong></a> and myself. <a href="https://www.linkedin.com/in/fergusondonald/"><strong>Don Furguson</strong></a>, then CTO of CA explained the problem to a team consisting of the five members from WSO2 and also folks from CA. I remember Paul breaking Don’s vision to technical terms we could understand and build a solution upon, and by the end of 3rd day of this assignment, we had a working prototype built! We in fact had built a thin slice of the solutions architecture. CA was highly impressed and became our first USD 1M customer. Their feedback was also immensely useful towards building our Carbon based products which were in their infancy at the time.</p><h3>First service award — 2011</h3><p>When I received the first 5 year service award, I couldn’t have imagined the journey WSO2 or myself would have taken in the next 15 years. We’ve been through so many up and downs, mainly ups, and the ride has always been thrilling.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/662/1*O3KlA74KjfLvLDPN_07XXQ.png" /><figcaption>Getting my 5 year service award — 2011 January</figcaption></figure><h3>Raspberry Pi cluster — 2013</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/640/0*isJPYW31WmeSV8UJ.png" /></figure><p>A team led by <a href="https://www.linkedin.com/in/sumedharubasinghe/"><strong>Sumedha Rubasinghe</strong></a> and <a href="https://www.linkedin.com/in/lakmal-warusawithana-b8191450/"><strong>Lakmal Warusawithana</strong></a> built a Raspberry Pi cluster that we used to run WSO2 technology, including the conference application. It was a memorable project for me because I helped build part of the electronics and wrote the command console application used to monitor and control the cluster. The creator of the Raspberry Pi, Eben Upton, delivered a keynote at this conference in the United Kingdom, which made the experience even more special. One of the key technical achievements of the project was trimming down the Carbon kernel so that it could run efficiently in a very resource limited environment.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*FMYGf51oQ3rnbuWO.png" /></figure><h3>Sabbatical at Vega — 2015</h3><p>Being one of the early members of WSO2 came with many firsts. The WSO2 sabbatical program, started in 2015, actually began with me. In a conversation with Sanjiva, he suggested that I take a break and spend a few months working on something I genuinely enjoyed. It was a three month period, and at first I had no idea what I could do outside my usual work. Knowing my passion for vehicles, Sanjiva recommended that I spend the time with Team Vega, which was building an electric supercar.</p><p>So in mid January 2015, I began my sabbatical and joined the <a href="https://www.vega.lk/">Vega</a> team for three months. It turned out to be an unforgettable experience. I was essentially working in a high tech garage filled with fascinating machines and an incredible group of people: software engineers, mechanical engineers, electronics engineers, steel fabricators, welders. It was a wonderful time.</p><p>I am grateful to Dr. <a href="https://www.linkedin.com/in/harsha-subasinghe-9898866b/"><strong>Harsha Subasinghe</strong></a>, the inventor and visionary behind the project, for giving me the opportunity to be part of it.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*K9W5vtCvvGxdYp2nVAF6CA.png" /><figcaption>Space frame chassis</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3cdD8sG3w_VOYLpx0mPLuw.png" /><figcaption>Team Vega lead by Dr. <a href="https://www.linkedin.com/in/beshankulapala/">Beshan Kulapala</a></figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*wqV13zTJvyvvkU4dISoPjQ.png" /><figcaption>Even the President of Sri Lanka visited our workshop</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*x0YBFzChC_Kt5zDy9kmwEA.jpeg" /><figcaption>The final product</figcaption></figure><h3>The move to microservices — MSF4J — 2015</h3><p>With microservices architecture (MSA) becoming a hot topic starting in 2014, we started looking at developing a truly container-native, lightweight, low resource consuming, quick bootup capable microservices framework. We named it Microservice Framework for Java (<a href="https://github.com/wso2/msf4j">MSF4J</a>). I wrote the initial version which was later taken over and improved by my team mate <a href="https://www.linkedin.com/in/samiyuru/"><strong>Samiyuru Senarathne</strong></a> who took it to the next level. We ran benchmark tests against many widely used frameworks at the time and came up on top.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*_G1OGi9jZ7ApelRy.png" /></figure><h3>Kubernetes Adoption — 2015</h3><p>WSO2 was one of the earliest adopters of Kubernetes and we had a hunch that it was going to be the de-facto way of orchestrating containers. Lakmal and I had the privilege of <a href="https://kubecon2015.sched.com/event/4WyJ/unleashing-k8s-to-reduce-complexities-of-an-entire-middleware-platform">presenting</a> at the very first <a href="https://kubecon2015.sched.com/">KubeCon</a> at San Francisco.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*0dkq3xS230Wg15N4OBlYuQ.png" /></figure><h3>Ballerina distributed transaction management — 2017</h3><p><strong>Prof. </strong><a href="https://www.linkedin.com/in/fleymann/"><strong>Frank Leymann</strong></a>, the first and only WSO2 fellow, had written a spec on distributed transaction management for the <a href="https://ballerina.io/">Ballerina</a> programming language, and I got pulled in to a couple of meetings to review it, after which I volunteered to take a crack at implementing it and wrote the first version. As always, Frank had thought through everything and included it in the spec, and I had a wonderful experience implementing the infection based distributed transaction flows.</p><h3>Cell-based Architecture and Cellery — 2018</h3><p>Paul and Asanka invented the concept of <a href="https://github.com/wso2/reference-architecture/blob/master/reference-architecture-cell-based.md">cell based architecture</a> in 2017 to address the challenge of logically grouping and deploying microservices and defining clear access control boundaries. In 2018, we held an architecture meeting in Dubai to explore how this model could be realized on Kubernetes. The result was <a href="https://github.com/wso2/cellery">Cellery</a>. It went through several code names along the way, starting with Distributed Integration Computer for Kubernetes (DICK), then Virtual Integration Computer for Kubernetes (VICK), and finally settling on Cellery to better reflect its focus on the cell architecture model. The initial renaming of DICK to VICK was due to my insistence since I was the head of the project :)</p><p>We had an elite team mentored by Paul, which was later led by <a href="https://www.linkedin.com/in/sinthuja-rajendran-b45209122/"><strong>Sinthuja Rajendran</strong></a> and included Deependra Ariyadewa, <a href="https://www.linkedin.com/in/hasinthaindrajee/"><strong>Hasintha Indrajee</strong></a>, <a href="https://www.linkedin.com/in/farasath/"><strong>Farasath Ahamed</strong></a>, <a href="https://www.linkedin.com/in/anuruddhal/"><strong>Anuruddha Liyanarachchi</strong></a>, <a href="https://www.linkedin.com/in/asma-jabir/"><strong>Asma Jabir</strong></a>, <strong>Miraj Abeysekara</strong>, <strong>Madhuka Wickramapala</strong>, <strong>Minoli Perera</strong>, <strong>Nadun de Silva</strong>, <a href="https://www.linkedin.com/in/isuruhaththotuwa/"><strong>Isuru Haththotuwa</strong></a>, <a href="https://www.linkedin.com/in/tharindujayathilake/"><strong>Tharindu Jayathilake</strong></a>, <strong>Madusha Gunasekara, </strong><a href="https://www.linkedin.com/in/anjana-supun/"><strong>Anjana Supun</strong></a><strong> (intern), </strong><a href="https://www.linkedin.com/in/shairam-sritharan/"><strong>Shairam Sritharan</strong></a><strong> (intern) </strong>and a few others. I remember the good times we had with this team in terms of R&amp;D, and also the almost weekly treats and team dinner outings.</p><h3>Choreo — 2020</h3><p>In January 2020, a small team was formed to quickly prototype the initial version of what was later to become Choreo and I was a part of this team and developed two of the control plane microservices, and also wrote the initial Docker Compose scripts and later the first version of the Kubernetes manifests where I introduced Kustomize to enable overlays for different environments. It was in this year that I got my CKAD and CKA certifications which gave me an indepth knowledge of the intricacies of the technology.</p><h3>From development to SRE — 2021</h3><p>During the COVID work from home period in late 2020, Sanjiva proposed forming a dedicated cloud team to take responsibility for everything WSO2 operated in the cloud, including Choreo, Asgardeo, private clouds and managed clouds, under a unified set of operating practices and compliance. At that time, we had different teams running systems following different toolkits and processes. I was asked to lead this new effort. It was a significant shift for me because my entire career until then had been in development and product architecture, with very little exposure to the operations side. This marked the beginning of the WSO2 SRE team, and I learned the craft on the job. It turned out to be a wonderful experience because it pushed me out of my comfort zone. Looking back, stepping out of my comfort zone from time to time is what allowed me to grow and explore new seas. I’m immensely thankful to the first SRE team which included <a href="https://www.linkedin.com/in/chamith/"><strong>Chamith Kumarage</strong></a>, <a href="https://www.linkedin.com/in/msmshariq/"><strong>Muhammed Shariq</strong></a>, <a href="https://www.linkedin.com/in/sratnaweera/"><strong>Sanjaya Ratnaweera</strong></a>, <a href="https://wso2.com/about/team/deependra-ariyadewa/"><strong>Deependra Ariyadewa</strong></a> and others from systems engineering and devops backgrounds for being pioneering members of the SRE team and helping me quickly learn this new area.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*X14p4bnYiHeCwsz21P1oxQ.png" /><figcaption>Part of the SRE team during a trip</figcaption></figure><p>Five years later, it is deeply satisfying to look back at the team we’ve built and the mature SRE practices we shaped and refined, with many leaders growing through that journey as well. In late 2023, the SRE team became part of the larger WSO2 Customer Success organization, which opened up even more opportunities for growth. Today we operate as two main branches within Customer Success, Customer Reliability Engineering (CRE) and Site Reliability Engineering (SRE), under the leadership of Ruwan Linton. I have an exceptional team with leaders of the likes of <a href="https://www.linkedin.com/in/ruwan-yatawara/"><strong>Ruwan Yatawara</strong></a>, Deependra Ariyadewa, <strong>Brayan Sylvester</strong>, <a href="https://www.linkedin.com/in/chalithasanyuja/"><strong>Chalitha Kulathunga</strong></a>, <a href="https://www.linkedin.com/in/vimukthiperera/"><strong>Vimukthi Perera</strong></a><strong> </strong>and others.</p><h3>2024 Reunion</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*aHiA2bsLF1yGHhE-a0SBPA.jpeg" /></figure><p>Some of the original team who were part of the 35 hour coding marathon met up at WSO2Con US 2024 to share our memories and talk about old times.</p><h3>Appreciation</h3><p>I want to express my heartfelt thanks to everyone I have worked with, learned from, and shared this journey with. I have been inspired, challenged, supported, and shaped by so many of you, and I value the friendships that have grown along the way. Thank you for being part of these twenty years. I may not be able to mention every name, but each of you has contributed to who I am today, and I am truly grateful.</p><h3>2025 and beyond</h3><p>As we look ahead, WSO2 is entering an exciting new chapter, especially with the growing industry shift toward cloud repatriation and more balanced deployment strategies. This change opens up a fresh set of engineering and operational challenges, and it is a space where we are already building our next wave of expertise. For me, this marks the beginning of another learning journey, one that will require rethinking assumptions, refining our platform, and helping customers run WSO2 technologies with greater control and efficiency. After twenty years here, I still find myself energized by what lies ahead. The landscape keeps evolving, and so will we.</p><p>As Simon Sinek reminds us, the infinite game is not about winning, but about staying in the game with a cause that inspires us. After twenty years at WSO2, I am grateful to still be playing, still learning, and still pushing toward a future that is always unfolding.</p><p><strong>P.S.:</strong> At the time of writing publishing this post, my country, Sri Lanka is going through a tough time after suffering perhaps the <a href="https://english.newsfirst.lk/2025/12/01/355-dead-366-missing-1-1-million-affected-in-sri-lanka%E2%80%99s-worst-flood-in-decades">worst natural disaster in its history</a>. At this time a full assessment of the damage hasn’t been carried out and there are still many people trapped and waiting rescue. So I’m not really in a mood to celebrate. We will need all the help to once again rise as a nation.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8d05f0849ea8" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Root Cause Analysis — Report template released]]></title>
            <link>https://medium.com/operational-sympathy/root-cause-analysis-report-template-released-de1add6345f8?source=rss-d32ea2ba6958------2</link>
            <guid isPermaLink="false">https://medium.com/p/de1add6345f8</guid>
            <category><![CDATA[sre]]></category>
            <category><![CDATA[rca]]></category>
            <category><![CDATA[operational-sympathy]]></category>
            <category><![CDATA[cloud]]></category>
            <dc:creator><![CDATA[Afkham Azeez]]></dc:creator>
            <pubDate>Tue, 28 Jan 2025 05:00:59 GMT</pubDate>
            <atom:updated>2026-02-19T06:46:33.530Z</atom:updated>
            <content:encoded><![CDATA[<h3>Root Cause Analysis — Report template released</h3><p>I have been <a href="https://afkham-azeez.medium.com/the-curious-case-of-the-leaking-land-rover-38e28758e6f5">talking</a> and <a href="https://afkham-azeez.medium.com/turning-wrenches-and-diagnosing-systems-28b63ec5ddbf">writing</a> about RCA for a while now trying to get everyone in general, and my teams in particular to carry out proper and high quality root cause analysis. One of the questions people have asked me about was how the analysis and findings should be documented. Based on the work we have been doing in the past and the gaps I’ve identified, I’ve created a 1.0 version of an RCA template which I’m releasing under the Apache Software Licence (ASL) 2.0.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fdocs.google.com%2Fdocument%2Fd%2F1w3TLoGAF_46xEYVQLooDvNVA-sidUhM4FqQvESRhBb8%2Fpreview%3Fembedded%3Dtrue&amp;display_name=Google+Docs&amp;url=https%3A%2F%2Fdocs.google.com%2Fdocument%2Fd%2F1w3TLoGAF_46xEYVQLooDvNVA-sidUhM4FqQvESRhBb8%2Fedit%3Ftab%3Dt.0&amp;image=https%3A%2F%2Flh7-us.googleusercontent.com%2Fdocs%2FAHkbwyKeytgTbsTrYveF9BpElK6krzhqp5lpU9muZQ4hoOTOxHMF97PYXubgkfDzRae7J3ePwcY30YXtdbXE26aWyyblo7P4zKyUK39g_OjjdPhu9OCWaCBS%3Dw1200-h630-p&amp;type=text%2Fhtml&amp;scroll=auto&amp;schema=google" width="700" height="530" frameborder="0" scrolling="no"><a href="https://medium.com/media/d91e1521555bbeaad8b3156aae420375/href">https://medium.com/media/d91e1521555bbeaad8b3156aae420375/href</a></iframe><p>You are free to download, make copies and modify this as you see fit. You could also comment on the Google doc itself if you have suggestions.</p><p>The template is self explanatory. Feedback welcome.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=de1add6345f8" width="1" height="1" alt=""><hr><p><a href="https://medium.com/operational-sympathy/root-cause-analysis-report-template-released-de1add6345f8">Root Cause Analysis — Report template released</a> was originally published in <a href="https://medium.com/operational-sympathy">operational-sympathy</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Turning Wrenches and Diagnosing Systems]]></title>
            <link>https://afkham-azeez.medium.com/turning-wrenches-and-diagnosing-systems-28b63ec5ddbf?source=rss-d32ea2ba6958------2</link>
            <guid isPermaLink="false">https://medium.com/p/28b63ec5ddbf</guid>
            <category><![CDATA[sre]]></category>
            <category><![CDATA[rca]]></category>
            <category><![CDATA[cloud]]></category>
            <category><![CDATA[operational-sympathy]]></category>
            <dc:creator><![CDATA[Afkham Azeez]]></dc:creator>
            <pubDate>Wed, 22 Jan 2025 08:59:58 GMT</pubDate>
            <atom:updated>2026-02-19T08:28:44.772Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/866/1*M-IjoP0nr1WU97-j2cfEcw.png" /></figure><p>From a young age, <a href="https://afkham-azeez.medium.com/the-journey-of-a-builder-b8fbadc18daf">I’ve been fascinated by how things work</a>, especially when it comes to vehicles. Over the years, this passion evolved into a lifelong hobby of tinkering with mechanics, undertaking DIY projects, and performing routine maintenance on my automobiles. The lessons I’ve learned in my garage often inspire the way I approach challenges in my day job, and vice versa. Whether diagnosing a mechanical fault or helping my teams troubleshoot a complex cloud deployment issue, I’ve found that treating symptoms without identifying the true underlying problem rarely works. Instead, taking the time to uncover and address the root cause not only solves the immediate issue but also prevents it from recurring. I’ve observed that people are keen to jump on fixing “something” instead of <a href="https://medium.com/@afkham-azeez/the-curious-case-of-the-leaking-land-rover-38e28758e6f5">taking a step back and asking why</a>, perhaps because they feel that it is a waste of time which they could spend on fixing something. Many years ago, I used to be like this too.</p><p>In my experience, the value of proper Root Cause Analysis (RCA) becomes most evident when things go wrong, whether in the garage, the workplace or any other instance in your life. One of the key principles I follow is to <strong><em>avoid rushing into solutions without a thorough understanding of the problem</em></strong>. This approach has saved me from unnecessary repairs and expenses with my vehicles and has helped my teams resolve intricate problems in the cloud environment. A thorough RCA requires persistence, skepticism, and a willingness to <a href="https://medium.com/@afkham-azeez/the-curious-case-of-the-leaking-land-rover-38e28758e6f5">dig deeper than surface-level symptoms</a>. It is this mindset that has driven me to question initial diagnoses, challenge assumptions, and ultimately find solutions that are both effective and lasting.</p><h4>The misdiagnosed check engine warning</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/804/1*Wmkat56tOrJER-rGFJuHNA.png" /><figcaption><a href="http://mmc-manuals.ru/manuals/pajero_iv/online/Service_Manual/2008_eu/13/html/M113350610019700ENG.HTM">P0628: Suction Control Valve Open</a>, <a href="http://mmc-manuals.ru/manuals/outlander_xl/online/Service_Manual/2011/13/html/M113350700053500ENG.HTM">P1272: Pressure Limiter Malfunction</a></figcaption></figure><p>One memorable example of the importance of RCA occurred when my vehicle developed a starter motor issue. After getting the motor repaired, I noticed new problems emerging — the engine didn’t sound as smooth as it used to, there was an occasional diesel injector noise, and the check engine light came on. Armed with my old and trusty <a href="https://www.ebay.com/b/ELM327-Vehicle-Diagnostic-Tools/179474/bn_116012156">ELM 327</a> adapter and the <a href="https://play.google.com/store/apps/details?id=org.prowl.torque&amp;hl=en">Torque OBD2</a> Android app, I scanned for error codes, which pointed to issues related to the suction control valve (SCV) which controls the diesel injector pump. Clearing the codes were a temporary solution, but the problem resurfaced, and the vehicle eventually went into limp mode. I had a gut feeling that the issue wasn’t as straightforward or serious as the SCV failing because it started showing up only after the starter repair, hence during the removal and refitting, some wiring may have got damaged or dislodged. Despite the garage diagnosing and confirming a faulty SCV in addition to a failed auto gear inhibitor switch which was not connected at all to the error codes, I insisted they dig deeper into the electrical wiring and sockets. My hunch proved correct — after further inspection, they found a loose socket causing the errors. Addressing it resolved the problem entirely, saving me from a costly and unnecessary repair.</p><h4>The mysterious oil leak</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*drXYamGdWtU7cUjw" /></figure><p>Another instance where RCA played a pivotal role was when I faced a mysterious and persistent oil leak in my vehicle. The oil drip on my garage floor wasn’t just a nuisance — it posed a serious concern because I frequently drive long distances late at night. A leaking engine could potentially leave me stranded on the road during those late hours, far from help. So, I took it to a mechanic. Instead of properly diagnosing the issue, they prematurely replaced the rocker cover seal and valve seals, which came at a relatively high price. However, the problem remained unresolved. The garage then suggested that the power steering oil rack seals might be leaking or, even worse, that the engine was burning oil due to wear! Unconvinced, I insisted they not make any further repairs until they could definitively pinpoint the root cause. Consulting another mechanic, he speculated that the leak was due to a faulty gasket in the oil cooler, but again, he couldn’t provide clear evidence. I remained persistent and asked them to inspect the source of the leak with a flashlight. After careful inspection, they discovered that the leak was coming from loose tubes feeding the power steering rack and the oil cooler. Tightening these tubes resolved the problem, saving me from unnecessary repairs and a significant expense.</p><p>Through these experiences, I’ve learned several valuable lessons that apply not only to vehicle repairs but to life and work in general. First and foremost, always dig deeper than the surface-level symptoms. Relying on quick fixes or initial diagnoses can lead to unnecessary costs and missed opportunities for more effective solutions. Secondly, persistence and skepticism are critical — just because an expert or a diagnostic tool suggests one path doesn’t mean it’s the right one. It’s important to trust your instincts and be willing to challenge assumptions. Thirdly, proper RCA is a time investment that pays off in the long run. Taking the extra time to thoroughly understand a problem can save far more time, money, and frustration than rushing to a solution. Finally, RCA isn’t just about solving one problem — it’s about building the mindset of continuous improvement. Whether fixing a mechanical issue or addressing complex cloud incidents, the goal is always the same: to fix the root causes and prevent the problem from recurring instead of masking the symptoms or worst, “fix” something unrelated.</p><p>In conclusion, whether it’s a leaking engine or a cloud service incident, the principles of Root Cause Analysis remain the same: take the time to understand the true source of the problem, and the solution will follow. By applying this approach, I’ve saved myself from unnecessary repairs, avoided costly mistakes, and gained a deeper appreciation for the power of thorough analysis. RCA isn’t just a technical tool — it’s a mindset that, when embraced, can lead to more efficient problem-solving and ultimately prevent recurring issues. As both a mechanic in my garage and a team lead in my professional role, I know that the key to lasting solutions lies not in quick fixes, but in solving the root causes once and for all.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=28b63ec5ddbf" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>