<?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 Dilukshi Sangeetha Wijayawardhana on Medium]]></title>
        <description><![CDATA[Stories by Dilukshi Sangeetha Wijayawardhana on Medium]]></description>
        <link>https://medium.com/@dilukshi1017?source=rss-b92ccbe88d13------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*njW5O6KTCT0XQTaA</url>
            <title>Stories by Dilukshi Sangeetha Wijayawardhana on Medium</title>
            <link>https://medium.com/@dilukshi1017?source=rss-b92ccbe88d13------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Thu, 19 Nov 2026 06:00:15 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@dilukshi1017/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[Networking Fundamentals: From Networks to Data Transmission]]></title>
            <link>https://medium.com/@dilukshi1017/networking-fundamentals-from-networks-to-data-transmission-fb51c099b186?source=rss-b92ccbe88d13------2</link>
            <guid isPermaLink="false">https://medium.com/p/fb51c099b186</guid>
            <category><![CDATA[internet]]></category>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[networking]]></category>
            <dc:creator><![CDATA[Dilukshi Sangeetha Wijayawardhana]]></dc:creator>
            <pubDate>Tue, 29 Sep 2026 07:23:56 GMT</pubDate>
            <atom:updated>2026-09-29T07:23:56.421Z</atom:updated>
            <content:encoded><![CDATA[<h3>Learning Computer Networking from Zero</h3><p>When we use the internet, we usually think about websites, YouTube, cloud applications, or social media. But behind every click, message, video, and application is something much more fundamental: <strong>Computer Networking</strong>.</p><p>Before learning advanced topics such as IP addresses, subnetting, routing, DNS, TCP/IP, firewalls, and cloud networking, we need to understand the basic idea:</p><blockquote><strong><em>How do computers and other devices communicate with each other?</em></strong></blockquote><p>This article covers networks, the Internet, connected devices, data, bits, signals, bandwidth, and throughput.</p><h3>1. What Is a Computer Network?</h3><p>A computer network is a group of two or more electronic devices connected together so they can communicate and share resources.</p><p>The connection can happen through:</p><ul><li>Network cables (Ethernet)</li><li>Wi-Fi</li><li>Fiber-optic cables</li><li>Cellular networks (4G, 5G)</li><li>Other wireless technologie</li></ul><pre>                              Office Network<br>                                    │<br>                            ┌───────┴───────┐<br>                            │    Switch     │<br>                            └───────┬───────┘<br>                               │    │    │<br>                               ▼    ▼    ▼<br>                             PC   Printer Server</pre><p>The computers can communicate with the server and share the printer because they are connected through a network.</p><blockquote><strong><em>A simple definition:</em></strong><em> A network allows devices to communicate and share resources. This is the foundation of modern IT.</em></blockquote><h3>2. Who Owns the Internet?</h3><p>A common question is: <em>Who owns the Internet?</em></p><p>The answer is surprisingly simple: <strong>No single person or company owns the entire Internet.</strong> The Internet is a <strong>network of networks</strong>.</p><p>Think about the world’s road system: no single company owns every road in every country. Local governments own city streets, while national highways connect cities together. The Internet works the same way.</p><figure><img alt="Internet" src="https://cdn-images-1.medium.com/max/600/1*_t1vma6OJSugWqtNc-5KkQ.jpeg" /><figcaption>A network of networks</figcaption></figure><h4>Internet Service Providers (ISPs)</h4><p>ISPs provide connectivity to homes, offices, and mobile phones using fiber cables, cell towers, and neighborhood routing gear.</p><h4>Backbone Providers</h4><p>Large telecommunications companies operate massive high-capacity lines across countries and oceans via deep-sea fiber cables.</p><h4>Technology Companies</h4><p>Companies such as Google, Microsoft, and Amazon run gigantic data centers packed with servers hosting websites, videos, and cloud storage.</p><h3>3. If Nobody Owns the Internet, Who Manages It?</h3><p>Because there is no single owner, computers around the world must follow <strong>common technical standards — </strong>just like traffic laws allow cars from different countries to drive safely.</p><ul><li><strong>ICANN (Internet Corporation for Assigned Names and Numbers):</strong> Coordinates domain names (like google.com) and numerical IP addresses so they never clash.</li><li><strong>IETF (Internet Engineering Task Force):</strong> Documents and develops the open rules (protocols) that dictate how computers send data across wires and airwaves.</li></ul><h3>4. What Devices Connect to Networks?</h3><p>Networks connect far more than just desktop computers. We categorize them into:</p><ol><li>Local networks (LAN)</li><li>Mobile devices</li><li>Connected home devices (Smart Home)</li><li>Internet of Things (IoT)</li></ol><h3>5. Local Networks — LAN</h3><p>A <strong>LAN (Local Area Network)</strong> connects devices within a limited physical area, such as a home, school computer lab, or office building.</p><figure><img alt="E.g.: LAN" src="https://cdn-images-1.medium.com/max/736/1*W_1vnN10q3LysDOvKzf2PA.jpeg" /><figcaption>Local Area Network</figcaption></figure><h3>Real-World Example: A School Computer Lab</h3><p>In a room with 30 student computers, 1 teacher computer, and 1 shared printer, all devices plug into a central switch. Every student can send their homework directly to the printer without physically moving files with a USB drive.</p><h3>6. Mobile Devices</h3><p>Modern networks allow devices to stay connected while moving:</p><ul><li><strong>Smartphones &amp; Tablets</strong></li><li><strong>Laptops</strong></li><li><strong>Smartwatches &amp; Fitness Trackers</strong></li></ul><p>These rely on wireless access points (Wi-Fi) or cell towers (4G/5G) so you can walk outside listening to a streaming podcast without interruption.</p><h3>7. Connected Home Devices &amp; IoT</h3><p>Traditional household appliances now feature network adapters to report data or accept remote commands.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/735/1*tUTfzA1pZOWyds0hOWlCBg.jpeg" /><figcaption>Smart Home — Automation</figcaption></figure><h3>Internet of Things (IoT)</h3><p>Beyond homes, factories and farms use sensors to gather data automatically:</p><ul><li><strong>Supermarket Refrigerators:</strong> A temperature sensor detects if a freezer fails, sending an immediate alert to a manager’s phone before food spoils.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/736/1*bozmgK-hAn1fQbbhMMjoyQ.jpeg" /><figcaption>Supermarket Refrigerators</figcaption></figure><ul><li><strong>Smart Agriculture:</strong> Soil moisture sensors trigger irrigation valves only when crops need water.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/736/1*6hZWXq6ZrgC9_ZQ-KKwWXw.jpeg" /><figcaption>Smart Agriculture</figcaption></figure><h3>8. Understanding Data: Bits and Bytes</h3><p>What actually travels through a network? <strong>Data.</strong></p><p>Computers do not read words, songs, or photos directly. They store and process everything using the <strong>binary system</strong> (0s and 1s).</p><ul><li><strong>Bit (Binary Digit):</strong> The smallest unit of information. Like a light switch:</li><li>0 = OFF</li><li>1 = ON</li></ul><p><strong>Byte: A group of 8 bits.</strong></p><pre>                             <strong> 1 Byte = 8 Bits</strong></pre><p>For example, the letter <strong>A</strong> in standard ASCII binary is:</p><p>01000001</p><p>When you type the word <strong>HELLO</strong>, your computer breaks each letter down into an 8-bit byte before sending it.</p><h3>9. How Are Bits Transmitted?</h3><p>To travel across a room or across an ocean, bits must be converted into physical signals over a medium:</p><ol><li><strong>Electrical Signals (Copper Cable / Ethernet):</strong> High and low electrical voltages represent 1s and 0s.</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*aXGan4euVdq2su6uBwUxjg.jpeg" /><figcaption>Electrical Signals (Copper Cable / Ethernet)</figcaption></figure><p><strong>2. Optical Signals (Fiber-Optic Cable):</strong> Rapid pulses of light flash through flexible glass strands over vast distances.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*vvmECre80DR7BnMvj5qWlw.jpeg" /><figcaption>Optical Signals (Fiber-Optic Cable)</figcaption></figure><p><strong>3. Wireless Signals (Radio Waves):</strong> Wi-Fi, cellular, and Bluetooth modulate electromagnetic waves through the air.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xx1a4L_ZcYFfMv2timjlIg.jpeg" /><figcaption>Wireless Signals (Radio Waves)</figcaption></figure><h3>10. Bandwidth, Throughput, and Latency</h3><p>These three terms are often confused, but they describe very different network properties.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ZrIq3LHNPE3dvWiROzlIpA.jpeg" /><figcaption>Bandwidth, Throughput, and Latency</figcaption></figure><h3>Bandwidth = The Width of the Pipe</h3><p>Bandwidth is the <strong>maximum theoretical capacity</strong> of your link — how much data <em>could</em> move through per second.</p><ul><li>Measured in <strong>Mbps</strong> (Megabits per second) or <strong>Gbps</strong> (Gigabits per second).</li><li>A wider highway lets more cars travel side-by-side at once.</li></ul><h3>Throughput = The Actual Water Flow</h3><p>Throughput is the <strong>actual amount of data</strong> that successfully reaches its destination each second.</p><ul><li>If your internet plan is 100 Mbps (bandwidth), but heavy network congestion or Wi-Fi interference slows you down, your actual speed might be 70 Mbps (throughput).</li></ul><h3>Latency = Travel Delay</h3><p>Latency is the time it takes for a single packet of data to make the round trip between your computer and the server.</p><ul><li>High bandwidth does not mean low latency. You could have a wide highway, but if the distance is far, travel still takes time.</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=fb51c099b186" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Database Backups: Protecting Your Data and Preparing for the Unexpected]]></title>
            <link>https://medium.com/@dilukshi1017/database-backups-protecting-your-data-and-preparing-for-the-unexpected-d2617f4028b1?source=rss-b92ccbe88d13------2</link>
            <guid isPermaLink="false">https://medium.com/p/d2617f4028b1</guid>
            <category><![CDATA[data-backup]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[automation]]></category>
            <category><![CDATA[database]]></category>
            <category><![CDATA[dbops]]></category>
            <dc:creator><![CDATA[Dilukshi Sangeetha Wijayawardhana]]></dc:creator>
            <pubDate>Tue, 22 Sep 2026 05:43:39 GMT</pubDate>
            <atom:updated>2026-09-22T05:43:39.535Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*7uRoVU37J8N7lHlGMxQMNQ.png" /></figure><p>Imagine this. You are working as a Database Engineer. It is 3:00 PM on Monday.</p><p>Your production database contains thousands or millions of customer records.</p><p>Suddenly, someone accidentally runs:</p><pre>DELETE FROM orders;</pre><p>The data is gone.</p><p>Or perhaps the server disk fails.</p><p>Or ransomware encrypts your database.</p><p>Or a software deployment corrupts important records.</p><p>The first question from management will probably be:</p><blockquote><strong><em>“Can we recover the data?”</em></strong></blockquote><p>This is where <strong>database backups</strong> become one of the most important responsibilities in DBOPS.</p><p>A database is not truly protected just because it is running successfully.</p><p>A production database needs a <strong>recovery plan</strong>.</p><p>And backups are one of the most important parts of that plan.</p><h3>1. What Is a Database Backup?</h3><p>A <strong>database backup is a copy of database data and related information that can be used to restore the database after data loss, corruption, or failure.</strong></p><p>In simple words:</p><blockquote><strong><em>Backup = A safety copy of your database.</em></strong></blockquote><p>For example, suppose we have:</p><pre>Production MySQL Database</pre><pre>customers<br>orders<br>payments<br>products</pre><p>We create a backup:</p><pre>BACKUP<br>                │<br>                ▼<br>┌─────────────────────────────┐<br>│      Production Database    │<br>│                             │<br>│ customers                   │<br>│ orders                      │<br>│ payments                    │<br>│ products                    │<br>└──────────────┬──────────────┘<br>               │<br>               │ Backup<br>               ▼<br>┌─────────────────────────────┐<br>│       Backup Storage        │<br>│                             │<br>│ database_backup.sql         │<br>│ / backup files              │<br>└─────────────────────────────┘</pre><p>If the production database fails, we can use the backup to restore the data.</p><h3>2. Why Do We Need Database Backups?</h3><p>Because <strong>things go wrong</strong>.</p><p>A database can be affected by many different problems.</p><h3>Human mistakes</h3><p>For example:</p><pre>DELETE FROM customers;</pre><p>without the correct WHERE condition.</p><h3>Hardware failure</h3><p>A production server’s disk can fail.</p><h3>Software failure</h3><p>A database upgrade or application deployment can introduce problems.</p><h3>Cyber attacks</h3><p>Ransomware or other attacks can damage or encrypt data.</p><h3>Data corruption</h3><p>Database files or records can become corrupted.</p><h3>Natural disasters</h3><p>Fire, flooding, power failure, or other disasters can affect physical infrastructure.</p><p>Therefore:</p><pre>Production Database<br>                     │<br>        ┌────────────┼────────────┐<br>        │            │            │<br>     Human        Hardware      Cyber<br>     Error         Failure       Attack<br>        │            │            │<br>        └────────────┼────────────┘<br>                     ▼<br>                 DATA LOSS<br>                     │<br>                     ▼<br>                  BACKUP<br>                     │<br>                     ▼<br>                  RECOVERY</pre><p>A backup gives the organization a way to recover.</p><h3>3. Backup vs Recovery</h3><p>These two words are closely related, but they are not the same.</p><h3>Backup</h3><p>Creating a copy of the data.</p><pre>Production DB<br>      │<br>      ▼<br>   BACKUP<br>      │<br>      ▼<br>Backup Storage</pre><h3>Recovery</h3><p>Using that backup and other recovery information to bring the database back.</p><pre>Database Failure<br>      │<br>      ▼<br>Find Backup<br>      │<br>      ▼<br>Restore<br>      │<br>      ▼<br>Verify Data<br>      │<br>      ▼<br>Database Available</pre><p>So:</p><blockquote><strong><em>Backup is the preparation. Recovery is the action.</em></strong></blockquote><p>This is an important DBOPS concept.</p><p>A company can have backups and still have a poor recovery process.</p><p>That is why Database Engineers should regularly <strong>test restoration</strong>.</p><h3>4. Main Categories of Database Backups</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*9kmNxesxgkh6nE0q5of9jw.jpeg" /></figure><p>There are several ways to classify backups.</p><p>The most common concepts are:</p><ol><li>Full Backup</li><li>Incremental Backup</li><li>Differential Backup</li><li>Logical Backup</li><li>Physical Backup</li></ol><p>There is also another important concept:</p><ol><li>Binary Logs / Transaction Logs for Point-in-Time Recovery</li></ol><p>Let’s understand them one by one.</p><h3>5. Full Backup</h3><p>A <strong>full backup</strong> contains a complete copy of the database at a particular point in time.</p><p>For example:</p><pre>Sunday 12:00 AM</pre><pre>Database<br>│<br>├── customers<br>├── orders<br>├── payments<br>├── products<br>└── users</pre><pre>↓</pre><pre>FULL BACKUP</pre><p>The backup contains everything needed from that backup set to reconstruct the database.</p><h3>Example</h3><p>Suppose our database is:</p><pre>100 GB</pre><p>A full backup might contain approximately:</p><pre>100 GB</pre><p>depending on compression and the backup method.</p><h3>Advantages</h3><ul><li>Simple concept</li><li>Easy to manage</li><li>Easier to restore</li><li>Complete backup set</li></ul><h3>Disadvantages</h3><ul><li>Requires more storage</li><li>Takes longer</li><li>Can consume significant I/O and network resources</li></ul><h3>When is it useful?</h3><p>A full backup is often used as the foundation of a backup strategy.</p><p>For example:</p><pre>Sunday<br>   │<br>   ▼<br>FULL BACKUP</pre><p>Then additional backups can capture changes after that.</p><h3>6. Incremental Backup</h3><p>An <strong>incremental backup stores changes made since the previous backup.</strong></p><p>For example:</p><pre>Sunday<br>Full Backup<br>     │<br>     ▼<br>Monday<br>Incremental<br>     │<br>     ▼<br>Tuesday<br>Incremental<br>     │<br>     ▼<br>Wednesday<br>Incremental</pre><p>Imagine:</p><pre>Sunday:<br>100 GB</pre><pre>Monday:<br>+ 5 GB changes</pre><pre>Tuesday:<br>+ 3 GB changes</pre><pre>Wednesday:<br>+ 4 GB changes</pre><p>The backups could conceptually contain:</p><pre>Sunday   → Full 100 GB<br>Monday   → 5 GB<br>Tuesday  → 3 GB<br>Wednesday→ 4 GB</pre><h3>Advantage</h3><p>Incremental backups are usually smaller and faster because they capture only changes since the previous backup.</p><h3>Disadvantage</h3><p>Restoration can be more complicated.</p><p>To restore Wednesday, you may need:</p><pre>Full Backup<br>     +<br>Monday Incremental<br>     +<br>Tuesday Incremental<br>     +<br>Wednesday Incremental</pre><p>If one required incremental backup is missing or unusable, the restore chain can be affected.</p><h3>7. Differential Backup</h3><p>A <strong>differential backup stores changes made since the last full backup.</strong></p><p>This is the important difference.</p><p>Suppose we take a full backup on Sunday:</p><pre>Sunday<br>FULL</pre><p>Then:</p><pre>Monday<br>Changes since Sunday → 5 GB</pre><pre>Tuesday<br>Changes since Sunday → 8 GB</pre><pre>Wednesday<br>Changes since Sunday → 12 GB</pre><p>So:</p><pre>Sunday<br>           FULL BACKUP<br>               │<br>       ┌───────┼────────┐<br>       │       │        │<br>     Monday  Tuesday  Wednesday<br>      5 GB     8 GB     12 GB</pre><p>Notice that Wednesday contains all changes since Sunday’s full backup, not just Wednesday’s changes.</p><h3>Advantage</h3><p>Restore can be simpler than a long incremental chain.</p><p>For example:</p><pre>Sunday Full<br>     +<br>Wednesday Differential<br>     =<br>Restore</pre><h3>Disadvantage</h3><p>Differential backups become larger as more data changes after the full backup.</p><h3>8. Incremental vs Differential</h3><p>This is a very common interview question.</p><pre>|              |      Incremental      |   Differential    |<br>| -----------  |     ---------------   | ----------------- |<br>| 🔄 Based on  |    Previous backup    | Last full backup  |<br>| 📦 Size      |    Usually smaller    | Usually grows     |<br>| ⚡ Speed     |    Usually faster     | Can become slower |<br>| 🔧 Restore   |    More complex       | Simpler           |<br>| 💾 Storage   |    Lower              | Higher            |</pre><h3>Simple example</h3><pre>Sunday → FULL</pre><pre>Monday<br>Incremental → changes since Sunday</pre><pre>Tuesday<br>Incremental → changes since Monday</pre><p>But:</p><pre>Sunday → FULL</pre><pre>Monday<br>Differential → changes since Sunday</pre><pre>Tuesday<br>Differential → changes since Sunday + Monday</pre><p>The easiest way to remember it:</p><blockquote><strong><em>Incremental = since the last backup</em></strong></blockquote><blockquote><strong><em>Differential = since the last full backup</em></strong></blockquote><h3>9. Logical Backup</h3><p>A <strong>logical backup</strong> represents database objects and their data in a logical format.</p><p>For MySQL, one common example is:</p><pre>mysqldump</pre><p>For example:</p><pre>mysqldump -u root -p company_db &gt; company_backup.sql</pre><p>The result may contain SQL statements such as:</p><pre>CREATE TABLE customers (...);</pre><pre>INSERT INTO customers VALUES (...);</pre><pre>INSERT INTO customers VALUES (...);</pre><p>You can later use the SQL file to recreate the database.</p><h3>Advantages</h3><ul><li>Easy to understand</li><li>Portable</li><li>Useful for migrations</li><li>Useful for smaller databases</li><li>Can selectively dump databases/tables</li></ul><h3>Disadvantages</h3><ul><li>Can be slower for very large databases</li><li>Restore can take significant time</li><li>Generates SQL rather than simply copying database files</li></ul><h3>10. Physical Backup</h3><p>A <strong>physical backup copies the actual database files</strong>.</p><p>Instead of saying:</p><blockquote><em>“Create SQL statements that represent the database.”</em></blockquote><p>we are effectively saying:</p><blockquote><em>“Copy the database’s underlying files.”</em></blockquote><p>Conceptually:</p><pre>MySQL Data Directory<br>        │<br>        │ Copy<br>        ▼<br>Backup Storage</pre><p>Physical backups are commonly important for large production databases because they can provide faster backup and restore operations compared with logical dumps, depending on the technology and environment.</p><p>Examples in the MySQL ecosystem include physical backup tools such as Percona XtraBackup.</p><h3>Advantages</h3><ul><li>Suitable for large databases</li><li>Can be faster for backup/restore</li><li>Preserves database files directly</li></ul><h3>Disadvantages</h3><ul><li>More complex</li><li>Usually less portable than logical SQL dumps</li><li>Requires understanding of the database storage format and backup tool</li></ul><h3>11. Logical vs Physical Backup</h3><p>Another useful comparison:</p><pre>| **Feature**    |      **Logical Backup**     |         **Physical Backup**            |<br>| -----------    | --------------------------- | ---------------------------------------|<br>| **Represents** |     Database logically      |          Physical database files       |<br>| **Example**    |        `mysqldump`          |              XtraBackup                |<br>| **Complexity** |    Easy to understand       |      More operationally complex        |<br>| **Best for**   |        Migrations           |       Large production systems         |<br>| **Selection**  | Can select specific objects | Typically works at file/database level |</pre><p>These aren’t necessarily competing choices.</p><p>A production organization may use different backup methods for different requirements.</p><h3>12. Binary Logs and Point-in-Time Recovery</h3><p>Now we reach one of the most important DBOPS concepts.</p><p>A backup taken at 12:00 AM does not automatically contain changes made at 10:00 AM.</p><p>So what happens if the database crashes at 3:00 PM?</p><p>This is where <strong>transaction logs</strong> become important.</p><p>In MySQL, these are commonly called <strong>binary logs (binlogs).</strong></p><p>They record database changes that can be used for replication and, when configured appropriately, recovery.</p><p>Think of it like this:</p><pre>12:00 AM<br>   │<br>   ▼<br>FULL BACKUP<br>   │<br>   │<br>   ├── 8 AM  → INSERT<br>   ├── 10 AM → UPDATE<br>   ├── 1 PM  → INSERT<br>   ├── 2 PM  → UPDATE<br>   └── 3 PM  → DELETE<br>                │<br>                ▼<br>             FAILURE</pre><p>If we have the full backup plus the required binary logs, we may be able to restore the database and replay changes up to a desired point.</p><p>This is called:</p><blockquote><strong><em>Point-in-Time Recovery (PITR)</em></strong></blockquote><h3>13. Real-World MySQL Recovery Example</h3><p>Imagine an online food delivery company.</p><p>The company has:</p><pre>Customers<br>Orders<br>Payments<br>Restaurants<br>Drivers</pre><p>At midnight:</p><pre>FULL BACKUP</pre><p>During the day:</p><pre>08:00 → New orders<br>10:00 → Payments<br>12:00 → New customers<br>14:00 → Driver updates<br>15:00 → Accidental DELETE</pre><p>Suppose someone accidentally deletes important order records at:</p><pre>15:05</pre><p>If the organization only has the midnight full backup, restoring it means losing legitimate changes from:</p><pre>00:00 → 15:05</pre><p>But if binary logs are available, the recovery process can potentially use:</p><pre>FULL BACKUP<br>      +<br>BINARY LOGS<br>      ↓<br>Replay changes<br>      ↓<br>Stop before accidental DELETE<br>      ↓<br>RECOVER</pre><p>This is why:</p><blockquote><strong><em>A backup strategy is not just about taking copies. It is about achieving the required recovery point.</em></strong></blockquote><h3>14. RPO and RTO</h3><p>When discussing backups professionally, two terms are extremely important.</p><h3>RPO — Recovery Point Objective</h3><p>RPO asks:</p><blockquote><strong><em>“How much data can the business afford to lose?”</em></strong></blockquote><p>Example:</p><pre>RPO = 15 minutes</pre><p>This means the organization wants the ability to recover to a point no more than approximately 15 minutes behind the failure, depending on the actual backup/logging design.</p><p>For a banking system, the acceptable data-loss window may be very small.</p><p>For a temporary internal reporting system, it might be larger.</p><h3>RTO — Recovery Time Objective</h3><p>RTO asks:</p><blockquote><strong><em>“How quickly must the system be restored?”</em></strong></blockquote><p>Example:</p><pre>RTO = 1 hour</pre><p>If the database goes down at:</p><pre>2:00 PM</pre><p>the organization wants the service restored by approximately:</p><pre>3:00 PM</pre><p>So:</p><pre>RPO → How much data can we lose?</pre><pre>RTO → How much time can we take to recover?</pre><p>These two requirements strongly influence the backup and recovery architecture.</p><h3>15. A Practical Backup Strategy</h3><p>A real organization might use something like:</p><pre>PRODUCTION DATABASE<br>                         │<br>             ┌───────────┼───────────┐<br>             │           │           │<br>             ▼           ▼           ▼<br>        Full Backup   Incremental   Binlogs<br>             │           │           │<br>             └───────────┼───────────┘<br>                         ▼<br>                  Backup Storage<br>                         │<br>                 ┌───────┴───────┐<br>                 │               │<br>              Local          Off-site<br>              Copy             Copy<br>                 │               │<br>                 └───────┬───────┘<br>                         ▼<br>                    Recovery</pre><p>The exact schedule depends on business requirements.</p><p>For example:</p><pre>Sunday<br>    Full Backup</pre><pre>Monday<br>    Incremental</pre><pre>Tuesday<br>    Incremental</pre><pre>Wednesday<br>    Incremental</pre><pre>Thursday<br>    Incremental</pre><pre>Friday<br>    Incremental</pre><pre>Saturday<br>    Incremental</pre><pre>Continuous<br>    Binary Logs</pre><p>This is only an example.</p><p>A real production strategy should be designed according to:</p><ul><li>RPO</li><li>RTO</li><li>database size</li><li>transaction volume</li><li>recovery requirements</li><li>storage cost</li><li>compliance requirements</li><li>infrastructure</li></ul><h3>16. Where Should Backups Be Stored?</h3><p>A common mistake is:</p><blockquote><em>“We have a backup on the same server, so we’re safe.”</em></blockquote><p>Not necessarily.</p><p>Imagine:</p><pre>Production Server<br>│<br>├── Database<br>│<br>└── Backup</pre><p>If the server’s disk fails:</p><pre>Database ❌<br>Backup   ❌</pre><p>Now the backup is useless.</p><p>A better approach is to maintain copies in separate locations.</p><p>For example:</p><pre>Production<br>                  │<br>                  ▼<br>             Backup System<br>                  │<br>          ┌───────┴────────┐<br>          ▼                ▼<br>     Local Storage     Cloud/Object<br>                         Storage</pre><p>For critical systems, organizations may also maintain an off-site or geographically separate copy.</p><h3>17. The 3–2–1 Backup Principle</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Z0NEyCoGJY7eOzAeZOhV3w.jpeg" /></figure><p>A widely used backup principle is:</p><h4>3 : Keep at least 3 copies of important data.</h4><h4>2 : Use at least 2 different types of storage/media.</h4><h4>1 : Keep at least 1 copy off-site.</h4><p>Conceptually:</p><pre>                 ORIGINAL<br>                    │<br>          ┌─────────┴─────────┐<br>          ▼                   ▼<br>      LOCAL COPY          SECOND COPY<br>                              │<br>                              ▼<br>                         OFF-SITE COPY</pre><p>The exact implementation depends on the organization’s infrastructure and requirements.</p><h3>18. Backup Is Not Complete Until Restore Is Tested</h3><p>This is one of the most important lessons for someone learning DBOPS.</p><p>Imagine a company says:</p><blockquote><em>“We have backups.”</em></blockquote><p>Then an incident happens.</p><p>The team tries to restore.</p><p>The backup is corrupted.</p><p>Or the password is missing.</p><p>Or the backup files are incomplete.</p><p>Or the restore takes 10 hours when the business expected 1 hour.</p><p>The organization technically had a backup.</p><p>But it did not have a <strong>reliable recovery capability</strong>.</p><p>Therefore:</p><blockquote><strong><em>A backup that has never been tested is only an assumption.</em></strong></blockquote><p>A good DBOPS process regularly performs restore tests.</p><p>For example:</p><pre>Backup<br>   │<br>   ▼<br>Restore to Test Environment<br>   │<br>   ▼<br>Verify Tables<br>   │<br>   ▼<br>Verify Row Counts<br>   │<br>   ▼<br>Verify Application<br>   │<br>   ▼<br>Record Recovery Time</pre><h3>19. Backup Security</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*QpvUuk0YvQEEKUz8RNf0lg.jpeg" /></figure><p>Backups contain valuable data.</p><p>Therefore, backups themselves must be protected.</p><p>Important practices include:</p><h3>Encryption</h3><p>Encrypt backups both during transfer and, where appropriate, at rest.</p><h3>Access control</h3><p>Only authorized people and systems should access backups.</p><h3>Separate credentials</h3><p>Do not expose backup credentials unnecessarily.</p><h3>Retention policies</h3><p>Do not keep every backup forever without a reason.</p><p>For example:</p><pre>Daily backups → 30 days<br>Weekly backups → 3 months<br>Monthly backups → 1 year</pre><p>The actual retention period should follow business and regulatory requirements.</p><h3>Immutable backups</h3><p>For important environments, immutable or otherwise protected backup copies can help prevent attackers from modifying or deleting backups.</p><h3>20. Common Backup Mistakes</h3><h3>Backup only on the same server</h3><p>Server failure can destroy both the database and backup.</p><h3>Never test restore</h3><p>A backup may exist but still be unusable.</p><h3>No backup monitoring</h3><p>A scheduled backup can fail silently.</p><h3>No retention policy</h3><p>Storage can eventually become full.</p><h3>No encryption</h3><p>Sensitive business data may be exposed.</p><h3>No recovery documentation</h3><p>During an incident, engineers may waste valuable time figuring out what to do.</p><h3>Confusing replication with backup</h3><p>Replication helps maintain copies and availability, but it does not replace backups.</p><p>For example:</p><pre>Primary DB<br>    │<br>    └── Replication ──&gt; Replica</pre><p>If someone accidentally deletes data on the primary and the change replicates immediately, the replica can also contain the deletion.</p><p>That is why:</p><blockquote><strong><em>Replication and backup solve different problems.\</em></strong></blockquote><h3>21. A Simple DBOPS Backup Architecture</h3><p>Putting everything together:</p><pre>USERS<br>                           │<br>                           ▼<br>                    APPLICATION<br>                           │<br>                           ▼<br>                 ┌─────────────────┐<br>                 │   MySQL Primary │<br>                 │   Production DB │<br>                 └────────┬────────┘<br>                          │<br>              ┌───────────┼───────────┐<br>              │           │           │<br>              ▼           ▼           ▼<br>          Full Backup  Incremental  Binary Logs<br>              │           │           │<br>              └───────────┼───────────┘<br>                          ▼<br>                 ┌──────────────────┐<br>                 │ Backup Repository│<br>                 └────────┬─────────┘<br>                          │<br>               ┌──────────┴──────────┐<br>               ▼                     ▼<br>         Local Backup          Off-site/Cloud<br>                                    Backup<br>                                      │<br>                                      ▼<br>                              Disaster Recovery</pre><p>This architecture helps answer an important question:</p><blockquote><strong><em>“What happens if production fails?”</em></strong></blockquote><p>Instead of simply hoping the database survives, we have a planned recovery path.</p><h3>Summary</h3><pre>DATABASE BACKUP<br>       │<br>       ├── Protect data<br>       │<br>       ├── Recover from failures<br>       │<br>       ├── Reduce downtime<br>       │<br>       ├── Support disaster recovery<br>       │<br>       ├── Meet business/compliance needs<br>       │<br>       └── Enable Point-in-Time Recovery<br>                    │<br>                    ▼<br>          FULL BACKUP + LOGS<br>                    │<br>                    ▼<br>              RESTORE<br>                    │<br>                    ▼<br>             VERIFY DATA<br>                    │<br>                    ▼<br>             RESUME SERVICE</pre><p>#DBOPS #DatabaseEngineering #Database #MySQL #DBA #DevOps #Backup #DisasterRecovery #PITR #Linux #CloudComputing #LearningInPublic</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d2617f4028b1" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[What is DBOPS?]]></title>
            <link>https://medium.com/@dilukshi1017/what-is-dbops-5489c5b03c27?source=rss-b92ccbe88d13------2</link>
            <guid isPermaLink="false">https://medium.com/p/5489c5b03c27</guid>
            <category><![CDATA[dbops]]></category>
            <category><![CDATA[database]]></category>
            <category><![CDATA[mysql]]></category>
            <category><![CDATA[database-administration]]></category>
            <category><![CDATA[technology]]></category>
            <dc:creator><![CDATA[Dilukshi Sangeetha Wijayawardhana]]></dc:creator>
            <pubDate>Mon, 21 Sep 2026 07:01:59 GMT</pubDate>
            <atom:updated>2026-09-21T07:01:59.130Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>From SQL Developer to Production Database Engineer</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qaqp6lcEEK3hlR_etU3msw.jpeg" /></figure><p>When developers start learning databases, education usually begins and ends with SQL syntax. We learn how to run SELECT * FROM users, insert dummy records, and connect a web app using an ORM.</p><p>In local development, queries return in single-digit milliseconds, disk space feels infinite, and data loss isn’t a catastrophe because a fresh sample script is only a terminal command away.</p><p>Then code reaches production.</p><p>Suddenly, a table holds 50 million rows instead of 50. Three slow queries spike CPU utilization to 100%, cascading into connection timeouts that bring down the entire API. At that scale, software engineering crosses paths with an essential, often overlooked discipline: <strong>Database Operations (DBOPS)</strong>.</p><h3>What Is DBOPS?</h3><p>In simple terms:</p><blockquote><strong><em>DBOPS is the practice of operating production databases safely, reliably, securely, and efficiently.</em></strong></blockquote><p>Writing SQL is only one part of database engineering.</p><p>A production database needs much more.</p><p>It needs:</p><ul><li>Monitoring</li><li>Backup</li><li>Recovery</li><li>Security</li><li>Performance optimization</li><li>Replication</li><li>Availability</li><li>Automation</li><li>Scaling</li><li>Incident response</li></ul><p>Think about it this way:</p><pre>                  PRODUCTION DATABASE<br>                           |<br>        -----------------------------------------<br>        |          |          |        |        |<br>     Monitor     Backup    Security  Performance<br>        |          |          |        |<br>      Alerts    Recovery   Access   Slow Queries<br>                           |<br>                       Availability<br>                           |<br>                         Scaling</pre><p>A developer may ask:</p><blockquote><em>“Does my query work?”</em></blockquote><p>A database engineer also needs to ask:</p><blockquote><em>“Is it fast enough?”</em></blockquote><blockquote><em>“What happens when the data grows?”</em></blockquote><blockquote><em>“What happens if the server fails?”</em></blockquote><blockquote><em>“Can we recover the data?”</em></blockquote><blockquote><em>“Who can access it?”</em></blockquote><blockquote><em>“How do we know when something goes wrong?”</em></blockquote><blockquote><em>“Can the system handle more traffic?”</em></blockquote><p>That change in thinking is what makes production database engineering different.</p><h3>SQL Is Only the Beginning</h3><p>Imagine we have a food delivery application.</p><p>A developer creates an orders table:</p><pre>CREATE TABLE orders (<br>    order_id INT PRIMARY KEY,<br>    customer_id INT,<br>    restaurant_id INT,<br>    amount DECIMAL(10,2),<br>    status VARCHAR(50)<br>);</pre><p>The table works perfectly.</p><p>During development, perhaps there are only 100 orders.</p><p>The query is fast.</p><p>Everything looks fine.</p><p>But six months later, the company has millions of orders.</p><p>Suddenly, a query that worked perfectly during development may become slow.</p><p>Then another problem appears.</p><p>CPU usage increases.</p><p>Database connections increase.</p><p>Users start experiencing timeouts.</p><p>The application becomes slower.</p><p>Now the question is no longer:</p><p><strong>“Does the SQL work?”</strong></p><p>The question becomes:</p><p><strong>“Can this database continue working reliably under real production traffic?”</strong></p><p>That is the world of DBOPS.</p><h3>DBA vs Database Engineer vs DBOPS</h3><p>These roles can overlap, especially in smaller companies.</p><p>But understanding their general focus helped me understand where DBOPS fits.</p><h3>1. Database Administrator — DBA</h3><p>A traditional DBA focuses heavily on operating and maintaining database systems.</p><p>Typical responsibilities include:</p><ul><li>Installing databases</li><li>Creating users</li><li>Managing permissions</li><li>Backup and restore</li><li>Database configuration</li><li>Monitoring</li><li>Performance tuning</li><li>Replication</li><li>Recovery</li></ul><p>A simplified view:</p><pre>                MySQL Server<br>                      |<br>        -------------------------------<br>        |       |        |       |     |<br>      Users  Access   Backup  Replica  Tuning</pre><p>The DBA makes sure the database environment is properly managed and maintained.</p><h3>2. Database Engineer</h3><p>A Database Engineer often works more closely with application development, database design, and data access patterns.</p><p>They may work on:</p><ul><li>Database schema design</li><li>Data modelling</li><li>SQL</li><li>Indexing</li><li>Query optimization</li><li>Data migration</li><li>Database architecture</li><li>Scaling</li><li>Automation</li></ul><p>For example, imagine we are building a food delivery platform.</p><p>A Database Engineer may think about relationships such as:</p><pre>Customer<br>   |<br>   | 1 : N<br>   ↓<br> Orders<br>   |<br>   | 1 : 1<br>   ↓<br>Payment</pre><p>They need to think about how the data should be structured and how applications will access it efficiently.</p><h3>3. DBOPS</h3><p>DBOPS brings database engineering together with <strong>operations and reliability</strong>.</p><p>A DBOPS engineer thinks about the database after it becomes a real production system.</p><p>A simplified picture looks like this:</p><pre>Developer<br>    |<br>    ↓<br>Application<br>    |<br>    ↓<br>Production Database<br>    |<br>    ├── Monitoring<br>    ├── Backup<br>    ├── Security<br>    ├── Replication<br>    ├── Recovery<br>    └── Performance</pre><p>And modern DBOPS can involve technologies such as:</p><pre>              DBOPS<br>                |<br>     ------------------------<br>     |       |       |      |<br>   MySQL   Linux   Cloud   Docker<br>     |       |       |      |<br>  Backup  Network   IAM   Kubernetes<br>     |       |       |<br>Replication Security Monitoring<br>                |<br>             CI/CD</pre><p>The exact responsibilities depend on the company.</p><p>But the mindset is similar:</p><blockquote><strong><em>Don’t just build the database. Operate it safely in the real world.</em></strong></blockquote><h3>Development Database vs Production Database</h3><p>This was one of the most important concepts for me to understand.</p><h3>Development</h3><p>A developer might have:</p><pre>Developer Laptop<br>       |<br>       ↓<br>Development MySQL<br>       |<br>       ↓<br>Test Data</pre><p>Maybe there are only:</p><pre>50 customers<br>100 orders<br>20 restaurants</pre><p>If something goes wrong, we can often recreate the database.</p><p>We can delete it.</p><p>Create it again.</p><p>Import test data.</p><p>Start over.</p><p>The impact is usually small.</p><h3>Production Is Different</h3><p>Now imagine a real food delivery company.</p><p>The production database may contain:</p><pre>1,000,000+ Customers<br>500,000+ Orders<br>Millions of Transactions<br>Payment Information<br>Restaurant Data<br>Driver Information</pre><p>Now imagine the database suddenly goes down.</p><pre>Customer<br>   ↓<br>Cannot place order</pre><pre>Driver<br>   ↓<br>Cannot receive jobs</pre><pre>Restaurant<br>   ↓<br>Cannot process orders</pre><pre>Company<br>   ↓<br>Business impact</pre><p>The database is no longer just a technical component.</p><p>It is part of the business.</p><p>That is why production database operations require much stronger planning around:</p><p><strong>Availability + Security + Backup + Recovery + Monitoring + Performance</strong></p><h3>Pillar 1: Monitoring</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/710/1*CJKC6gqx54-krkTjBZ4LzA.jpeg" /></figure><p>You cannot sit in front of a database and watch it all day.</p><p>So production databases need monitoring.</p><p>We may monitor:</p><pre>CPU<br>Memory<br>Disk Space<br>Connections<br>Query Latency<br>Slow Queries<br>Replication Lag<br>Errors<br>Locks<br>Availability</pre><p>For example:</p><pre>Disk Usage = 92%<br>       |<br>       ↓<br>Monitoring System<br>       |<br>       ↓<br>     ALERT<br>       |<br>       ↓<br>DBOPS Engineer<br>       |<br>       ↓<br>Investigate</pre><p>The important idea is that monitoring should help us discover problems <strong>before they become major incidents</strong>.</p><p>Imagine a server has 8% disk space remaining.</p><p>If nobody is watching it, the database could eventually run out of storage.</p><p>But with monitoring, an alert can be generated early.</p><p>The engineer can investigate and take action.</p><h3>Pillar 2: Backup</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*5CiwZgGtJU_UNZBrC0w1jA.jpeg" /></figure><p>A backup is a copy of data that can be used for recovery.</p><p>Now imagine someone accidentally runs:</p><pre>DELETE FROM orders;</pre><p>That single command could be extremely dangerous in production.</p><p>Without a suitable recovery strategy, getting the deleted data back may be difficult or impossible.</p><p>Production environments can use different backup strategies depending on their requirements.</p><p>A simplified example:</p><pre>              Database<br>                 |<br>        -------------------<br>        |                 |<br>   Full Backup        Binary Logs<br>        |                 |<br>        -----------+-------<br>                   |<br>                   ↓<br>              Recovery</pre><p>For example:</p><pre>Sunday<br>  ↓<br>Full Backup</pre><pre>Monday<br>  ↓<br>Database Changes</pre><pre>Tuesday<br>  ↓<br>Database Changes</pre><pre>Wednesday<br>  ↓<br>Accidental DELETE</pre><p>With an appropriate backup and binary-log strategy, it may be possible to recover the database to a point before the mistake.</p><p>This leads to another important concept.</p><h3>Backup Is Not Recovery</h3><p>This was one of the most important lessons for me.</p><p>Creating a backup does not automatically mean we can recover from it.</p><p>We need to test restoration.</p><p>A simple recovery process can look like:</p><pre>Production Database<br>        |<br>        ↓<br>      Backup<br>        |<br>        ↓<br>    Test Restore<br>        |<br>        ↓<br>   Verify Data<br>        |<br>        ↓<br>   Recovery Ready</pre><p>Imagine a company discovers that its backups are corrupted <strong>during a real disaster</strong>.</p><p>That is the worst time to discover the problem.</p><p>Therefore:</p><blockquote><strong><em>A backup is only useful if we know how to restore it successfully.</em></strong></blockquote><p>Or even more simply:</p><blockquote><strong><em>Backup creates the copy. Recovery proves that the copy works.</em></strong></blockquote><h3>Pillar 3: Performance Engineering</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3d8IdOlZQBk36B7tE9pn4A.jpeg" /></figure><p>A query can be correct and still be a bad production query.</p><p>Imagine a restaurant search:</p><pre>SELECT *<br>FROM restaurants<br>WHERE city = &#39;Colombo&#39;;</pre><p>Maybe the database contains only 100 restaurants.</p><p>No problem.</p><p>But what if the table contains millions of rows?</p><p>Now the database needs to determine how efficiently it can find the required records.</p><p>This is where tools such as EXPLAIN become important.</p><p>For example:</p><pre>EXPLAIN<br>SELECT *<br>FROM restaurants<br>WHERE city = &#39;Colombo&#39;;</pre><p>We can investigate things such as:</p><ul><li>Index usage</li><li>Execution plan</li><li>Rows scanned</li><li>Join behavior</li><li>Locks</li><li>CPU</li><li>Memory</li><li>Disk I/O</li></ul><p>A simplified troubleshooting flow:</p><pre>               Slow Query<br>                   |<br>                   ↓<br>                EXPLAIN<br>                   |<br>                   ↓<br>          Check Execution Plan<br>                   |<br>                   ↓<br>             Check Indexes<br>                   |<br>                   ↓<br>           Optimize Query<br>                   |<br>                   ↓<br>              Test Again<br>                   |<br>                   ↓<br>               Monitor</pre><p>The goal is not simply:</p><p><strong>“Make the query work.”</strong></p><p>The goal is:</p><p><strong>“Make the query work efficiently at production scale.”</strong></p><h3>Pillar 4: Availability</h3><p>Availability means the database should be available when the application needs it.</p><p>Imagine a customer placing an order:</p><pre>Customer<br>    |<br>    ↓<br>Mobile App<br>    |<br>    ↓<br>Backend<br>    |<br>    ↓<br>Database ❌</pre><p>The application may not be able to complete the order.</p><p>This is why production systems may use techniques such as:</p><ul><li>Replication</li><li>Failover</li><li>High-availability architectures</li><li>Backups</li><li>Disaster recovery</li><li>Monitoring</li></ul><p>A simplified architecture could look like:</p><pre>                 Application<br>                      |<br>                      ↓<br>                 Primary DB<br>                      |<br>                 Replication<br>                 /          \<br>                ↓            ↓<br>           Replica 1     Replica 2</pre><p>The exact architecture depends on the database technology and business requirements.</p><p>The important idea is:</p><blockquote><strong><em>A production database should be designed with failure in mind.</em></strong></blockquote><h3>Pillar 5: Security</h3><p>Database security is another major part of DBOPS.</p><p>One common principle is:</p><h3>Least Privilege</h3><p>Simply put:</p><blockquote><strong><em>Give an application or user only the permissions it actually needs.</em></strong></blockquote><p>For example:</p><pre>Application<br>     |<br>     ↓<br>Database User<br>     |<br>     ↓<br>Limited Permissions<br>     |<br>     ↓<br>Only Required Operations</pre><p>Instead of giving an application unlimited database access, we should restrict what it can do.</p><p>Other important security areas include:</p><ul><li>Authentication</li><li>Access control</li><li>Encryption</li><li>Private networking</li><li>Secure credentials</li><li>Auditing</li></ul><p>Security is not something to think about only after an incident.</p><p>It should be part of the database design from the beginning.</p><h3>A Real-World DBOPS Scenario</h3><p>Let’s bring everything together.</p><p>Imagine we are operating the database for a food delivery application.</p><p>The application architecture may look like:</p><pre>Customer<br>   |<br>   ↓<br>Mobile Application<br>   |<br>   ↓<br>Backend API<br>   |<br>   ↓<br>MySQL Production</pre><p>But behind the database, there is an entire operational layer:</p><pre>                  MySQL Production<br>                                |<br>        ------------------------------------------------<br>        |             |             |          |       |<br>    Monitoring     Backup       Security   Performance<br>        |             |             |          |<br>      Alerts       Recovery      Access     Indexes<br>                                |<br>                           Replication</pre><p>Now imagine it is 3:00 PM.</p><p>The application suddenly becomes slow.</p><p>What does the database engineer do?</p><p>Not randomly restart the server.</p><p>They investigate.</p><p>A simplified troubleshooting process could be:</p><pre>1. Check database availability<br>             ↓<br>2. Check CPU / Memory<br>             ↓<br>3. Check disk usage<br>             ↓<br>4. Check active connections<br>             ↓<br>5. Check running queries<br>             ↓<br>6. Check slow query logs<br>             ↓<br>7. Run EXPLAIN<br>             ↓<br>8. Check indexes<br>             ↓<br>9. Identify the root cause<br>             ↓<br>10. Fix and verify<br>             ↓<br>11. Continue monitoring</pre><p>This is where database engineering becomes much more interesting than simply writing SQL.</p><h3>The DBOPS Mindset</h3><p>I started to notice a difference between the questions I asked as a developer and the questions I need to ask when thinking about production databases.</p><p>A developer might ask:</p><blockquote><strong><em>Does my query work?</em></strong></blockquote><p>A production database engineer also asks:</p><blockquote><em>Is it safe?</em></blockquote><blockquote><em>Is it fast?</em></blockquote><blockquote><em>What happens when the data grows?</em></blockquote><blockquote><em>What happens if the server fails?</em></blockquote><blockquote><em>Can we recover the data?</em></blockquote><blockquote><em>Who has access?</em></blockquote><blockquote><em>How will we detect a problem?</em></blockquote><blockquote><em>How will we scale the system?</em></blockquote><blockquote><em>Can we automate this process?</em></blockquote><p>These questions represent the <strong>DBOPS mindset</strong>.</p><h3>What Technologies Should a DBOPS Engineer Learn?</h3><p>DBOPS is not a single technology.</p><p>It is a combination of database, operating system, cloud, automation, and monitoring knowledge.</p><p>A possible learning path looks like:</p><pre>                    DBOPS<br>                      |<br>       --------------------------------<br>       |              |               |<br>    Database        Linux           Cloud<br>       |              |               |<br>    MySQL          Commands          AWS/GCP<br>    SQL            Processes          IAM<br>    Indexes        Networking         Storage<br>    Backup         Security           VM<br>    Recovery<br>       |<br>   Replication<br>       |<br>  High Availability<br>       |<br>   Performance<br>       |<br>   Monitoring<br>       |<br>  Docker / Kubernetes<br>       |<br>      CI/CD</pre><blockquote><strong><em>How databases are designed, deployed, monitored, protected, recovered, and scaled in real production environments.</em></strong></blockquote><p>A production database engineer needs to think about:</p><pre>              DATABASE ENGINEERING<br>                         |<br>       -----------------------------------------<br>       |        |        |        |      |     |<br>      SQL   Security  Backup  Performance  HA  Scaling<br>                 |        |        |<br>              Recovery  Monitoring</pre><p>#DBOPS #DatabaseEngineering #MySQL #DatabaseAdministrator #DevOps #Linux #CloudComputing #SQL #DatabasePerformance #TechLearning</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5489c5b03c27" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Managing the Data: The Brain of an RPA Robot]]></title>
            <link>https://medium.com/@dilukshi1017/managing-the-data-the-brain-of-an-rpa-robot-d0b0f5a4b9ae?source=rss-b92ccbe88d13------2</link>
            <guid isPermaLink="false">https://medium.com/p/d0b0f5a4b9ae</guid>
            <category><![CDATA[robotics]]></category>
            <category><![CDATA[rpa-solutions]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[automation]]></category>
            <category><![CDATA[beginners-guide]]></category>
            <dc:creator><![CDATA[Dilukshi Sangeetha Wijayawardhana]]></dc:creator>
            <pubDate>Fri, 21 Aug 2026 08:44:33 GMT</pubDate>
            <atom:updated>2026-08-21T08:44:33.699Z</atom:updated>
            <content:encoded><![CDATA[<p>Understanding variables, static data, and dynamic data — the simple way a robot remembers what it is working with.</p><p>In my previous article, I explained the basic building blocks of an RPA robot: <strong>workflows, triggers, activities, elements, and attributes</strong>.</p><p>Once we understand those building blocks, another question comes up:</p><blockquote><strong><em>How does a robot remember information while it is working?</em></strong></blockquote><p>Imagine telling a robot:</p><blockquote><em>“Read the student’s name, remember the marks, open the website, enter the details, and then move to the next student.”</em></blockquote><p>The robot needs somewhere to temporarily keep that information.</p><p>This is where <strong>variables</strong> come in.</p><p>You can think of variables as the robot’s <strong>small storage boxes</strong>.</p><pre>RPA BOT<br>                 │<br>        ┌────────┼────────┐<br>        ↓        ↓        ↓<br>     Student   Marks    Date<br>      Name<br>        │        │        │<br>        ↓        ↓        ↓<br>    &quot;Kamal&quot;     85   &quot;2026-08-21&quot;</pre><p>The values inside these boxes can change while the workflow is running. Let’s understand this with a simple, real-world example.</p><h3>1. What Is a Variable?</h3><p>A <strong>variable</strong> is simply a place where a program or automation stores information so it can use that information later.</p><p>If this sounds technical, think about a cardboard box on your desk. You can put a piece of paper inside the box:</p><pre>┌──────────────────┐<br>│  student_name    │<br>│                  │<br>│     &quot;Kamal&quot;      │<br>└──────────────────┘</pre><p>Here:</p><p><strong>student_name</strong> is the variable.</p><p><strong>&quot;Kamal&quot;</strong> is the value stored inside it.</p><p>Later, we can change the value:</p><pre>student_name = &quot;Nimal&quot;</pre><p>Now the same box contains:</p><pre>┌──────────────────┐<br>│  student_name    │<br>│                  │<br>│     &quot;Nimal&quot;      │<br>└──────────────────┘</pre><p>The box stayed the same.</p><p>The <strong>value inside the box changed</strong>.</p><p>That is the basic idea of a variable.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*T-61urPRARsZV2BsP-iACA.jpeg" /></figure><h3>2. Why Does an RPA Bot Need Variables?</h3><p>Without variables, it would be difficult for a bot to work with changing information.</p><p>Let’s take a simple example.</p><p>Imagine an Excel file contains:</p><pre>| Student | Marks |<br>| ------- | ----: |<br>| Kamal   |    85 |<br>| Nimal   |    92 |<br>| Amal    |    78 |</pre><p>Our bot needs to process each student.</p><p>The workflow might look like this:</p><pre>                            Excel<br>                              ↓<br>                      Read Student Name<br>                              ↓<br>                     Store in student_name<br>                              ↓<br>                          Read Marks<br>                              ↓<br>                        Store in marks<br>                              ↓<br>                         Open Website<br>                              ↓<br>                      Enter student_name<br>                              ↓<br>                         Enter marks<br>                              ↓<br>                            Submit</pre><p>When the bot processes Kamal:</p><pre>student_name = &quot;Kamal&quot;<br>marks = 85</pre><p>Then it moves to Nimal:</p><pre>student_name = &quot;Nimal&quot;<br>marks = 92</pre><p>Then Amal:</p><pre>student_name = &quot;Amal&quot;<br>marks = 78</pre><p>The same variables are being reused.</p><p>This is one of the reasons variables are so important in automation.</p><h3>3. Variables Are Like the Robot’s Memory</h3><p>I like to think of variables as <strong>temporary memory for the robot</strong>.</p><p>For example, a bot processing an invoice might have:</p><pre>invoice_number = &quot;INV-10025&quot;<br>customer_name = &quot;ABC Company&quot;<br>amount = 25000<br>invoice_date = &quot;2026-08-21&quot;</pre><p>The robot can then use those values in different activities.</p><p>For example:</p><pre>                              Read Invoice<br>                                    ↓<br>                           Store invoice_number<br>                                    ↓<br>                           Store customer_name<br>                                    ↓<br>                              Store amount<br>                                    ↓<br>                              Open Database<br>                                    ↓<br>                             Save information</pre><p>The variable allows information to travel from one activity to another.</p><pre>                                  Activity 1<br>                                     ↓<br>                                  Read data<br>                                     ↓<br>                                  Variable<br>                                     ↓<br>                                  Activity 2<br>                                     ↓<br>                                  Use data</pre><p>So a useful way to remember it is:</p><blockquote><strong><em>A variable stores information that the robot needs to use later.</em></strong></blockquote><h3>4. Variables in a Real RPA Workflow</h3><p>Let’s imagine we are building a simple employee onboarding automation.</p><p>The company has an Excel file:</p><pre>| Employee | Email                                     | Department |<br>| -------- | ----------------------------------------- | ---------- |<br>| Kamal    | [kamal@email.com](kamal@email.com)        | IT         |<br>| Nimal    | [nimal@email.com](nimal@email.com)        | Finance    |<br>| Amal     | [amal@email.com](amal@email.com)          | HR         |</pre><p>The bot needs to create an account for each employee.</p><p>The workflow could be:</p><pre>                                Start<br>                                 ↓<br>                             Read Excel<br>                                 ↓<br>                         Get employee name<br>                                 ↓<br>                             Get email<br>                                 ↓<br>                          Get department<br>                                 ↓<br>                     Store values in variables<br>                                 ↓<br>                           Open HR system<br>                                 ↓<br>                           Create account<br>                                 ↓<br>                        Enter employee details<br>                                 ↓<br>                               Submit<br>                                 ↓<br>                           Next employee</pre><p>During one loop:</p><pre>employee_name = &quot;Kamal&quot;<br>email = &quot;kamal@email.com&quot;<br>department = &quot;IT&quot;</pre><p>During the next loop:</p><pre>employee_name = &quot;Nimal&quot;<br>email = &quot;nimal@email.com&quot;<br>department = &quot;Finance&quot;</pre><p>The workflow remains the same.</p><p>Only the <strong>data inside the variables changes</strong>.</p><p>This is where variables become especially powerful.</p><h3>5. Static vs Dynamic Variables</h3><p>Not all information changes during automation.</p><p>Some information stays the same.</p><p>Other information changes while the bot is running.</p><p>This gives us a useful distinction:</p><pre>Variables<br>    │<br>    ├── Static Data<br>    │<br>    └── Dynamic Data</pre><h3>6. Static Variables</h3><p>A <strong>static variable</strong> contains information that normally stays the same during a particular workflow.</p><p>For example, imagine a company has one main website:</p><pre>https://company-example.com</pre><p>The bot always needs to open this website.</p><p>We could store it as:</p><pre>website_url = &quot;https://company-example.com&quot;</pre><p>During the workflow, the value doesn’t change.</p><pre>website_url<br>     ↓<br>https://company-example.com</pre><p>Other examples could include:</p><pre>company_name = &quot;ABC Company&quot;<br>department = &quot;Finance&quot;<br>file_path = &quot;C:/Reports/&quot;</pre><p>The exact meaning of “static” can depend on the workflow. Something that is static during one run could be changed later when the automation is configured differently.</p><p>So I think of static data simply as:</p><blockquote><strong><em>Information that remains the same while the workflow is running.</em></strong></blockquote><h3>7. Dynamic Variables</h3><p>Now let’s look at the opposite.</p><p>A <strong>dynamic variable</strong> changes while the workflow is running.</p><p>The best example is a loop.</p><p>Imagine our Excel file contains:</p><pre>| Student | Marks |<br>| ------- | ----: |<br>| Kamal   |    85 |<br>| Nimal   |    92 |<br>| Amal    |    78 |</pre><p>The bot processes one student at a time.</p><p>At the beginning:</p><pre>current_student = &quot;Kamal&quot;<br>current_marks = 85</pre><p>After the first loop:</p><pre>current_student = &quot;Nimal&quot;<br>current_marks = 92</pre><p>After the second loop:</p><pre>current_student = &quot;Amal&quot;<br>current_marks = 78</pre><p>The variable names stay the same.</p><p>But their <strong>values change</strong>.</p><pre>           LOOP 1<br>current_marks = 85<br>     ↓<br>           LOOP 2<br>current_marks = 92<br>     ↓<br>           LOOP 3<br>current_marks = 78</pre><p>That is dynamic data.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*KR97pyeiDlNIAwXm1URW1A.jpeg" /></figure><h3>8. Static vs Dynamic: A Simple Comparison</h3><p>Here is the easiest way I remember the difference.</p><pre>| Static                                     | Dynamic                     |<br>| ------------------------------------------ | --------------------------- |<br>| Usually stays the same during the workflow | Changes during the workflow |<br>| Company website                            | Current student             |<br>| Company name                               | Current marks               |<br>| Fixed folder path                          | Current invoice             |<br>| Fixed email address                        | Current customer            |<br>| Configuration value                        | Data from the current loop  |</pre><p>For example:</p><pre>STATIC<br>                                website_url<br>                                     ↓<br>                                Same website<br>                                     ↓<br>                                Same website<br>                                     ↓<br>                                Same website</pre><p>While:</p><pre>DYNAMIC<br>                                current_student<br>                                      ↓<br>                                    Kamal<br>                                      ↓<br>                                    Nimal<br>                                      ↓<br>                                     Amal</pre><h3>9. Variables + Loops = Powerful Automation</h3><p>This is where things become really interesting.</p><p>A variable becomes extremely useful when we combine it with a <strong>loop</strong>.</p><p>Suppose we have 1,000 customers.</p><p>We don’t want to create 1,000 separate workflows.</p><p>Instead, we create one workflow:</p><pre>                              Read Customer<br>                                    ↓<br>                        Store customer in variable<br>                                    ↓<br>                             Process customer<br>                                    ↓<br>                           Move to next customer<br>                                    ↓<br>                              Update variable<br>                                    ↓<br>                           Process next customer</pre><p>For example:</p><pre>Customer 1 → Kamal<br>Customer 2 → Nimal<br>Customer 3 → Amal<br>Customer 4 → Saman<br>...<br>Customer 1000</pre><p>The same workflow handles all of them.</p><p>This is one of the ideas that makes automation scalable.</p><h3>10. Variables Can Store Different Types of Data</h3><p>Variables don’t only store names.</p><p>Depending on the automation tool and workflow, we can work with different types of information.</p><p>For example:</p><h4>Text</h4><pre>customer_name = &quot;Kamal&quot;</pre><h4>Number</h4><pre>marks = 85</pre><h4>True / False</h4><pre>payment_completed = True</pre><h4>Date</h4><pre>invoice_date = &quot;2026-08-21&quot;</pre><h4>List</h4><pre>students = [&quot;Kamal&quot;, &quot;Nimal&quot;, &quot;Amal&quot;]</pre><h4>Dictionary / structured data</h4><pre>customer = {<br>    &quot;name&quot;: &quot;Kamal&quot;,<br>    &quot;city&quot;: &quot;Colombo&quot;,<br>    &quot;age&quot;: 25<br>}</pre><p>The exact variable types available depend on the RPA platform, but the main idea is the same:</p><blockquote><strong><em>Different kinds of information need to be stored and processed in different ways.</em></strong></blockquote><h3>11. Variables in ElectroNeek</h3><p>When working with RPA platforms such as <strong>ElectroNeek</strong>, variables become part of the workflow.</p><p>You might have variables such as:</p><pre>customerName<br>invoiceNumber<br>invoiceAmount<br>filePath<br>websiteURL<br>currentRow</pre><p>Activities can use those variables as inputs.</p><p>For example:</p><pre>                              Read Excel<br>                                   ↓<br>                            Get Customer Name<br>                                   ↓<br>                          Store in customerName<br>                                   ↓<br>                              Open Website<br>                                   ↓<br>                             Use customerName<br>                                   ↓<br>                           Enter Customer Name</pre><p>This makes the workflow reusable.</p><p>Instead of hard-coding:</p><pre>Kamal</pre><p>inside the workflow, we can use:</p><pre>customerName</pre><p>Then the bot can process Kamal, Nimal, Amal, or thousands of other customers.</p><p>when variables are combined with conditions and loops, our robot can handle much more realistic business processes.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d0b0f5a4b9ae" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Anatomy of a Robot: Understanding the Building Blocks of RPA]]></title>
            <link>https://medium.com/@dilukshi1017/the-anatomy-of-a-robot-understanding-the-building-blocks-of-rpa-d0e11c12e7a1?source=rss-b92ccbe88d13------2</link>
            <guid isPermaLink="false">https://medium.com/p/d0e11c12e7a1</guid>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[automation]]></category>
            <category><![CDATA[tech]]></category>
            <category><![CDATA[rpa-solutions]]></category>
            <category><![CDATA[roboticprocessautomation]]></category>
            <dc:creator><![CDATA[Dilukshi Sangeetha Wijayawardhana]]></dc:creator>
            <pubDate>Thu, 20 Aug 2026 05:15:04 GMT</pubDate>
            <atom:updated>2026-08-20T05:15:04.821Z</atom:updated>
            <content:encoded><![CDATA[<p>A robot may look simple from the outside, but behind every successful automation is a carefully designed workflow.</p><p>In my previous article, I introduced RPA (Robotic Process Automation) and explained how software robots can perform repetitive computer tasks. But after understanding what RPA is, another question comes to mind:</p><h3>How do we actually build a robot?</h3><p>An RPA bot doesn’t magically know what to do. We have to give it clear instructions. Just like a teacher prepares a lesson plan before teaching a class, an automation developer creates a workflow that tells the robot what to do, when to do it, and where to do it.</p><p>To understand how this works, we need to learn some basic RPA vocabulary. The five important building blocks are:</p><pre>                      Workflow<br>                         ↓<br>                      Trigger<br>                         ↓<br>                     Activity<br>                         ↓<br>                      Element<br>                         ↓<br>               Attributes / Properties</pre><p>Let’s understand each one with a simple example.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*t0cs_BlJ9fe1mUA-LNQZVw.jpeg" /></figure><h3>1. What is a Workflow?</h3><p>A workflow is the complete sequence of steps that an RPA bot follows to finish a task.</p><p>Think about a cooking recipe. A recipe might say:</p><pre>                    Take ingredients<br>                            ↓<br>                     Cut vegetables<br>                            ↓<br>                      Heat the pan<br>                            ↓<br>                     Cook vegetables<br>                            ↓<br>                          Serve</pre><p>An RPA workflow works in a similar way. For example, imagine a company wants to process customer information from an Excel file. The workflow could be:</p><pre>                      Start<br>                        ↓<br>                      Open Excel<br>                        ↓<br>             Read customer information<br>                        ↓<br>                   Open website<br>                        ↓<br>             Enter customer information<br>                        ↓<br>                    Click Submit<br>                        ↓<br>                    Check result<br>                        ↓<br>                    Save result<br>                        ↓<br>                       End</pre><p>This entire sequence is the workflow.</p><p>A simple way to remember it:</p><p><strong>Workflow = the complete plan of what the robot needs to do.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*r9z_GoHDxKSYX1JWRdI-eQ.jpeg" /></figure><h3>2. What is a Trigger Point?</h3><p>A workflow needs something to tell the robot: <em>“Start working now.”</em> This is called a trigger.</p><p>I like to think of a trigger as an alarm clock for a robot. Your alarm rings at 6:00 AM and tells you: <em>“Wake up!”</em> A trigger tells the bot: <em>“Start the workflow.”</em></p><p>For example, a company might want a bot to start when a new email arrives.</p><pre>                  New email arrives<br>                        ↓<br>                      Trigger<br>                        ↓<br>                   Start the bot<br>                        ↓<br>                    Read email</pre><p>Another example is a scheduled trigger.</p><pre>              Every day at 8:00 AM<br>                        ↓<br>                      Trigger<br>                        ↓<br>                    Start bot</pre><p>Other possible triggers include a new file appearing in a folder, or a user manually clicking a button to start the bot.</p><p>A simple way to remember it:</p><p><strong>Trigger = What tells the robot to start?</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*rk_XPbpO-ZasC96QUVXpjQ.jpeg" /></figure><h3>3. What is an Activity?</h3><p>Once the robot starts, it needs to perform individual tasks. Each individual task is called an activity.</p><p>If the workflow is the complete cooking recipe, the activities are the individual steps like “chop the vegetables” or “stir the soup.”</p><p>For example: “Open browser” is one activity. “Enter username” is another activity. “Click Login” is another activity.</p><p>So a workflow can look like:</p><pre>Workflow<br>   │<br>   ├── Open Browser       ← Activity<br>   ├── Open Website       ← Activity<br>   ├── Enter Username     ← Activity<br>   ├── Enter Password     ← Activity<br>   └── Click Login        ← Activity</pre><p>An activity normally represents one specific operation the bot needs to perform. An RPA tool provides activities for opening applications, clicking, typing, reading Excel, moving files, sending emails, and repeating actions. So when building an RPA workflow, we are simply connecting many small activities together.</p><h3>4. What is an Element?</h3><p>Now we have a workflow and activities. But there is another important question. Suppose we tell the bot: <em>“Click.”</em></p><p>The bot might ask: <em>“Click what?”</em></p><p>That’s where an element comes in. An element is the specific object or target that the bot interacts with on the screen.</p><p>For example, imagine a login page:</p><pre>                  ┌──────────────────────────┐<br>                  │       Login              │<br>                  │                          │<br>                  │ Username: [___________]  │ ← Element<br>                  │                          │<br>                  │ Password: [___________]  │ ← Element<br>                  │                          │<br>                  │        [ LOGIN ]         │ ← Element<br>                  │                          │<br>                  └──────────────────────────┘</pre><p>If the activity is “Click”, the element could be the “Login button”.</p><p>If the activity is “Type”, the element could be the “Username textbox”.</p><p>So we can think about it like this:</p><p><strong>Activity + Element = Actual interaction</strong></p><h4>(Click) + (Login Button) = Bot clicks Login</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*D6ZMfaen8qSkFVehiHXy3Q.jpeg" /></figure><h3>5. What are Attributes or Properties?</h3><p>Now comes a slightly more technical concept. Imagine there are two buttons on a website that both say “Submit”. The robot needs to know exactly <em>which</em> Submit button it should click.</p><p>This is where the element’s attributes or properties help identify and control it. For example, a web element might have hidden code information such as:</p><p>id = &quot;loginButton&quot;</p><p>class = &quot;btn-primary&quot;</p><p>text = &quot;Login&quot;</p><p>These properties help the automation tool uniquely identify the correct element. You don’t need to memorize every possible property at the beginning.</p><p><strong>The important idea is:</strong> Attributes/properties describe an element and help the automation tool identify it perfectly.</p><h3>Putting Everything Together</h3><p>Now let’s connect the concepts. Imagine we want to automate a simple login.</p><ul><li><strong>TRIGGER:</strong> The bot receives a command to start.</li><li><strong>WORKFLOW:</strong> The complete process begins.</li><li><strong>ACTIVITY:</strong> Open Website</li><li><strong>ACTIVITY:</strong> Enter Username ➔ <strong>ELEMENT:</strong> Username textbox</li><li><strong>ACTIVITY:</strong> Enter Password ➔ <strong>ELEMENT:</strong> Password textbox</li><li><strong>ACTIVITY:</strong> Click ➔ <strong>ELEMENT:</strong> Login Button (identified by its <strong>PROPERTIES</strong>)</li><li><strong>SUCCESS!</strong></li></ul><p>This is the basic anatomy of an RPA automation.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*vhj6Erw7r_oxoV2mF38V5g.jpeg" /></figure><h3>Why Do We Need These Concepts?</h3><p>At first, these words might seem unnecessary. But when an automation fails, understanding them becomes very important.</p><p>Imagine your bot stops and says: <em>“Element not found.”</em></p><p>What does that mean? It could mean the website changed, the element’s properties changed, the page has not finished loading, or the bot is looking in the wrong place.</p><p>Now you can start looking for the real issue instead of simply saying: <em>“The bot doesn’t work.”</em></p><p>This is the exact difference between simply testing a bot to see if it works, and actually debugging it to fix the problem! Don’t just build the bot. Learn how to understand why it fails.</p><p>Learning RPA is not about memorizing hundreds of activities in an automation tool. The most important thing is learning how to think in workflows. We take a large human task and break it into small, clear, machine-executable steps.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d0e11c12e7a1" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Meet Your New Digital Assistant: A Beginner’s Guide to RPA]]></title>
            <link>https://medium.com/@dilukshi1017/meet-your-new-digital-assistant-a-beginners-guide-to-rpa-e14d184ec63c?source=rss-b92ccbe88d13------2</link>
            <guid isPermaLink="false">https://medium.com/p/e14d184ec63c</guid>
            <category><![CDATA[roboticprocessautomation]]></category>
            <category><![CDATA[productivity]]></category>
            <category><![CDATA[automation]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[software-development]]></category>
            <dc:creator><![CDATA[Dilukshi Sangeetha Wijayawardhana]]></dc:creator>
            <pubDate>Wed, 19 Aug 2026 07:39:35 GMT</pubDate>
            <atom:updated>2026-08-19T07:39:35.647Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Fx5Vam513j9AQWm_uwNiNg.jpeg" /></figure><h3>How to let a software robot handle repetitive computer work so you can focus on more meaningful tasks.</h3><p>Have you ever spent hours copying data from an Excel sheet into a website?</p><p>You read a value from Excel, switch to another application, type it, click a button, go back to Excel, and repeat the same process again and again.</p><p>It is boring. It is repetitive. And honestly, it is the kind of work where your brain starts asking, <strong>“Can’t someone — or something — just do this for me?”</strong></p><p>The good news is: <strong>yes.</strong></p><p>That is where <strong>RPA (Robotic Process Automation)</strong> comes in.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Df41Yjmx85fYEGcbD7pAIQ.jpeg" /></figure><h3>01. What is RPA?</h3><p>RPA stands for <strong>Robotic Process Automation</strong>.</p><p>Before hearing the word “robot”, you might imagine a physical robot walking around an office. But RPA is not that kind of robot.</p><p>An RPA bot is a <strong>software robot</strong>.</p><p>It runs on a computer and performs tasks that a human normally does using applications.</p><p>For example, imagine a person doing this:</p><pre>Open Excel<br>     ↓<br>Read customer name<br>     ↓<br>Open website<br>     ↓<br>Enter customer name<br>     ↓<br>Enter email<br>     ↓<br>Click Submit<br>     ↓<br>Go back to Excel<br>     ↓<br>Repeat 1,000 times</pre><p>An RPA bot can follow the same workflow automatically.</p><p>So, in simple words:</p><p><strong>RPA is a technology that allows software robots to perform repetitive, rule-based computer tasks automatically.</strong></p><h3>02. How Does RPA Work?</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*pZWoHcOWOc6YJOhB1gMqfQ.jpeg" /></figure><p>One reason RPA is interesting is that a bot can interact with computer applications in a way that is similar to how humans interact with them.</p><p>A bot can:</p><ul><li>Open applications</li><li>Click buttons</li><li>Type text</li><li>Read information</li><li>Copy and paste data</li><li>Work with Excel files</li><li>Move files</li><li>Send emails</li><li>Interact with websites</li><li>Call APIs</li><li>Process information</li><li>Make decisions based on predefined rules</li></ul><p>For example:</p><pre>Excel<br>  ↓<br>Read customer<br>  ↓<br>Open website<br>  ↓<br>Enter customer details<br>  ↓<br>Click Submit<br>  ↓<br>Check result<br>  ↓<br>Save result</pre><p>The important thing is that the bot is not simply “clicking randomly.”</p><p>We create a <strong>workflow</strong> that tells the bot what to do.</p><p>A workflow can contain:</p><p><strong>Trigger → Actions → Conditions → Loops → Result</strong></p><p>For example:</p><pre>Start<br>  ↓<br>Read Excel<br>  ↓<br>For each customer<br>  ↓<br>Open website<br>  ↓<br>Enter details<br>  ↓<br>Submit<br>  ↓<br>Was it successful?<br>    ↙        ↘<br>  Yes        No<br>   ↓          ↓<br>Continue    Log error</pre><p>This is very similar to giving instructions to a new employee, but instead of a human employee, we are giving the instructions to a software bot.</p><h3><strong>03. Human vs. RPA Bot</strong></h3><p>Let’s imagine a simple data-entry task.</p><p>A human might need to process hundreds of records manually.</p><pre>👩 Human<br><br>Read Excel<br>   ↓<br>Type data<br>   ↓<br>Click<br>   ↓<br>Check<br>   ↓<br>Repeat<br>   ↓<br>Repeat<br>   ↓<br>Repeat</pre><p>After several hours, the person can become tired or lose concentration.</p><p>An RPA bot can follow the same process continuously:</p><pre>🤖 RPA Bot<br><br>Read Excel<br>   ↓<br>Type data<br>   ↓<br>Click<br>   ↓<br>Check<br>   ↓<br>Repeat<br>   ↓<br>Repeat<br>   ↓<br>Repeat</pre><p>The bot does not get bored by repetitive work.</p><p>However, this does <strong>not</strong> mean RPA is perfect. A poorly designed automation can still make mistakes, especially when the input data is wrong or the application changes.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qsoTX2W6Dcdc4pWu1nhwag.jpeg" /></figure><h3>04. Real-World Magic: Where Can We Use RPA?</h3><p>RPA is especially useful when a process is <strong>repetitive, rule-based, and happens frequently</strong>.</p><p>Let’s look at a few simple examples.</p><p><strong>🧑‍🏫 Use Case 1: A School or Education System</strong></p><p>Imagine a teacher has the monthly mathematics marks of <strong>50 students</strong> in an Excel file.</p><p>The school also has a web portal where the teacher needs to enter each student’s marks.</p><p>Without automation:</p><pre>Excel<br> ↓<br>Read Kamal&#39;s marks<br> ↓<br>Open website<br> ↓<br>Enter marks<br> ↓<br>Save<br> ↓<br>Go back to Excel<br> ↓<br>Read Nimal&#39;s marks<br> ↓<br>Repeat...</pre><p>This could take a long time.</p><p>With RPA:</p><pre>Excel<br>   ↓<br>Read student<br>   ↓<br>Read marks<br>   ↓<br>Open school portal<br>   ↓<br>Enter marks<br>   ↓<br>Save<br>   ↓<br>Next student</pre><p>The bot can process the records automatically according to the rules we give it.</p><p>This is a simple example of how <strong>Excel automation + web automation</strong> can work together.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/620/1*Yr-79U2PqwY2BLTGsszPHQ.jpeg" /></figure><h3>💼 Use Case 2: Data Migration</h3><p>Consider a company that has thousands of customer records in an old system.</p><p>The company decides to move to a new system.</p><p>Someone might have to manually copy:</p><pre>Customer Name<br>Email<br>Phone Number<br>Address<br>Order Information</pre><p>from the old system into the new one.</p><p>For thousands of records, this can become a huge task.</p><p>An RPA bot can help automate the process:</p><pre>Old System<br>     ↓<br>Read Customer Data<br>     ↓<br>Validate Data<br>     ↓<br>New System<br>     ↓<br>Enter Customer Data<br>     ↓<br>Verify<br>     ↓<br>Log Result</pre><p>This type of automation can save employees from spending many hours on repetitive data-entry work.</p><h3>📧 Use Case 3: Processing Emails</h3><p>Imagine a company receives hundreds of emails every day.</p><p>Some emails contain customer requests or documents.</p><p>An RPA workflow could:</p><pre>New Email<br>   ↓<br>Read Email<br>   ↓<br>Check Subject<br>   ↓<br>Download Attachment<br>   ↓<br>Save File<br>   ↓<br>Update Database<br>   ↓<br>Send Confirmation</pre><p>For example, if an email contains an invoice, the automation could download the invoice and send it for further processing.</p><p>This is where RPA can also work together with technologies such as <strong>OCR, AI, APIs, and databases</strong>.</p><h3>05. The Big Advantages of RPA</h3><p>Why would a company spend time and money building automation?</p><p>There are several reasons.</p><h4>Less Repetitive Work</h4><p>Employees don’t have to spend their entire day doing the same clicks and data entry.</p><p>Instead, they can spend more time on tasks that require:</p><ul><li>Problem solving</li><li>Communication</li><li>Creativity</li><li>Decision making</li></ul><h4><strong>Speed</strong></h4><p>A bot can process large numbers of records much faster than a person in many repetitive tasks.</p><p>For example:</p><pre>Human<br>100 records<br>       ↓<br>Several hours<br><br>RPA<br>100 records<br>       ↓<br>Potentially much faster</pre><p>The actual time depends on the application, workflow, network, and complexity of the process.</p><h4><strong>Consistency</strong></h4><p>A bot follows the same workflow every time.</p><p>For example:</p><pre>Read data<br> ↓<br>Validate<br> ↓<br>Enter data<br> ↓<br>Submit<br> ↓<br>Log result</pre><p>It doesn’t decide to skip a step because it is tired or distracted.</p><h4>Better Use of Employee Time</h4><p>One of the biggest benefits isn’t simply “robots replace people.”</p><p>A better way to think about it is:</p><blockquote><strong><em>Let software handle repetitive work so people can focus on work that needs human thinking.</em></strong></blockquote><p>Humans and automation can work together.</p><h3>06. But RPA Is Not Magic</h3><p>It is important to understand that RPA has limitations.</p><p>RPA works best when the process is predictable.</p><p>For example:</p><pre>If:<br>Same website<br>Same buttons<br>Same process<br>Same rules<br>Structured data</pre><p>RPA can be a good choice.</p><p>But imagine:</p><pre>Website completely changes<br>+<br>Unstructured information<br>+<br>Different process every time<br>+<br>Human judgment required</pre><p>The automation can become much harder to maintain.</p><p>This is why modern automation often combines RPA with other technologies such as:</p><p><strong>Python + APIs + AI + OCR + databases + web automation.</strong></p><h3>🤖 So, What Is an RPA Bot?</h3><p>An <strong>RPA bot</strong> is simply the software program that performs the automated workflow.</p><p>For example:</p><pre>RPA Bot<br>   │<br>   ├── Open Excel<br>   ├── Read data<br>   ├── Open website<br>   ├── Enter data<br>   ├── Click Submit<br>   ├── Check result<br>   └── Save result</pre><p>You can think of the bot as a <strong>digital assistant</strong>.</p><p>You give it clear instructions, and it follows them.</p><h3>🔄 What Is a Workflow?</h3><p>A <strong>workflow</strong> is the sequence of steps that the bot follows.</p><p>For example:</p><pre>START<br>  ↓<br>Read Excel<br>  ↓<br>Get customer<br>  ↓<br>Open website<br>  ↓<br>Enter information<br>  ↓<br>Submit<br>  ↓<br>Check result<br>  ↓<br>Save result<br>  ↓<br>NEXT CUSTOMER<br>  ↓<br>END</pre><p>When I started learning RPA, this was one of the most important concepts for me to understand:</p><blockquote><strong><em>Automation is not just about clicking buttons. It is about designing a complete workflow.</em></strong></blockquote><h3><strong>🌱 Where Do We Go From Here?</strong></h3><p>RPA is much bigger than simply automating clicks.</p><p>A modern automation system might look like this:</p><pre>                 Business Process<br>                       ↓<br>                     RPA<br>                       ↓<br>        ┌──────────────┼──────────────┐<br>        ↓              ↓              ↓<br>     Python         Playwright       API<br>        ↓              ↓              ↓<br>       AI             Web          Database<br>        ↓<br>       OCR</pre><p>This is why I find RPA interesting.</p><p>You don’t need to know everything on day one.</p><p>You can start with a simple workflow:</p><pre>Excel → Read → Process → Save</pre><p>Then gradually move to:</p><pre>Excel → Website → API → Database</pre><p>And eventually:</p><pre>Document<br>   ↓<br>OCR<br>   ↓<br>AI<br>   ↓<br>Python<br>   ↓<br>RPA<br>   ↓<br>Database / Business Application</pre><p>RPA is not about building a physical robot.</p><p>It is about <strong>teaching software how to perform repetitive digital work</strong>.</p><p>The first step is simply learning how to look at a boring manual process and ask:</p><blockquote><strong><em>“Which steps can I automate?”</em></strong></blockquote><p>Once you start thinking that way, you are already starting to think like an automation engineer.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e14d184ec63c" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>