<?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 Marcin Choiński on Medium]]></title>
        <description><![CDATA[Stories by Marcin Choiński on Medium]]></description>
        <link>https://medium.com/@marcin_84085?source=rss-32ce1ef861dd------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/2*34X97oqokxnVAIUEd1XHBg.jpeg</url>
            <title>Stories by Marcin Choiński on Medium</title>
            <link>https://medium.com/@marcin_84085?source=rss-32ce1ef861dd------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 10:37:30 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@marcin_84085/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[Is Data Mess our next failed approach to implementation of a Data Ecosystem?]]></title>
            <link>https://medium.com/@marcin_84085/is-data-mess-our-next-failed-approach-to-implementation-of-a-data-ecosystem-8cc1effb125f?source=rss-32ce1ef861dd------2</link>
            <guid isPermaLink="false">https://medium.com/p/8cc1effb125f</guid>
            <category><![CDATA[data-fabric]]></category>
            <category><![CDATA[data-mesh]]></category>
            <category><![CDATA[olap]]></category>
            <category><![CDATA[data-warehouse]]></category>
            <category><![CDATA[data-management]]></category>
            <dc:creator><![CDATA[Marcin Choiński]]></dc:creator>
            <pubDate>Wed, 19 Oct 2022 20:00:44 GMT</pubDate>
            <atom:updated>2022-10-19T21:51:17.217Z</atom:updated>
            <content:encoded><![CDATA[<h4><strong>…and what’s the next step after that?</strong></h4><h4><strong>History repeats itslef</strong></h4><p>…and it’s no different with Data. Since the introduction of OLAP we are trying to get the Enterprise Data Management and our Data Ecosystems right, but we keep failing. From the helicopter view there seems to be a certain pattern we follow that may suggest what’s the next thing after Data Mesh. Is it a Data Mess?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/958/1*BtQJVf4qB_1v0cqniSVEsw.png" /></figure><h4><strong>The battle between centralization and decentralization</strong></h4><p>When OLAP came in and made it possible to get more value from the Data, the lack of best practices led to multiple <strong>decentralized</strong> heterogenous solutions in different departments and managers proving to one another that they have the right Data (a.k.a. Data Silos). This ended up with companies <strong>centralizing</strong> their efforts and Data teams (a.k.a. BICC, CoE) and comming up with best practices around creating Enterprise Data Warehouses (EDW).</p><p>These didn’t last the test of the Digital era and exponential growth of the amount and different types of Data available to be collected and processed. The lucky ones could afford Data Warehouse Appliances but those also didn&#39;t rise up to all the business expectations. Luckily Google published its papers on Google File System and MapReduce, which helped Doug Cutting and Mike Cafarella turn Apache Nutch into Apache Hadoop. That started the Big Data era and allowed almost every business to processes massive amounts of Data on a commodity harware and with no or relatively little licence costs.</p><p>Unfortunatelly, these solutions known as Data Lakes, though technically centralized, allowed multiple teams to store and process their Data in a logically <strong>decentralized</strong> manner. Lack of best practices quickly turned Data Lakes into Data Swamps with massive amounts of unused and undocumented Data schemas and objects (a.k.a. Dark Data). In addition Data Lakes in early stages lacked alignemnt with EDW. A natural step forward was to come up with best practices, clean the mess and assure single, <strong>centralized</strong> logical view and management approach to those heterogenous analytical Data repositories. This ended up with the Logical Data Warehouse (LDW) / Modern Data Warehouse notion. These were further elevated to the Data Fabric approach.</p><p>Data Fabric in theory addresses all the lessons learned from the previous failures BUT requires skills, resources and maturity that very few (if not close to none) companies have or can afford. Acknowledging the fact that for many companies this is a utopian approach, almost impossible to effectively execute, <a href="https://www.linkedin.com/in/zhamak-dehghani/"><strong>Zhamak Dehghani</strong></a><strong> </strong>came up with (or put together) the idea of a <a href="https://martinfowler.com/articles/data-mesh-principles.html">Data Mesh</a> — a <strong>decentralized</strong> sociotechnical approach to building and managing complex Data ecosystems. Drawing from the Domain Driven Design, it’s based on four key concepts: Domain ownership of Data, Data as a Product, Self-service Data platform and Federated computational governance.</p><p><strong>Is a Data Mess next?</strong></p><p>Even though <a href="https://www.linkedin.com/in/zhamak-dehghani/">the author of the Data Mesh</a> and the whole Data comunity are doing their best to come up with best practices around the concept, dollars to donuts we will fail first. Likewise with OLAP and Data Lakes, we will end up with a Data Mess first. It will take time, blood, sweat and tears, before we get it right, as such <strong>decentralisation</strong>, likewise Data Fabric, will require a lot of skill, resources and maturity of people working with the Data in their domains. In addition (as the history teaches us) I would project that we will also need to find a proper balance <strong>between the centralization and decentralization </strong>of our Data Ecosystems. Even if the next revised concept that will take us out of the Data Mess is just a proper implementation of the Data Mesh approach, again dollars to donuts someone will come up with a new buzzword for it. Time will show.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8cc1effb125f" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>