<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: Aurora</title>
    <description>The latest articles on DEV Community by Aurora (@datalearnerchou).</description>
    <link>https://dev.to/datalearnerchou</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4253427%2F36f1d1c8-6682-45c1-ba15-fba11984a723.jpg</url>
      <title>DEV Community: Aurora</title>
      <link>https://dev.to/datalearnerchou</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/datalearnerchou"/>
    <language>en</language>
    <item>
      <title>From Psychology to Metadata: My First Day Learning Apache Gravitino</title>
      <dc:creator>Aurora</dc:creator>
      <pubDate>Thu, 19 Nov 2026 10:02:46 +0000</pubDate>
      <link>https://dev.to/datalearnerchou/from-psychology-to-metadata-my-first-day-learning-apache-gravitino-362</link>
      <guid>https://dev.to/datalearnerchou/from-psychology-to-metadata-my-first-day-learning-apache-gravitino-362</guid>
      <description>&lt;p&gt;I’m coming from a psychology background with almost zero technical foundation. Recently, I started learning about &lt;a href="https://github.com/apache/gravitino" rel="noopener noreferrer"&gt;Apache Gravitino&lt;/a&gt;. Instead of jumping straight into installation and getting lost in terminal errors, I decided to spend day one just understanding the big picture.&lt;/p&gt;

&lt;p&gt;In psychology, we talk a lot about &lt;strong&gt;mental schemas&lt;/strong&gt;—the cognitive maps our brains use to organize scattered information so we can make sense of the world without getting overwhelmed. Reading through the docs today, it clicked for me that a metadata system is basically a shared cognitive map for a company’s data.&lt;/p&gt;

&lt;p&gt;Here are my learning notes from day one.&lt;/p&gt;

&lt;h3&gt;
  
  
  What problem is Gravitino trying to solve?
&lt;/h3&gt;

&lt;p&gt;In modern data and AI setups, assets end up scattered across completely different tools:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Structured tables live in databases like MySQL or PostgreSQL&lt;/li&gt;
&lt;li&gt;Large-scale analytical tables sit in data lakes (Hive, Iceberg, Paimon, etc.)&lt;/li&gt;
&lt;li&gt;Unstructured files (like images or raw training data) are stored on HDFS or S3&lt;/li&gt;
&lt;li&gt;Real-time data streams flow through systems like Kafka&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Because these tools are separate, the &lt;strong&gt;metadata&lt;/strong&gt;—the "labels" and descriptions telling you what data exists, where it lives, and who is allowed to touch it—gets fragmented too. That makes finding and safely using data really messy, especially when AI projects need to pull from multiple sources at once.&lt;/p&gt;

&lt;p&gt;Gravitino’s idea is to act as a unified metadata layer (a federated metadata lake). It connects to these different sources and manages their metadata under one roof, &lt;strong&gt;without&lt;/strong&gt; having to copy or move the actual data.&lt;/p&gt;

&lt;h3&gt;
  
  
  The core hierarchy I learned
&lt;/h3&gt;

&lt;p&gt;To organize everything, Gravitino uses a pretty clean four-level structure:&lt;br&gt;
&lt;/p&gt;

&lt;pre data-lang="mermaid"&gt;&lt;code&gt;graph LR
    A[Metalake] --&amp;gt; B[Catalog]
    B --&amp;gt; C[Schema]
    C --&amp;gt; D[Tables]
    C --&amp;gt; E[Filesets]
    C --&amp;gt; F[Topics]
    C --&amp;gt; G[Models]&lt;/code&gt;&lt;/pre&gt;



&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Metalake:&lt;/strong&gt; The top-level container (think of it as the whole workspace for an organization or environment).&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Catalog:&lt;/strong&gt; Groups different types of data sources together.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Schema:&lt;/strong&gt; Organizes assets inside a catalog, similar to folders.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Assets:&lt;/strong&gt; The actual things teams work with—relational &lt;strong&gt;tables&lt;/strong&gt;, unstructured &lt;strong&gt;filesets&lt;/strong&gt;, Kafka &lt;strong&gt;topics&lt;/strong&gt;, and AI &lt;strong&gt;models&lt;/strong&gt;.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;One thing that stood out to me is that it doesn't just stop at traditional database tables. Including filesets, messaging topics, and AI models in the same hierarchy makes a lot of sense for how people actually build data + AI projects today.&lt;/p&gt;

&lt;h3&gt;
  
  
  My takeaway so far
&lt;/h3&gt;

&lt;p&gt;Gravitino isn’t trying to be a giant storage bucket that swallows all your data. It’s more like a single directory that lets you see, organize, and manage everything across different systems—without moving the actual files.&lt;/p&gt;

&lt;p&gt;I’m still at the very beginning of this journey. Next, I plan to spin it up locally and see how these concepts actually look in practice.&lt;/p&gt;




&lt;p&gt;&lt;em&gt;Any tips for a non-CS beginner diving into open-source data infra?&lt;/em&gt;&lt;/p&gt;




&lt;p&gt;&lt;em&gt;P.S. By the way, I had Paigu Niangao (crispy fried pork chop with chewy rice cakes) at my campus cafeteria today—it was delicious! Definitely good fuel for learning metadata.&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F8xm9bzjuoz2gxu7xvdaw.jpg" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F8xm9bzjuoz2gxu7xvdaw.jpg" alt="Paigu Niangao - Crispy pork chop with rice cakes" width="800" height="1067"&gt;&lt;/a&gt;&lt;/p&gt;

</description>
      <category>opensource</category>
      <category>ai</category>
      <category>beginners</category>
      <category>kafka</category>
    </item>
  </channel>
</rss>
