<?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 Johannes Engelke on Medium]]></title>
        <description><![CDATA[Stories by Johannes Engelke on Medium]]></description>
        <link>https://medium.com/@quablab?source=rss-c48431be7c54------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*i-XimhFYWqiL98JmxMrTGw.jpeg</url>
            <title>Stories by Johannes Engelke on Medium</title>
            <link>https://medium.com/@quablab?source=rss-c48431be7c54------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Fri, 20 Nov 2026 04:59:39 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@quablab/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[Monitor Services in Istio with Metricbeat]]></title>
            <link>https://medium.com/@quablab/monitor-services-in-istio-with-metricbeat-85c531fa93dd?source=rss-c48431be7c54------2</link>
            <guid isPermaLink="false">https://medium.com/p/85c531fa93dd</guid>
            <category><![CDATA[metricbeat]]></category>
            <category><![CDATA[istio]]></category>
            <category><![CDATA[kubernetes]]></category>
            <category><![CDATA[spring-boot]]></category>
            <category><![CDATA[elastic-stack]]></category>
            <dc:creator><![CDATA[Johannes Engelke]]></dc:creator>
            <pubDate>Thu, 23 Jun 2022 12:57:55 GMT</pubDate>
            <atom:updated>2022-06-23T12:57:55.813Z</atom:updated>
            <content:encoded><![CDATA[<p>The de-facto standard for metrics in kubernetes clusters is prometheus. This might be a problem in services meshes, especially if mutual TLS is activated. Direct access to the prometheus endpoints is not possible any more as istio is build for a service to service communication and direct access to pods and containers is blocked. The purpose of this article is to describe how to scrape process metrics using metricbeat in a istio service mesh with enabled mutual TLS.</p><h3>Architecture Overview</h3><p>If MTLS is enabled in istio, it is hard to directly scrape the metrics endpoint. To achieve this, the agent calling the metrics endpoint needs to present a <a href="https://istio.io/latest/docs/ops/integrations/prometheus/#option-2-customized-scraping-configurations">proper certificate to the istio proxy</a>. The setup can be complex and some monitoring tools are not supporting it. To overcome this problem, the istio-proxy has the capability to include (or merge) the application metrics into the proxy-metrics. This will allow the monitoring agent to call the istio-proxy metrics endpoint to get the proxy and application metrics at the same time.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*zbBKrjS2JP9FF2wQ6MV63A.png" /><figcaption>Application metric integration in istio-proxy</figcaption></figure><p>To enable metric merging, istio is looking for <a href="https://github.com/helm/charts/tree/master/stable/prometheus#scraping-pod-metrics-via-annotations">prometheus.io pod annotations</a>. The benefit of this setup is the simplicity. As soon as the metrics are exposed, we can simply call the isto-proxy metrics endpoint from within the mesh. No further configuration is needed. The drawback, metricsbeat has to be part of the services mesh. This might create problems calling services outside the mesh. Beside that, the application can overwrite existing istio metrics (e.g. istio_requests_total) and an eventually running prometheus instance has to be configured to ignore scrape annotations (or use custom ones).</p><h3>Configuration</h3><p>As an example we are using a simple spring boot application. Due to the actuator it is possible to expose health and readiness endpoints as well as metrics in prometheus format by configuration.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/d225a117aff6beb43eb677410415cb00/href">https://medium.com/media/d225a117aff6beb43eb677410415cb00/href</a></iframe><p>The actuator is exposed on port :8081. The prometheus metrics are available under /manage/prometheus. Health and readiness endpoints under /manage/health. The application is deployed to kubernetes using the following configuration:</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/5f9699d72ac55ade29670c25d5e53ca1/href">https://medium.com/media/5f9699d72ac55ade29670c25d5e53ca1/href</a></iframe><p>The prometheus.io/scrape, prometheus.io/pathand prometheus.io/port annotations are pointing the istio-proxy to the metrics endpoint, which should be included in the pods metrics. In the next step, we can instruct metricbeat to scrape the metrics endpoint.</p><p>Metricbeat has a dedicated module for istio, which is gathering a lot of insides from the services mesh. As this module is covering the standard istio-proxy metrics already, only services with custom metrics needs to be scraped. At the same time, all istio related metrics can be ignored, to not store them twice.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/26ea47d3d76b206ff12505e1897bb312/href">https://medium.com/media/26ea47d3d76b206ff12505e1897bb312/href</a></iframe><p>The configuration is instructing metricbeat to only scrape pods with the prometheus.io/scrape annotation set to true. If the annotation is set, metricbeat will call the istio-proxy metrics endpoint. The filter will exclude all istio related metrics as they are already stored.</p><h3>Summary</h3><p>Using standard prometheus metrics and annotations helps us to build an easy to use solution which can be easily adapted to other monitoring tools. Due to the annotations the solution can be consumed in a self services and generic way. The only drawback which is currently unaddressed is the fact, that services can overwrite istio metrics. This problem is mitigated, as the scraping configuration is ignoring istio metrics in favor of the native metricbeat-istio module.</p><h3>Resources</h3><ul><li>Istio Prometheus Documentation: <a href="https://istio.io/latest/docs/ops/integrations/prometheus/">https://istio.io/latest/docs/ops/integrations/prometheus/</a></li><li>Autodiscovery Metricbeat: <a href="https://www.elastic.co/guide/en/beats/metricbeat/current/configuration-autodiscover.html">https://www.elastic.co/guide/en/beats/metricbeat/current/configuration-autodiscover.html</a></li><li>Metricbeat Prometheus Module: <a href="https://www.elastic.co/guide/en/beats/metricbeat/current/metricbeat-metricset-prometheus-collector.html">https://www.elastic.co/guide/en/beats/metricbeat/current/metricbeat-metricset-prometheus-collector.html</a></li><li>Spring Actuator: <a href="https://docs.spring.io/spring-metrics/docs/current/public/prometheus">https://docs.spring.io/spring-metrics/docs/current/public/prometheus</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=85c531fa93dd" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Growth — My Third Month at DPS]]></title>
            <link>https://medium.com/digital-product-school/growth-my-third-month-at-dps-2a01a6d6ed76?source=rss-c48431be7c54------2</link>
            <guid isPermaLink="false">https://medium.com/p/2a01a6d6ed76</guid>
            <category><![CDATA[digital-product-school]]></category>
            <category><![CDATA[agile]]></category>
            <category><![CDATA[lean-startup]]></category>
            <category><![CDATA[design-thinking]]></category>
            <category><![CDATA[product-management]]></category>
            <dc:creator><![CDATA[Johannes Engelke]]></dc:creator>
            <pubDate>Tue, 04 Aug 2020 15:26:42 GMT</pubDate>
            <atom:updated>2020-08-31T09:14:24.048Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*sBLzSh-EDIOyug9nLniwig.jpeg" /></figure><h3>Growth — My Third Month at DPS</h3><p>The third and last month at Digital Product School was focused on implementation. The validated ideas and click dummies have been transferred into working code and first test customers could use the final software. For the product managers, it was time to calm down, start to work on the business model, and think about how to grow the idea.</p><p>For a startup, the business model is probably the most interesting topic as this is the point, where an idea is turned into revenue. In the past, the revenue model was pretty straight forward. The company is selling a good and the consumer/user is paying for it. For many products, even for classic licensed based software packages, this is still a valid model. For multi-tenant cloud applications, there are many more options to generate revenue. Just to mention a few examples:</p><ul><li>Facebook itself is for free, they are not selling to the user, they are selling to companies who like to have more visibility</li><li>PatientsLikeMe offers a free service to patients but charges pharma companies to get access to the data and to improve their products</li><li>Bosch offers vehicle diagnostic services and recommends partner companies to fix the findings</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/676/1*8H6SCw5q0FCyFfq1mEBdqg.png" /></figure><p>When we analyzed our potential customers, we figured out, that we have at least three involved parties and some of them would be willing to pay for our service. This is great as we not only have one revenue stream, we have at least two.</p><p>At this stage, we had sorted out two major questions: What to build and how to make money out of it. The next stop find customers and grow.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-3fuX6FDqEBBbWyxU4EkWQ.jpeg" /></figure><p>For startups, you start with one customer at a time and grow step by step. Just keep in mind if you have one customer in the first year and two in the second you have had 100% growth. This allows you to stay focused on the first customer and later do a few adjustments to fit the needs of the second one and so on. You stay lean and agile.</p><p>In big corporations, your product is exposed to thousands of customers from day one. The benefit is, you can grow really quickly and bring your idea in front of many users. Instead of convincing every customer one by one you can set trends and generate a run on your product. Have a look at how Apple positioned the iPhone. Nobody was looking for such a phone, but Apple has been able to create a completely new market with a lot of demand. The downside is, your product has to be generic and stable enough to serve all those customers well. This is the place where company projects are slowed down as you have to forecast the different cases and build a product that is able to deal with most of them. Finding the right balance to release as early as possible and be ready for the demand is the key factor for corporations.</p><p>What are the learnings out of the last three months? What are the things I will keep in mind going back to my daily work at SAP? First of all, I have learned many tools, tips, and tricks I could use in my daily routine. But more importantly, I had time to reflect on the strengths and weaknesses startups and corporations are facing. Where are the benefits of one model vs the other? I am still not finished in understanding and learning but keeping them in mind will allow me to better leverage the strength of a corporation and take some ideas from startups to iron out some of the weaknesses of a big corporation.</p><p><strong>This is the third article in a series about my time at the Digital Product School.</strong></p><ul><li><em>If you are interested in learning more about the Digital Product School or even apply for a training slot: </em><a href="https://digitalproductschool.io/"><em>https://digitalproductschool.io/</em></a></li><li><em>If you are interested in the first article: </em><a href="https://leaks.digitalproductschool.io/my-first-month-at-digital-product-school-43951f14a5fb"><em>My First Month at Digital Product School</em></a></li><li><em>If you are interested in the second article: </em><a href="https://leaks.digitalproductschool.io/validation-my-second-month-at-dps-5596cb94f6b4"><em>Validation — My Second Month at DPS</em></a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=2a01a6d6ed76" width="1" height="1" alt=""><hr><p><a href="https://medium.com/digital-product-school/growth-my-third-month-at-dps-2a01a6d6ed76">Growth — My Third Month at DPS</a> was originally published in <a href="https://medium.com/digital-product-school">Digital Product School</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Validation — My Second Month at DPS]]></title>
            <link>https://medium.com/digital-product-school/validation-my-second-month-at-dps-5596cb94f6b4?source=rss-c48431be7c54------2</link>
            <guid isPermaLink="false">https://medium.com/p/5596cb94f6b4</guid>
            <category><![CDATA[digital-product-school]]></category>
            <category><![CDATA[design-thinking]]></category>
            <category><![CDATA[validation]]></category>
            <category><![CDATA[product-management]]></category>
            <dc:creator><![CDATA[Johannes Engelke]]></dc:creator>
            <pubDate>Tue, 04 Aug 2020 15:24:04 GMT</pubDate>
            <atom:updated>2020-08-06T09:43:29.901Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*JekLkvP1RwHVRFjjs3-pjg.jpeg" /></figure><h3>Validation — My Second Month at DPS</h3><p>In the retrospective, the first month was mainly about digging into the problem space. Understand the different stakeholder and their needs, goals, and challenges. At the end of the first month, we had a rough idea of how a product could look like. From Product Management perspective the second month was all about validating if our idea is addressing the needs of the users.</p><p>The final validation of an Idea is if the Product we are building is used by paying customers and is exceeding the revenue expectations. As it is expensive and time-consuming to develop a product to production level, it is important to find proof of an idea in advance.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xFg_PANx6n0B9P7Phz_MlA.jpeg" /></figure><p>Every product is based on assumptions. For instance, a car is based on the assumption, that people like to travel individually over long distances. This is a validated assumption, as people are doing it. New and innovative products are based on other assumptions, which are sometimes not as obvious. To predict the success of a product, it is important to uncover the assumptions the product is based on and validate if they are true.</p><p>So what are assumptions? Assumptions are guesses about how users (people, customers) will behave, what problems they are facing, or certain needs they have. One example of an assumption we had was:</p><blockquote>Customers like to attend in-store events</blockquote><p>How can we validate this assumption? The easiest way is to ask potential customers if they would be interested in such an event. This can be done in form of an interview or online survey. Another way is to execute small experiments. There are many different forms of experiments. E.g. create a landing page and check how many people are interacting with the page. Another option is to fake a certain functionality and check how many people trying to use it.</p><p>For the assumption mentioned above, we decided to ask customers what they think. It turns out, that our assumption was not detailed enough. There are many store events and dependent on the type of event the answer was a different one. This generated a better understanding of the problem domain and helped us improve our product.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/600/0*LzpvVibFHQb18o5_.jpg" /></figure><p>Most of the techniques described here can be applied inside the corporate context of SAP, too. Everybody can start validating assumptions and start changing his behavior based on it. The problem is, doing it is relentless hard as it is like asking somebody to smack you in the face with a baseball bat. If you ask the right questions there is always the chance that the answer will destroy your new and shiny idea. The good thing is, as it is just an idea in this early stage, you throw away a few post-its and can search for a new one.</p><p>If we are going through the pain of doing assumption validation right, we should start to harvest and manage the results better. The sources of feedback are endless. Using these resources and encountering them inside our products will make us stronger and better.</p><p><strong>This is the second article in a series about my time at the Digital Product School.</strong></p><ul><li><em>If you are interested in learning more about the DPS or even apply for a training slot: </em><a href="https://digitalproductschool.io/"><em>https://digitalproductschool.io/</em></a></li><li><em>If you are interested in the first article: </em><a href="https://leaks.digitalproductschool.io/my-first-month-at-digital-product-school-43951f14a5fb"><em>My First Month at Digital Product School</em></a></li><li><em>If you are interested in the third article: </em><a href="https://leaks.digitalproductschool.io/growth-my-third-month-at-dps-2a01a6d6ed76"><em>Growth — My Third Month at DPS</em></a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5596cb94f6b4" width="1" height="1" alt=""><hr><p><a href="https://medium.com/digital-product-school/validation-my-second-month-at-dps-5596cb94f6b4">Validation — My Second Month at DPS</a> was originally published in <a href="https://medium.com/digital-product-school">Digital Product School</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[My First Month at Digital Product School]]></title>
            <link>https://medium.com/digital-product-school/my-first-month-at-digital-product-school-43951f14a5fb?source=rss-c48431be7c54------2</link>
            <guid isPermaLink="false">https://medium.com/p/43951f14a5fb</guid>
            <category><![CDATA[customer-experience]]></category>
            <category><![CDATA[digital-product]]></category>
            <category><![CDATA[education]]></category>
            <category><![CDATA[digital-product-school]]></category>
            <category><![CDATA[product-management]]></category>
            <dc:creator><![CDATA[Johannes Engelke]]></dc:creator>
            <pubDate>Mon, 08 Jun 2020 07:49:55 GMT</pubDate>
            <atom:updated>2020-06-08T09:27:50.263Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="The team canvas we use to colaborate in times of COVID-19" src="https://cdn-images-1.medium.com/max/1024/1*-ys4mGZdY7njBop7GEDIIQ.png" /></figure><p>At the beginning, I like to give you some context. Since a couple of years, I am working as a Product Manager at SAP Customer Experience. It is an interesting and challenging job that I really love to-do. Nevertheless, at some point I had the feeling I need to step out of my current role to get some new perspectives. This was a point in time, when a co-worker pointed my interest to a Fellowship position at the Digital Product School (DPS).</p><p>The DPS is an education format for young professionals to teach them the tools and skills required to create digital products. In the course of 3 months the participants are working on a challenge given by e.g a company (if they are in a company team) and getting introduced to the required tools and methods to solve the challenge in a learning by doing fashion. This enables them to come up with new ideas without the restrictions a corporate environment normally brings.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*HsIDd2haHeG5MLgpRoENVA.jpeg" /><figcaption>Mini stories</figcaption></figure><p>The first exercise in the DPS was the Design sprint. I would explain a Design Sprint as a Design Thinking process squeezed in one week. We started digging into the problem space, did first user interviews and synthesised the feedback. In the second half of the week, we started creating ideas how to address the problem and build a first prototype which we could test at the end of the week with a potential user. The beauty of the Design Sprint is, the entire team has an overview of the Design Thinking process and we learned to put the user in the center of all our activities.</p><p>In the following weeks, we have been able to leverage the knowledge we gained during the design sprint and reiterate over the process to improve our ideas and prototypes. During this process we have been introduced to many new tools and methods, which are helping us in the process of generating and validating new ideas.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*H2K401aUNvV8oUPGl0rP7g.png" /></figure><p>When it comes to my personal key takeaway: The role of a product manager is different dependent on the context, he is working in. For example, inside a big corporation, a Product Manager is way more involved in internal meetings, than he would be inside a lean startup. The reason is, in a startup he can fokus much more on the customer it self. Inside a corporate environment you are serving way more customers. In case of SAP you have about 440k for example. That means most of the time, you are syncing with different departments like marketing, service and support, education, sales etc. to assure they are aware of the new thing and know how to answer customer requests appropriately, as this cannot be done by the product management team allone.</p><p>Beside that, a lot of development effort goes into topics like scalability, internationalization and integration. This enables you to serve many customers right from the beginning but brings the danger of loosing the focus on the customer. The tools, methods and mindset, set DPS introduced are helping to keep the focus.</p><p>If you are interested in learning more about the DPS or even apply for a training slot: <a href="https://digitalproductschool.io/">https://digitalproductschool.io/</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=43951f14a5fb" width="1" height="1" alt=""><hr><p><a href="https://medium.com/digital-product-school/my-first-month-at-digital-product-school-43951f14a5fb">My First Month at Digital Product School</a> was originally published in <a href="https://medium.com/digital-product-school">Digital Product School</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>