<?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 Florence Nwizugbe on Medium]]></title>
        <description><![CDATA[Stories by Florence Nwizugbe on Medium]]></description>
        <link>https://medium.com/@nwizugbeflorence?source=rss-aff89d81f065------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*tvkBW4PVKEvMiF26MT4pgw.jpeg</url>
            <title>Stories by Florence Nwizugbe on Medium</title>
            <link>https://medium.com/@nwizugbeflorence?source=rss-aff89d81f065------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 04:39:43 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@nwizugbeflorence/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[The Part of Security I Care About Happens Before the Configuration]]></title>
            <link>https://medium.com/@nwizugbeflorence/the-part-of-security-i-care-about-happens-before-the-configuration-52d78876e245?source=rss-aff89d81f065------2</link>
            <guid isPermaLink="false">https://medium.com/p/52d78876e245</guid>
            <dc:creator><![CDATA[Florence Nwizugbe]]></dc:creator>
            <pubDate>Sat, 24 Oct 2026 07:01:02 GMT</pubDate>
            <atom:updated>2026-10-24T07:01:02.878Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/800/0*kiDU5alxtNvZECV4.jpg" /></figure><p>For a while, a lot of my technical writing has leaned toward how-to content: how I configured something, how I deployed something, how I fixed a problem, how I got a security control working. I still enjoy that kind of writing, and I’ll keep sharing walkthroughs, especially when I build something I think others will find useful.</p><p>But I’ve realized that the part of technology I enjoy most happens before the configuration. It’s the problem. The messy, imperfect, real-world problem where there isn’t necessarily a single correct answer.</p><p>Take a company that gets acquired. Now there are two identity environments, two sets of administrators, legacy applications, different security policies, and different levels of technical debt. Meanwhile, the business teams still need everything to work. Someone has to answer: <em>how do we reduce the security risk without breaking the business?</em></p><p>That question is much more interesting to me than “how do I configure this feature?”</p><h3>What I want to write about</h3><p>I want to spend more time solving security problems under real-world constraints. That means asking:</p><ul><li>Why would we choose one architecture over another, and what assumptions are we making?</li><li>What happens when the ideal solution isn’t possible, and what risks are we accepting?</li><li>What does a particular control actually solve? What does it not solve?</li><li>Where does the technology stop and the business decision begin?</li></ul><h3>A taste of what that looks like</h3><p>I’ve been working on an identity-security architecture around exactly this kind of post-acquisition scenario: two disconnected Active Directory forests, one Microsoft Entra tenant, Zero Trust controls, identity threat detection, and an attack simulation.</p><p>I could simply write “Here’s how I configured Cloud Sync.” And I’ll probably still show that in a walkthrough. But the more interesting story is the thinking behind it. The textbook answer is to consolidate the forests, but in the real world that can take years, and the business can’t wait that long. So the question becomes: what can you do <em>now</em>, without consolidation? You can bring both identities into one tenant, apply Conditional Access, and get visibility through identity threat detection. What you can’t do is make the legacy risk disappear. Choosing that path means accepting it, at least for a while, and being honest about what you’re accepting.</p><p>That’s the part I want to write about: not just the control, but the trade-off.</p><h3>My projects will still be part of the story</h3><p>I’m not abandoning technical walkthroughs. I’ll still take you into the lab and show exactly how something was built. But I want those walkthroughs to sit alongside the bigger questions: why did I build it this way, what problem was I trying to solve, what did I learn, and what would I change in a real environment?</p><p>Security architecture is usually a negotiation between risk, business requirements, time, and what is actually possible. Those constraints are where I think the most interesting engineering decisions happen.</p><p>So, going forward, expect a little less “here’s what I configured,” and a little more “here’s the problem, here’s how I thought about it, here’s the trade-off, here’s what I built, and here’s what I learned.”</p><p>That’s the direction I’m taking my writing: solving security problems under real-world constraints.</p><p><strong>What’s the messiest constraint you’ve ever had to design a security control around?</strong> I’d like to hear it, and it might become a future post.</p><p>If you’d like to connect with me and follow more of my work, you can find me on <a href="https://www.linkedin.com/in/florence-nwizugbe.">LinkedIn</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=52d78876e245" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Building Your First Analytics Rule in Microsoft Sentinel: A Step-by-Step Guide]]></title>
            <link>https://medium.com/@nwizugbeflorence/building-your-first-analytics-rule-in-microsoft-sentinel-a-step-by-step-guide-2ead0a0fa9f3?source=rss-aff89d81f065------2</link>
            <guid isPermaLink="false">https://medium.com/p/2ead0a0fa9f3</guid>
            <category><![CDATA[cloud-security]]></category>
            <category><![CDATA[analytics]]></category>
            <category><![CDATA[microsoft-azure]]></category>
            <category><![CDATA[cloud-computing]]></category>
            <category><![CDATA[sentinel]]></category>
            <dc:creator><![CDATA[Florence Nwizugbe]]></dc:creator>
            <pubDate>Mon, 15 Jun 2026 14:32:44 GMT</pubDate>
            <atom:updated>2026-10-08T09:02:13.657Z</atom:updated>
            <content:encoded><![CDATA[<p>Analytics Rules: What They Are and How to Create Your First One</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/268/0*FGOYe2z-6mDF5oiC" /></figure><p>Your organisation is generating thousands of logs every single day; sign-ins, device activity, application events, network traffic. Microsoft Sentinel is collecting all of it.</p><p>But here is the thing: Sentinel collecting logs is not the same as Sentinel <em>protecting</em> you.</p><p>Without something actively watching those logs and raising a flag when something looks wrong, you just have a very expensive storage account. That “something” is called an <strong>Analytics Rule</strong> — and it is the heartbeat of any Sentinel deployment.</p><p>In this post, I am going to break down what analytics rules are, why they matter, and walk you through creating your very first one from scratch.</p><h4><strong>What Are Analytics Rules?</strong></h4><p>Think of Microsoft Sentinel as a security camera system. The cameras are recording everything but if no one is watching the footage, a break-in could happen and nobody would know until it was too late.</p><p>Analytics rules are the security guard watching the footage. They continuously check your log data and ask one question: <em>does anything here look like a threat?</em> When the answer is yes, Sentinel raises an alert and creates an <strong>incident</strong> for your team to investigate.</p><blockquote>An <strong>alert</strong> is the raw signal — Sentinel saying <em>“something suspicious just happened.”</em> An <strong>incident</strong> is what your security team actually investigates. Think of the alert as the smoke detector going off, and the incident as the case file opened to find out why. One incident can contain multiple alerts, keeping your team focused instead of overwhelmed.</blockquote><p>Without analytics rules, Sentinel sees everything but does nothing. With them, it becomes an active detection system.</p><p>Here is how it works: the rule runs a query against your logs on a schedule. If that query finds something that matches your conditions say, a user failing to sign in 10 times in an hour Sentinel fires an <strong>alert</strong>. That alert then becomes an <strong>incident</strong> that your security team can open and investigate.</p><p>Analytics rules help detect things like:</p><ul><li>Multiple failed sign-in attempts</li><li>Privilege escalation</li><li>Dormant accounts suddenly becoming active</li><li>Suspicious administrative actions</li><li>Data exfiltration attempts</li></ul><h4><strong>Types of Analytics Rules</strong></h4><p>There are three types of analytics rules in Sentinel. For this guide, we are focusing on <strong>Scheduled Rules</strong> — the most commonly used type and the best starting point for beginners. Here is a quick overview of all four so you know what exists:</p><p><strong>Scheduled Rules</strong> — Run a KQL query at set intervals (every 5 minutes, every hour, etc.) and trigger an alert when the query returns results. This is what we are building today.</p><p><strong>NRT (Near Real-Time) Rules</strong> — Similar to scheduled rules but process events almost immediately after ingestion. Useful when speed of detection is critical.</p><p><strong>Microsoft Security Rules</strong> — Use detections from Microsoft’s own security products like Defender for Endpoint or Defender for Identity. No custom query writing needed.</p><h4><strong>Scenario: Detecting Multiple Failed Sign-In Attempts</strong></h4><p>For this walkthrough, we are building a rule that detects users generating multiple failed sign-in attempts within a short window. This pattern is commonly associated with brute-force attacks, password spraying, and credential stuffing; all very real and very common threats.</p><p>Let’s build it.</p><p><strong>Step 1 — Navigate to Analytics</strong></p><p>In the Microsoft Sentinel portal, open your Sentinel workspace. From the left-hand menu, navigate to <strong>Configuration</strong> → <strong>Analytics</strong>.</p><p>This is where all your existing rules and rule templates live.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*BJ1nlsqoK41i3sKSbnTpIA.png" /></figure><p><strong>Step 2 — Create a New Scheduled Query Rule</strong></p><p>Click <strong>+ Create</strong> → <strong>Scheduled query rule</strong>.</p><p>Sentinel will open the Analytics Rule Wizard — a tabbed form where you configure everything about your rule.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*lcVp2nEGeGKjoRYm-Hm6iQ.png" /></figure><p><strong>Step 3 — General Settings</strong></p><p>Fill in the following:</p><p><strong>Name:</strong> Multiple Failed Sign-In Attempts</p><p><strong>Description:</strong> Detects users generating multiple failed Azure AD sign-in attempts within a specified timeframe.</p><p><strong>Severity:</strong> Medium</p><p><strong>MITRE ATT&amp;CK Mapping:</strong> Select <em>Credential Access</em> → <em>T1110 — Brute Force</em></p><p>The MITRE mapping is just a tag that tells Sentinel and your analysts what category of attack this rule is looking for. It adds useful context to your incidents later.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xgHaNm_lylK4aEeocPNWZw.png" /></figure><p><strong>Step 4 — Define the Detection Logic</strong></p><p>Click the <strong>Set rule logic</strong> tab. This is where you write the KQL query that powers your detection.</p><p>Paste in the following:</p><p>kql</p><pre>SigninLogs<br>| where ResultType != 0<br>| summarize FailedAttempts = count() by UserPrincipalName<br>| where FailedAttempts &gt;= 5</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/659/1*L_FKA0ykQROjCHqkNv8ntw.png" /></figure><p>Here is what each line does in plain English:</p><ul><li>SigninLogs — look in the sign-in logs table</li><li>where ResultType != 0 — filter for failed sign-ins only (a result of 0 means success)</li><li>summarize FailedAttempts = count() by UserPrincipalName — count how many failures each user has</li><li>where FailedAttempts &gt;= 5 — only return users with 5 or more failures</li></ul><blockquote><strong><em>Note:</em></strong><em> The SigninLogs table requires the Microsoft Entra ID connector to be enabled in your Sentinel workspace. If you don’t see results, check your data connectors first.</em></blockquote><p><strong>Step 5 — Query Scheduling</strong></p><p>Still on the same tab, scroll down to <strong>Query scheduling</strong>. Set:</p><ul><li><strong>Run query every:</strong> 5 minutes</li><li><strong>Lookup data from the last:</strong> 5 hours</li></ul><p>This means Sentinel checks your logs every 5 minutes, scanning the past hour of data each time. The shorter the interval, the faster you detect but it also uses more resources, so find a balance that works for your environment.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/536/1*FN0GTdPmu89womLvSTLmZQ.png" /></figure><p><strong>Step 6 — Entity Mapping</strong></p><p>This step is easy to overlook but makes a big difference in your investigation experience.</p><p>Entity mapping tells Sentinel which fields in your query results correspond to real-world objects like users, IP addresses, or devices. When Sentinel knows this, it can enrich your incidents with additional context automatically.</p><p>For this rule, map:</p><ul><li><strong>Entity type:</strong> Account → <strong>Field:</strong> UserPrincipalName</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/661/1*YmLwzLmfX95ZMX0Ekru6kQ.png" /></figure><p><strong>Step 7 — Incident Settings</strong></p><p>Click the <strong>Incident settings</strong> tab. Make sure <strong>Create incidents from alerts triggered by this analytics rule</strong> is enabled.</p><p>You can also configure <strong>Alert grouping</strong> here — this controls whether multiple alerts get bundled into a single incident or each fires separately. Leave this at the default for now.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/660/1*exmrgj3Ua2BlfLSNSrgkdw.png" /></figure><p><strong>Step 8 — Review and Create</strong></p><p>Click <strong>Review + create</strong>. Sentinel will validate your rule configuration. If everything looks good, click <strong>Create</strong>.</p><p>Your rule is now live.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/663/1*rZ-TjVgDK2rV9CTLlY02mg.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/619/1*DBxjxRcJc21vg9qOgsnXwA.png" /></figure><p><strong>Viewing Your Rule in Action</strong></p><p>Your new rule will now appear in the <strong>Active rules</strong> list on the Analytics page. You can see its name, severity, the last time it ran, and its current status.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/879/1*sb1Y1hTupwdRSIYvLz-hPA.png" /></figure><p>When the rule triggers meaning a user accumulates 5 or more failed sign-ins within the lookback window, Sentinel creates an incident in the <strong>Incidents</strong> blade. Opening that incident shows you exactly which user triggered it, the timeline of events, and any related activity.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*bdizv02K8-149vbqtod5bA.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/1*1lt2M9CCFLXhT0_zj0jtQw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Zhg4NDgIjArPedl2WvmI6w.png" /></figure><blockquote><strong><em>Tip:</em></strong><em> Use the </em><strong><em>Test with current data</em></strong><em> button inside the rule logic tab to preview what your query returns before the rule goes live. It is a great way to validate your detection without waiting for a real event.</em></blockquote><p><strong>Why This Matters</strong></p><p>Before you had this rule, Sentinel was collecting your sign-in data and doing nothing with it. Now it is actively watching for a specific attack pattern and will alert your team the moment it spots it; automatically, every 5 minutes, around the clock.</p><p>That is the shift analytics rules create. They move you from reactive (finding out about an attack after the damage is done) to proactive (catching it while it is happening).</p><p>As you get more comfortable, you can build rules for dozens of other scenarios — privilege escalation, lateral movement, data exfiltration. Each one adds another layer to your detection coverage.</p><p><strong>Bonus: Using Microsoft’s Built-In Rule Templates</strong></p><p>Now that you have built a rule from scratch, you understand exactly what goes into one; the query, the scheduling, the incident settings, all of it. That knowledge matters because Sentinel actually comes loaded with hundreds of ready-made templates you can use instead of starting from zero every time.</p><p>To access them, go to <strong>Analytics</strong> and click the <strong>Rule templates</strong> tab.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*lBW6YRLhMnkaLDczDC9Rng.png" /></figure><p>What you are looking at is a library of <strong>244 pre-built detection rules</strong> created by Microsoft’s security researchers. Each one is mapped to a MITRE ATT&amp;CK technique, tagged with a severity level, and ready to deploy. They cover a wide range of threats like credential attacks, persistence, lateral movement, data exfiltration, and more.</p><p>You can also filter by severity, tactic, or data source to find rules relevant to your environment quickly.</p><p>To use a template, simply click on it and select <strong>Create rule</strong>. Sentinel will pre-fill the entire wizard with the template’s configuration, you can use it as it is or tweak it to fit your needs.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xjs1ZXQTMWXzLTloQ5pNWQ.png" /></figure><blockquote><strong><em>Tip:</em></strong><em> Templates are a great way to learn. Open a few and study the KQL queries inside them. It is one of the fastest ways to improve your detection writing skills.</em></blockquote><p><strong>What’s Next</strong></p><p>You have now got a working detection rule. But what happens after Sentinel raises an incident? Does someone on your team have to manually triage every single one?</p><p>Not if you set up <strong>Automation Rules</strong> correctly. You can automatically assign incidents, change their severity, and suppress known false positives without writing a single line of code.</p><p><em>If you’d like to connect with me or discuss cloud security projects, you can find me on </em><a href="https://www.linkedin.com/in/florence-nwizugbe"><em>LinkedIn</em></a><em>.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=2ead0a0fa9f3" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[I Built a Full SIEM Lab on Azure — Here’s Everything That Broke (and What I Learned)| Part 5]]></title>
            <link>https://medium.com/@nwizugbeflorence/i-built-a-full-siem-lab-on-azure-heres-everything-that-broke-and-what-i-learned-part-5-7b48b6a76fbc?source=rss-aff89d81f065------2</link>
            <guid isPermaLink="false">https://medium.com/p/7b48b6a76fbc</guid>
            <category><![CDATA[linux]]></category>
            <category><![CDATA[microsoft]]></category>
            <category><![CDATA[cloud-computing]]></category>
            <category><![CDATA[azure]]></category>
            <category><![CDATA[siem]]></category>
            <dc:creator><![CDATA[Florence Nwizugbe]]></dc:creator>
            <pubDate>Mon, 06 Apr 2026 14:32:45 GMT</pubDate>
            <atom:updated>2026-04-06T14:32:45.876Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*6z1zSVNdH9OVXmuy1ZhBdg.png" /></figure><p>Integrating Linux Systems With Microsoft Sentinel.</p><p>Every environment I’ve encountered in certifications was clean and homogeneous — all Windows or all cloud. But real enterprise networks are always mixed. Linux machines run web servers, databases, containerized workloads, and critical backend infrastructure. If those machines aren’t feeding logs into your SIEM, they are a blind spot that an attacker can exploit with confidence.</p><p>This part closes that gap. The goal: deploy a Linux VM, join it to the Active Directory domain, configure syslog forwarding, and get Linux events flowing into Sentinel. What sounds like a straightforward checklist turned into a multi-blocker debugging session that taught me more about Linux networking than any single exercise before it.</p><h3>Architecture Overview</h3><p>Unlike Windows, Linux doesn’t use WMI or the Windows Event Log system. Linux log forwarding relies on <strong>rsyslog</strong> — a daemon that captures log events and routes them to files or remote destinations using facility-based filtering.</p><p>Sentinel collects Linux logs through the <strong>Syslog data connector</strong>, backed by an AMA-based Data Collection Rule:</p><pre>Ubuntu VM (FlorenceCTF.local domain member)<br>    └── rsyslog daemon<br>          └── local4 facility → /var/log/local4.log<br>                └── AMA Agent<br>                      └── Syslog DCR<br>                            └── FlorenceCTF-Workspace → Syslog table</pre><p>For the domain join to work, the Linux VM also needs to be a member of the <strong>FlorenceCTF.local</strong> Active Directory domain — the same domain the Windows endpoints belong to. This makes it a properly managed endpoint rather than an isolated machine with no identity context.</p><h3>Step-by-Step Implementation</h3><h4>Step 1 — Provisioning the Linux VM</h4><p>I navigated to <strong>Azure Portal → Virtual Machines → Create → Azure Virtual Machine</strong> and configured the VM under the Basics, Disk, Networking, and Management tabs.</p><p>Key decisions made during provisioning:</p><ul><li><strong>OS:</strong> Ubuntu Server LTS</li><li><strong>Authentication type:</strong> SSH public key — more secure than password authentication, especially in a lab where domain credentials are in active use</li><li><strong>Virtual Network:</strong> Same VNet as the existing Windows endpoints — this is non-negotiable for the domain join to work</li><li><strong>Network Security Group:</strong> Pre-configured to allow inbound and outbound traffic on ports <strong>53 (DNS)</strong>, <strong>88 (Kerberos)</strong>, and <strong>389 (LDAP)</strong> — all required for Active Directory communication</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/939/1*My6-NFJZZhyyqCLMq87MmQ.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/939/1*nuzJkbmt8Z9A3cqrRyFKFQ.png" /></figure><p>After configuring the VM, Azure prompted me to download the SSH private key before completing deployment. I downloaded it immediately — once you miss this prompt, the key is gone and you’d need to generate a new one.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/638/1*miAHmQIcra9nJWG5wOSX3A.png" /></figure><p>The VM deployed successfully. Next step: connect to it</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/939/1*02OOkcmYDyK0OuHDEutXUA.png" /><figcaption>VM Deployed</figcaption></figure><h4>Step 2 — SSH into the VM and Verify Network Connectivity</h4><p>Once the VM was deployed, I connected to it via SSH using the private key downloaded during provisioning:</p><pre>ssh -i &lt;path-to-private-key&gt; azureuser@&lt;VM-public-IP&gt;</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/940/1*bO5k9oiNM3sj8cDeiQbFzQ.png" /></figure><p>Connected successfully. First order of business; check whether the Linux VM could reach the Domain Controller. I ran a basic ping test using the DC’s hostname:</p><pre>ping CTFVM-DC</pre><p>It failed. The VM could not resolve the DC’s hostname at all. The problem was the DNS. The VM was using a public DNS resolver that had no knowledge of the internal FlorenceCTF.local domain.</p><p>I’ll come back to this. First, I updated the system packages and installed the tools needed for the domain join.</p><pre>sudo apt update &amp;&amp; sudo apt upgrade -y</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/940/1*GTcUjRJRh_LGdv2hBv8uDQ.png" /></figure><p>Then installed the required packages:</p><pre>sudo apt install -y realmd sssd sssd-tools adcli packagekit</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/940/1*YaWeb-I_8Z9-85MhLCl8SA.png" /></figure><p>With the packages in place, I ran the realm discovery command — which queries the domain controller via DNS and lists domain configuration details:</p><pre>realm discover FlorenceCTF.local</pre><p>· This command queries the domain controller via DNS and lists domain configuration details.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/827/1*lVeJL1EMNLPCbtoQIqKs7w.png" /></figure><p><strong>Error 1: realm: No such realm found</strong></p><p>Exactly as expected given what we already saw with the ping. The VM’s DNS resolver was pointing at a public server; it couldn’t resolve an internal domain name. No DNS resolution means no realm discovery. No realm discovery means no domain join.</p><p><strong>The fix:</strong> Update /etc/resolv.conf to point at the Domain Controller&#39;s private IP address.</p><pre>sudo nano /etc/resolv.conf</pre><p>Added the following line:</p><pre>nameserver 10.3.1.4</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/940/1*-qDA61tL3mDnH1O-6QvIDQ.png" /></figure><p>Saved and exited. Then immediately re-tested:</p><p>After the above configurations, the ping and nslookup started working. The VM could now resolve the DC’s hostname.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/792/1*48yeHB32p2eZyc-ZIAbjtg.png" /></figure><p>· Re-ran realm discover, which then successfully detected the domain.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/895/1*qz1zJ2aNz_GzncOhukrcqw.png" /></figure><h4>Step 3 — Join the Linux VM to the Domain</h4><p>With the domain now discoverable, I ran the join command:</p><pre>sudo realm join --user=&lt;admin-user&gt; &lt;domain-name&gt;</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/908/1*p5fe2_JgQhFgXZlXjlZMNw.png" /></figure><p><strong>Error 2: Insufficient Permissions</strong></p><pre>realm join: Insufficient permissions to join the domain</pre><p>This one is deceptive. The error says “insufficient permissions” ;but the credentials were correct. The actual problem had nothing to do with permissions.</p><p>What was happening: Kerberos — the authentication protocol that Active Directory relies on — requires the joining machine’s own hostname to be resolvable from within the domain. The Linux VM’s hostname wasn’t mapped in /etc/hosts, so when the DC tried to verify the machine during the handshake, it couldn&#39;t resolve it. Kerberos failed silently and surfaced it as a permissions error.</p><p><strong>The fix:</strong> Add an explicit hostname entry to /etc/hosts.</p><pre>sudo nano /etc/hosts</pre><p>Added the following line beneath the existing entries:</p><pre>10.4.1.5  CTFVM-Lin</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/940/1*jzyL9NtXbS1YxBQD5dLlLw.png" /></figure><p>· Saved and exited.</p><p>Verified that the hostname now resolves correctly:</p><pre>ping -c 3 CTFVM-Lin</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/895/1*gEYSg8lBfBgQrN5nZTD-8Q.png" /></figure><p>· Re-ran the join command</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/939/1*R5t9OyBPo-z0Lq2Ni5ymxQ.png" /></figure><p>Join successful. The Linux VM was now a member of the FlorenceCTF.local Active Directory domain.</p><blockquote><strong><em>This is one of the most important lessons in the entire series:</em></strong><em> “Insufficient permissions” in a Linux domain join almost never means your credentials are wrong. It almost always means Kerberos can’t complete its handshake — and the first thing to check is hostname resolution, not your password.</em></blockquote><h4>Step 4 — Install and Configure rsyslog</h4><p>With the VM on the domain, the next step was setting up <strong>rsyslog</strong> — the daemon responsible for capturing and routing log messages on the Linux system.</p><p>Installed rsyslog:</p><pre>sudo apt install -y rsyslog</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/940/1*ZgWiLYi8QkEr42S3zXW-Tg.png" /></figure><p>Enabled it to start automatically and started it immediately:</p><pre>sudo systemctl enable rsyslog<br>sudo systemctl start rsyslog</pre><pre>sudo systemctl status rsyslog</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/939/1*usga3zRDnQzKQIValFuQtQ.png" /></figure><h4>Step 5 — Configure the local4 Log Facility</h4><p>rsyslog uses a facility-based system to categorise log messages. Facilities like auth, syslog, and daemon are reserved for system use. The local0 through local7 facilities are available for custom use — I used <strong>local4</strong> to keep lab log messages clearly separated from system messages.</p><p>Created a custom configuration file:</p><pre>sudo nano /etc/rsyslog.d/10-local4.conf</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/739/1*0PTa7xsZs6ygLlHOgjw1lQ.png" /></figure><p>Added the following line:</p><pre>local4.* /var/log/local4.log</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/940/1*NbyrAwJWf2AEm_svm5mO-g.png" /></figure><p>This tells rsyslog to route all messages from the local4 facility — regardless of severity level — to /var/log/local4.log.</p><p>Restarted rsyslog to apply the new configuration:</p><pre>sudo systemctl restart rsyslog</pre><p>Now I needed to test whether it was actually working. I sent a manual test message using the logger command:</p><pre>logger -p local4.info &quot;Test message from Florence and FlorenceCTF.local&quot;</pre><p>Then checked the log file:</p><pre>cat /var/log/local4.log</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/642/1*ioOthYjjMrP8H-WUJdH0nQ.png" /></figure><p>Nothing. The file was empty.</p><p>I checked rsyslog’s own logs to understand what was going wrong:</p><pre>journalctl -u rsyslog --since &quot;5 minutes ago&quot;</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/939/1*lXUQ_QXe5fskL8Zq8Etavg.png" /></figure><p>The output confirmed a <strong>permission error</strong>. The rsyslog daemon runs under a specific system user, and that user did not have write access to /var/log/local4.log. Every log message was being silently dropped because rsyslog couldn&#39;t write to the file.</p><p><strong>The fix:</strong> Update the file permissions so rsyslog can write to it.</p><pre>sudo chmod 666 /var/log/local4.log</pre><p>The message appeared. The local4 facility was working correctly.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/803/1*ZjB8NG9aKaQqlF32N5tCsQ.png" /></figure><h4>Step 6 — Automate Log Generation with Cron</h4><p>A single test message confirms the configuration works once. But for Sentinel validation purposes, I needed continuous log generation — something that would keep producing messages on a schedule so I could verify the full pipeline from rsyslog to Sentinel over time.</p><p>I set up a <strong>cron job</strong> to fire a syslog message every 30 minutes automatically.</p><p>Opened the cron scheduler:</p><pre>sudo crontab -e</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/830/1*kvYrfgcZEqp128c5lXXdSw.png" /></figure><p>Added the following line:</p><pre>*/30 * * * * /usr/bin/logger -p local4.info &quot;Test message from Florence and FlorenceCTF.local&quot;</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/940/1*XfrM8_kbi74GwOlF5I5fxQ.png" /></figure><p>· Saved and exited.</p><p>After waiting 30 minutes, I checked the log file:</p><pre>cat /var/log/local4.log</pre><p>A new entry had appeared at the 30-minute mark. The cron job was running correctly and the rsyslog pipeline was holding up.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/939/1*I0TFihgCiudgUxy6LLGngA.png" /></figure><h3>Step 7 — Connect the Syslog Data Connector in Sentinel</h3><p>With rsyslog producing logs reliably on the Linux VM, the next step was pulling those logs into Sentinel.</p><ol><li>In Sentinel, navigated to <strong>Content Management → Content Hub</strong></li><li>Searched for <strong>Syslog</strong> and installed the <strong>Syslog via AMA</strong> connector</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/940/1*pnk2-kXhp1sL-916NZojdA.png" /></figure><p>3. Opened the connector page and clicked <strong>Create Data Collection Rule</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/940/1*q7XnLm3lpJ0L2jDenF1Cuw.png" /></figure><p>4. Configured the DCR:</p><ul><li><strong>Data source:</strong> Selected the Linux VM</li><li><strong>Log facility:</strong> Selected local4 — matching exactly what was configured in rsyslog</li><li><strong>Destination:</strong> FlorenceCTF-Workspace</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/940/1*oCM5DhyNlPd40kRh-EVh2Q.png" /></figure><ul><li>Clicked <strong>Review + Create</strong></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/940/1*46bQzBsNLtlH3V21bI4MDA.png" /></figure><p>The DCR was created. The AMA agent was confirmed as installed and healthy on the Linux VM.</p><h4>Step 8 — Verify Log Ingestion in Sentinel</h4><p>· Navigated to <strong>Sentinel → Logs</strong> and ran the following KQL query to confirm the syslog messages were arriving:</p><pre>Syslog<br>| where Facility == &quot;local4&quot;<br>| where TimeGenerated &gt; ago(1h)<br>| order by TimeGenerated desc</pre><p>The messages generated by the cron job were appearing in the Sentinel Syslog table — flowing from the Linux VM through rsyslog, through the AMA agent, through the DCR, and into the workspace.</p><p>The pipeline was complete!</p><h3>Lessons Learned</h3><ul><li><strong>DNS first, always.</strong> Every Linux domain issue in this part traced back to name resolution. Before debugging anything else on a Linux machine that can’t reach a domain, check the nameserver.</li><li><strong>“Insufficient permissions” is Kerberos lying to you.</strong> It almost always means hostname resolution failed, not that your credentials are wrong. Check /etc/hosts before you change anything else.</li><li><strong>rsyslog drops messages silently.</strong> If a log file isn’t populating, run journalctl -u rsyslog and check file permissions — the daemon won&#39;t tell you it can&#39;t write.</li></ul><h3>Security Insight</h3><p>Linux machines get compromised quietly because they’re rarely monitored well. SSH brute force, cron-based persistence, and local log tampering are all trivially easy to execute on an unmonitored Linux host. Forwarding logs to a centralised SIEM means an attacker who clears /var/log/ on the machine can&#39;t touch what&#39;s already in Sentinel. That&#39;s the entire point — visibility that survives the compromise.</p><h3>Conclusion — End of Series</h3><p>Five parts. Two modules. Every blocker documented.</p><p>We went from an empty Azure subscription to a fully instrumented SIEM environment — Entra ID identity logs, Defender alerts, NSG network flows, Windows endpoint events, and now Linux syslog — all flowing into a single workspace, all queryable with KQL, all validated with real data.</p><p>The honest truth about this series: the configuration steps are the easy part. The blockers are where the real learning happens. The diagnostic settings wall, the DNS failure, the Kerberos misdirection, the silent rsyslog permission drop — none of those come from a study guide. They come from actually building the thing.</p><p>If this series helped you, the next step is going deeper on detection; writing analytics rules, hunting with KQL, and building incident response workflows on top of the data you’ve just ingested.</p><p><strong>For that, I’ll point you to a resource I found genuinely useful:</strong></p><h4><strong>Further Reading</strong></h4><p><a href="https://www.amazon.com/Microsoft-Sentinel-Responding-Cloud-Native-Operations-ebook/dp/B0GRWN9XDB/"><strong>Microsoft Sentinel: A Practical Guide to Designing, Deploying, Detecting, Hunting, and Responding with Microsoft Sentinel, the Cloud-Native SIEM for Modern Security Operations</strong></a></p><p>If you want to go beyond the infrastructure layer and into detection engineering, threat hunting, and incident response with Sentinel — this book covers exactly that.</p><p><em>If you’d like to connect with me or discuss cloud security projects, you can find me on </em><a href="https://www.linkedin.com/in/florence-nwizugbe"><em>LinkedIn</em></a><em>.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7b48b6a76fbc" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[I Built a Full SIEM Lab on Azure — Here’s Everything That Broke (and What I Learned)| Part 4]]></title>
            <link>https://medium.com/@nwizugbeflorence/i-built-a-full-siem-lab-on-azure-heres-everything-that-broke-and-what-i-learned-part-4-cd3058c70b8a?source=rss-aff89d81f065------2</link>
            <guid isPermaLink="false">https://medium.com/p/cd3058c70b8a</guid>
            <category><![CDATA[cloud-computing]]></category>
            <category><![CDATA[cyberse]]></category>
            <category><![CDATA[microsoft-sentinel]]></category>
            <category><![CDATA[siem]]></category>
            <category><![CDATA[azure]]></category>
            <dc:creator><![CDATA[Florence Nwizugbe]]></dc:creator>
            <pubDate>Mon, 23 Mar 2026 10:32:27 GMT</pubDate>
            <atom:updated>2026-03-23T10:32:27.594Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/700/0*b1ioa8HCLwhWk5CU.png" /></figure><p>Here’s a trap that catches a lot of people building out a SIEM for the first time: they configure the connectors, see a green “Connected” badge, and move on assuming the data is flowing.</p><p>It often isn’t.</p><p>Connector status reflects whether the configuration was accepted — not whether data is arriving, whether the agent is healthy, or whether the correct log categories are being forwarded. The only way to know for certain is to query the workspace and look for records. This part is about building that verification habit — and the agent infrastructure that makes endpoint monitoring possible.</p><h3>Architecture Overview</h3><p>Cloud-native connectors forward logs automatically through diagnostic settings. But Windows endpoint security events — logon activity, process creation, privilege use — require an agent running on each machine.</p><p>The modern Microsoft agent is the <strong>Azure Monitor Agent (AMA)</strong>, governed by a <strong>Data Collection Rule (DCR)</strong> that defines source machines, event types, and the destination workspace.</p><pre>Domain Controller ──┐<br>                    ├──→ Azure Monitor Agent (AMA)<br>Workstation 1 ──────┤         │<br>                    │         ▼<br>Workstation 2 ──────┘   Data Collection Rule (DCR)<br>                              │<br>                              ▼<br>                    FlorenceCTF-Workspace → SecurityEvent table</pre><h3>Step-by-Step Implementation</h3><h3>1. Deploy the Azure Monitor Agent via Data Collection Rule</h3><ul><li>Navigate to <strong>Microsoft Sentinel</strong></li><li>Go to <strong>Configuration → Data Connectors</strong></li><li>Search for and select <strong>Windows Security Events via AMA</strong></li><li>Click <strong>Install</strong></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/1*yyqe4HAEtKmF2suzwn2QWg.png" /></figure><ul><li>Under the <strong>Instructions</strong> tab, click <strong>Create Data Collection Rule (DCR)</strong></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*eGsWRFDDbf_gNxarKXjTZg.png" /></figure><p>Configure the Data Collection Rule:</p><ul><li><strong>Name:</strong> (based on your naming convention)</li><li><strong>Subscription:</strong> Select your active subscription</li><li><strong>Resource Group:</strong> Choose your existing resource group</li><li><strong>Region:</strong> Ensure it matches your workspace region</li><li><strong>Destination:</strong> Select your Log Analytics workspace (<em>FlorenceCTF-Workspace</em>)</li></ul><p>Under <strong>Resources</strong>:</p><ul><li>Click <strong>Add resources</strong></li></ul><p>Select all endpoints to monitor:</p><ul><li>Domain Controller</li><li>Workstations</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/865/1*wIuKT6w65mYoTIsPkx8GNQ.png" /></figure><ul><li>Click <strong>Review + Create</strong></li><li>Click <strong>Create</strong> to deploy the Data Collection Rule</li></ul><p>Azure automatically installs the AMA extension on each selected VM.</p><ul><li>Navigate to your Workspace → <strong>Agents</strong> to confirm the agent installed successfully.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*VcrvRgDUPwuEpvTXqSQ3dg.png" /></figure><h3>2. Verify Agent Communication</h3><p>This is to confirm that all connected endpoints (Domain Controller and Workstations) are actively communicating with Microsoft Sentinel via the Azure Monitor Agent (AMA) and reporting to the correct Log Analytics workspace.</p><h4>Heartbeat Query</h4><p>The Heartbeat query is the first thing I check after onboarding any machine. It tells me whether my endpoints are actually talking to Microsoft Sentinel — because if they’re not, nothing else matters.</p><ul><li>Navigate to Microsoft Sentinel → Log Analytics Workspace → Logs.</li><li>Select Table → Click on Log Management →Scroll down and click Run on the Heartbeat.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/978/1*8XmjMWipyJkYRCeMyijYvQ.png" /></figure><ul><li>You can also run the query below in the Log Analytics query editor of your workspace.</li><li>Query:</li></ul><pre>// Show last heartbeat time per machine in the last 24 hours<br>Heartbeat<br>| where TimeGenerated &gt; ago(24h)<br>| summarize LastSeen = max(TimeGenerated) by Computer, OSType, _ResourceId<br>| order by LastSeen desc</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*wkmwWb5HkrXDTMymNhXHLg.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*utB-Yt6PAa4_bsK0VYrc2Q.png" /><figcaption>Result of the Heartbeat Query</figcaption></figure><p>Every enrolled machine should appear with a recent timestamp. If a machine is missing, it likely means the agent is not reporting; and this should be investigated before proceeding.</p><p><strong>Note:</strong> You can adjust the time range in your query to retrieve more results or view historical activity.</p><h3>3. Validate Data Ingestion</h3><p>Data ingestion was validated by running KQL queries across key log sources. This confirmed that events were successfully flowing into Microsoft Sentinel and were available for analysis.</p><h4>Sign-in Logs</h4><p>Authentication activity was verified using sign-in logs.</p><pre>SigninLogs<br>| where TimeGenerated &gt; ago(24h)<br>| top 10 by TimeGenerated desc<br></pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*voWmzgi4aZL1c7FmPI2NUg.png" /></figure><h4>Audit Logs</h4><p>Administrative and configuration changes were validated using audit logs.</p><pre>AuditLogs<br>| where TimeGenerated &gt; ago(7d)<br>| top 10 by TimeGenerated desc</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*eLQtNkV7lSajWyQ3cKIdZw.png" /></figure><h4>NSG Logs</h4><p>Network traffic and rule activity were validated using NSG logs.</p><pre>AzureDiagnostics<br>| where ResourceType == &quot;NETWORKSECURITYGROUPS&quot; or Category == &quot;NetworkSecurityGroupFlowEvent&quot;<br>| where TimeGenerated &gt; ago(24h)<br>| take 50</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*VL1MP1K203OIaKPmoCOTow.png" /></figure><p><strong>Note:</strong> Each query should return records. If no results are returned, it means no data is being ingested — and that’s a problem to fix immediately, not after building detections on top of empty tables.</p><h3>Real-World Blocker Encountered</h3><p>After deploying the DCR and confirming the AMA extension showed as installed, the SecurityEvent query returned no results. The heartbeat was reporting correctly, but the security events table was empty. It created real uncertainty: was the DCR misconfigured? Was there a permissions issue? Was the agent silently failing?</p><h3>Resolution</h3><p>The issue was <strong>ingestion latency</strong>. The AMA, once installed, takes 5–15 minutes before events begin appearing in the workspace. Rather than reconfiguring the DCR, I waited a full 15 minutes and re-ran the query.</p><p>Records appeared. The pipeline was functioning correctly the entire time.</p><p>The practical lesson: when an agent is freshly deployed, always wait at least 15 minutes before diagnosing an ingestion failure. Premature troubleshooting can lead to unnecessary configuration changes and wasted effort while chasing what is actually a timing issue.</p><h3>Lessons Learned</h3><ul><li><strong>Green icons are not evidence. KQL results are.</strong> Build query-based validation as a non-negotiable step in every Sentinel deployment.</li><li><strong>The Heartbeat table is your first diagnostic tool.</strong> If a machine isn’t in Heartbeat, it’s not sending anything — agent communication is the prerequisite for everything else.</li><li><strong>Patience is part of the process.</strong> Cloud telemetry pipelines have inherent latency. Waiting 15 minutes is not wasted time; it’s correct process.</li></ul><h3>Security Insight</h3><p>The <strong>SecurityEvent</strong> table is one of the most critical data sources in Microsoft Sentinel. It contains Windows security events that form the foundation for most endpoint-based detections.</p><p>Understanding key Windows Event IDs is essential for building meaningful security analytics:</p><ul><li><strong>4624</strong> — Successful logon</li><li><strong>4625</strong> — Failed logon attempt</li><li><strong>4648</strong> — Logon using explicit credentials</li><li><strong>4672</strong> — Special privileges assigned to a logon session</li><li><strong>4688</strong> — New process creation</li></ul><p>Detection engineers write analytics rules against this table to identify brute force, pass-the-hash, privilege escalation, and lateral movement. Without this table populated and validated, none of those detections fire.</p><h3>Conclusion</h3><p>The pipeline is validated. Every data source — Entra ID, Defender, NSG, and Windows endpoints — has confirmed records in the workspace. The SIEM is no longer just deployed; it’s ingesting real data and ready to be queried.</p><p>But the environment is still exclusively Windows and cloud. Most production environments have Linux. In Part 5, that changes.</p><h3>What’s Next in the Series</h3><p><strong>Part 5 </strong>covers the most technically complex part of this project so far: deploying a Linux VM, joining it to the Active Directory domain through two separate blockers, configuring rsyslog with a custom facility, automating log generation with cron, and connecting the Syslog data connector in Sentinel.</p><p><em>If you’d like to connect with me or discuss cloud security projects, you can find me on </em><a href="https://www.linkedin.com/in/florence-nwizugbe"><em>LinkedIn</em></a><em>.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=cd3058c70b8a" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[I Built a Full SIEM Lab on Azure — Here’s Everything That Broke (and What I Learned)| Part 3]]></title>
            <link>https://medium.com/@nwizugbeflorence/i-built-a-full-siem-lab-on-azure-heres-everything-that-broke-and-what-i-learned-part-3-5d1ffbfd8707?source=rss-aff89d81f065------2</link>
            <guid isPermaLink="false">https://medium.com/p/5d1ffbfd8707</guid>
            <category><![CDATA[azure]]></category>
            <category><![CDATA[logs]]></category>
            <category><![CDATA[linux]]></category>
            <category><![CDATA[microsoft-sentinel]]></category>
            <category><![CDATA[cloud-computing]]></category>
            <dc:creator><![CDATA[Florence Nwizugbe]]></dc:creator>
            <pubDate>Mon, 16 Mar 2026 15:19:12 GMT</pubDate>
            <atom:updated>2026-04-25T09:28:03.549Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*hI8mQ2R47r8SJF8dtpbVZw.png" /></figure><p>An enabled Sentinel workspace with no connected data sources is like installing a security camera with no feed. The platform is ready; but it’s watching nothing.</p><p>The goal of this part is straightforward: connect Sentinel to the three most important cloud-native data sources in the environment — <strong>Microsoft Entra ID</strong> (identity logs), <strong>Microsoft Defender for Cloud</strong> (security alerts), and <strong>Azure NSG flow logs</strong> (network traffic). Together, these three sources give you visibility into who is signing in, what threats the platform is detecting, and what traffic is crossing your network.</p><p>What I didn’t expect was a hard platform limit that stopped one of the connectors cold.</p><h3>Architecture Overview</h3><p>Data connectors are the ingestion layer of Sentinel. Each connector establishes a pipeline between a Microsoft service and the Log Analytics workspace, using <strong>Diagnostic Settings</strong> — a mechanism where Azure forwards log categories from a resource to a destination workspace.</p><pre>Microsoft Entra ID ──────────→ SigninLogs, AuditLogs, ProvisioningLogs<br>Microsoft Defender for Cloud ──→ Security Alerts, Recommendations<br>Azure NSG Flow Logs ─────────→ NetworkSecurityGroupEvent, NSGRuleCounter<br>              │<br>              └──── All flowing into: FlorenceCTF-Workspace (Sentinel)</pre><p>Each connector is installed from Sentinel’s <strong>Content Hub</strong> and activated through its own configuration page.</p><h3>Step-by-Step Implementation</h3><h3>Connector A: Microsoft Entra ID</h3><ol><li><strong>Sentinel → Configuration → Data Connectors → Content Hub</strong></li><li>Search for and install <strong>Microsoft Entra ID</strong></li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*jWYgGk4JtnjLiGO6Mc7pFw.png" /></figure><p>3. Open the connector page and select log categories: Sign-in Logs, Audit Logs, Provisioning Logs</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*4Gj1bIgITWC7vx7wTQ35OQ.png" /></figure><p>4. Click <strong>Apply Changes</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*7UNGIjonLL2_mChpUchp1Q.png" /></figure><h3>Connector B: Microsoft Defender for Cloud</h3><ol><li><strong>Sentinel → Data Connectors → Content Hub → Microsoft Defender for Cloud</strong></li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Iph7hOGZjo2hyBYMCh_JDQ.png" /></figure><p>2. Click <strong>Install</strong></p><p>3. Configure the <strong>Subscription-level</strong> connection</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*jiOFW3u7hApv38H4X7Jjjw.png" /></figure><p>4. Click <strong>Connect</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/947/1*ZYwxbXAC2zbIDWp-PlFg1g.png" /></figure><p>The data connector is now showing connected.</p><h3>Connector C: Azure NSG Flow Logs</h3><ol><li>Enable <strong>Network Watcher</strong> in East US</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/939/1*Qx9UCYkbjgwNTx0QacqfRg.png" /></figure><p>2. For each NSG: <strong>Diagnostic Settings → Add diagnostic setting</strong></p><p>3. Select <strong>Send to Log Analytics workspace</strong> → FlorenceCTF-Workspace</p><p>4. Enable NetworkSecurityGroupEvent and NetworkSecurityGroupRuleCounter</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/939/1*JLlRpvCibHwL8ujPmO5yuQ.png" /></figure><p>5. Save</p><h3>Real-World Blocker Encountered</h3><p>While configuring the <strong>Microsoft Entra ID</strong> connector and clicking <strong>Apply Changes</strong>, the configuration failed:</p><pre>Error: The resource has reached its maximum limit of 5 diagnostic settings.</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*J9gQaN0QEb7jYF7ISMi7rw.png" /></figure><p>Azure enforces a hard cap of <strong>5 diagnostic settings per resource</strong>. My Entra ID tenant — used across multiple prior lab exercises — already had 5 settings configured, leaving zero room for Sentinel to add its own. The Sentinel UI gave no warning before the attempt. It only surfaced the error after trying to apply.</p><h3>Resolution</h3><ol><li>Navigate to: <strong>Microsoft Entra ID → Monitoring → Diagnostic settings</strong></li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*kDM2CsYW9Kc4INNKU_F2CQ.png" /></figure><p>2. Review all existing settings — identify unused, duplicate, or settings pointing to deleted workspaces</p><p>3. Delete one redundant entry to free up a slot</p><p>4. Return to the Sentinel connector page and re-apply the configuration</p><p>After clearing the slot, the connector applied successfully and logs began flowing within minutes.</p><h3>Lessons Learned</h3><ul><li><strong>Always audit diagnostic settings before configuring connectors</strong> in any shared or existing Azure tenant. The limit of 5 is a hard platform constraint that Sentinel does not warn you about upfront.</li><li><strong>Shared tenants accumulate technical debt.</strong> Old lab exercises, test configurations, and forgotten connectors pile up in diagnostic settings. Clean them up proactively.</li><li><strong>Connector status is not the same as data flow.</strong> After resolving the blocker, I waited 5 minutes and ran a KQL query to confirm logs were actually arriving. Status indicators update faster than reality.</li></ul><h3>Security Insight</h3><p>The three connectors configured here form the <strong>detection trifecta</strong> of any Azure-based security program:</p><ul><li><strong>Entra ID logs</strong> tell you <em>who</em> is accessing your environment and from where. Sign-in failures, risky sign-ins, and impossible travel detections all start here.</li><li><strong>Defender for Cloud alerts</strong> tell you <em>what threats</em> the platform has identified — misconfigured resources, malware detections, lateral movement attempts.</li><li><strong>NSG flow logs</strong> tell you <em>how</em> traffic is moving. East-west movement between subnets, unexpected outbound connections, and port scanning all appear in this data.</li></ul><p>Without all three, you have partial visibility. A sophisticated attacker will move through the gap you left uncovered.</p><h3>Conclusion</h3><p>Three connectors are now active. The environment has visibility into identity, security posture, and network traffic. But seeing data appear in the portal is not the same as confirming it in the workspace. In Part 4, I deploy agents to the Windows endpoints and use KQL to validate every log source before trusting any of it.</p><h3>What’s Next in the Series</h3><p><strong>Part 4 </strong>covers deploying the <strong>Azure Monitor Agent (AMA)</strong> to the Domain Controller and workstations, creating Data Collection Rules, running heartbeat queries, and using KQL to confirm that every log type is present and queryable in the workspace.</p><p><em>If you’d like to connect with me or discuss cloud security projects, you can find me on </em><a href="https://www.linkedin.com/in/florence-nwizugbe"><em>LinkedIn</em></a><em>.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5d1ffbfd8707" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[I Built a Full SIEM Lab on Azure — Here’s Everything That Broke (and What I Learned)| Part 2]]></title>
            <link>https://medium.com/@nwizugbeflorence/i-built-a-full-siem-lab-on-azure-heres-everything-that-broke-and-what-i-learned-part-2-dca60c5e1e9d?source=rss-aff89d81f065------2</link>
            <guid isPermaLink="false">https://medium.com/p/dca60c5e1e9d</guid>
            <category><![CDATA[soc-analyst]]></category>
            <category><![CDATA[cybersecurity]]></category>
            <category><![CDATA[cloud-security]]></category>
            <category><![CDATA[siem]]></category>
            <category><![CDATA[azure]]></category>
            <dc:creator><![CDATA[Florence Nwizugbe]]></dc:creator>
            <pubDate>Mon, 09 Mar 2026 11:34:59 GMT</pubDate>
            <atom:updated>2026-04-24T20:33:29.705Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/303/1*yD49toEEz_luixtOC71IOQ.png" /></figure><p>There’s a gap that every aspiring SOC analyst and cloud security engineer eventually has to cross: the gap between understanding how a SIEM works in theory and actually knowing how to build one.</p><p>Most lab guides hand you a pre-configured workspace and skip straight to the alerts. But in a real enterprise environment, you inherit messy tenants, hit permission walls, and discover that “connected” doesn’t always mean “working.” I wanted to build Microsoft Sentinel the hard way — from an empty Azure subscription, configuring every layer myself, and documenting every blocker I hit along the way.</p><p>This first part is about the thing nobody talks about because it isn’t glamorous: setting up the workspace that everything else depends on. Get this wrong and nothing downstream works.</p><h3>Prerequisites</h3><p>Before starting the lab, a few foundational resources were already in place to allow the environment to generate and send logs to the SIEM.</p><ul><li><strong>Azure Subscription</strong> — Required to create and manage resources such as virtual machines, networking components, and the Log Analytics Workspace<strong> </strong>used by Microsoft Sentinel.</li><li><strong>Microsoft Entra ID Tenant</strong> — Needed to collect identity-related logs such as Sign-in Logs, Audit Logs, and Provisioning Logs.</li><li><strong>Virtual Machines</strong> — Three Windows VMs and one<strong> </strong>Linux machine<strong> </strong>were deployed to act as endpoints that generate security and system logs. I made use of the VMs I spinned up in this <a href="https://medium.com/@nwizugbeflorence/building-an-on-premise-active-directory-infrastructure-from-scratch-part-2-installing-and-3affd8f19b11?source=your_stories_outbox---writer_outbox_published-----------------------------------------">post</a> .</li><li><strong>Basic Networking</strong> — The virtual machines were connected to an Azure<strong> </strong>Virtual Network to enable communication and generate network-related logs.</li><li><strong>Administrative Permissions</strong> — Required to configure Microsoft Sentinel, enable data connectors, and install agents on the machines.</li></ul><p>With these components in place, the lab environment can simulate a small infrastructure where multiple systems generate logs that feed into the SIEM for monitoring and analysis.</p><h3>Architecture Overview</h3><p>Microsoft Sentinel is not a standalone product — it’s a layer built on top of <strong>Azure Log Analytics</strong>. Before Sentinel can ingest a single log, you need a <strong>Log Analytics Workspace (LAW)</strong> deployed in the right region, attached to the right subscription, and sized appropriately.</p><p>The architecture for this lab looks like this:</p><pre>Azure Subscription<br>  └── Resource Group: CaptureTheFlag<br>        └── Log Analytics Workspace: FlorenceCTF-Workspace (East US)<br>              └── Microsoft Sentinel (enabled on top of LAW)</pre><p>Everything in this series — data connectors, agents, Linux syslog — feeds into this single workspace. Choosing the right region matters: placing the workspace in the same region as your resources avoids cross-region data transfer costs and latency.</p><h3>Step-by-Step Implementation</h3><h3>Step 1: Creating the Log Analytics Workspace</h3><p>Before Sentinel can do anything, it needs a home. That home is a <strong>Log Analytics Workspace (LAW)</strong> — the data store that Sentinel sits on top of and queries.</p><p>I navigated to <strong>Azure Portal → Create a resource → Log Analytics workspace</strong> and configured it as follows:</p><ul><li><strong>Subscription:</strong> My active Azure subscription</li><li><strong>Resource Group:</strong> CaptureTheFlag (pre-existing lab resource group)</li><li><strong>Workspace Name:</strong> FlorenceCTF-Workspace</li><li><strong>Region:</strong> East US — matching all my other resources to avoid cross-region latency</li><li><strong>Pricing Tier:</strong> Pay-as-you-go</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/783/1*AIyau8UuR8NuYVw31gHGJg.png" /></figure><p>After clicking <strong>Review + Create</strong>, click create.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/782/1*sdUIULdR-2qQGwyv3udkvQ.png" /></figure><p>The deployment took about 2–4 minutes. It’s a small step, but it’s the foundation everything else rests on.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/801/1*lbGEx4H4EdBNcMCsDxWEdA.png" /><figcaption>Deployment in progress…</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/938/1*qyISfWbzDwC4H4FueKnQVQ.png" /><figcaption>Deployment succeeded..</figcaption></figure><h3>Step 2: Adding Microsoft Sentinel to the Workspace</h3><p>Once the workspace was live, adding Sentinel was surprisingly simple:</p><ol><li>Search for <strong>Microsoft Sentinel</strong> in the portal</li><li>Click <strong>Create</strong></li><li>Select the FlorenceCTF-Workspace</li><li>Click <strong>Add</strong></li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/711/1*p9LwkJU729iLyM40REC_Xw.png" /></figure><p>That’s it. What you’re really doing here is enabling the Sentinel feature set on top of the Log Analytics workspace — it’s not a separate resource, it’s an extension that unlocks analytics rules, incident management, threat intelligence, and workbooks. Within seconds, the Sentinel dashboard appeared, empty and waiting for data.</p><h3>Real-World Blocker Encountered</h3><p>The blocker in this step wasn’t an error message — it was a decision. When creating the workspace, there’s a temptation to accept whichever region defaults. I nearly deployed to West US before catching that all my other lab resources — VMs, NSGs, VNets — were in <strong>East US</strong>.</p><p>A workspace in the wrong region creates two problems: cross-region egress costs for log data, and potential data residency compliance issues. There’s no warning in the portal. It defaults and moves on.</p><h3>Resolution</h3><p>Before clicking Create, cross-check your workspace region against the region of every resource you plan to monitor. In this case, the fix was a single dropdown change to <strong>East US</strong> before deploying. A five-second check that prevents a resource rebuild later.</p><h3>Lessons Learned</h3><ul><li>A Log Analytics Workspace is not optional infrastructure — it <em>is</em> the SIEM. Sentinel is the intelligence layer; the workspace is the data layer. They are inseparable.</li><li>Name your workspace intentionally. FlorenceCTF-Workspace maps directly to the project, making it easy to identify in a subscription with multiple workspaces.</li><li>Region alignment isn’t a best practice — it’s a cost and compliance requirement. Get it right at deployment, not after.</li></ul><h3>Security Insight</h3><p>The Log Analytics Workspace is one of the most sensitive resources in your Azure environment. Anyone with <strong>Contributor</strong> or <strong>Log Analytics Contributor</strong> access can query all ingested data — including sign-in logs, audit trails, and security events. Apply <strong>RBAC</strong> to your workspace from day one:</p><ul><li><strong>Security Readers</strong> → read-only access to logs and queries</li><li><strong>Security Contributors</strong> → can create analytics rules and workbooks</li></ul><p>No one should have Owner access to the workspace in production unless strictly necessary. Treat the workspace like a vault. Its contents represent your entire visibility into the environment.</p><h3>Wrapping Up</h3><p>An empty Sentinel workspace is just a container. But it’s the most important container in your cloud security stack. Getting the foundation right — correct region, intentional naming, RBAC applied — means everything built on top of it will work cleanly. The workspace is live. Sentinel is enabled. Now it needs data.</p><h3>What’s Next in the Series</h3><p>In <strong>Part 3</strong>, I connect three critical data sources — Microsoft Entra ID, Defender for Cloud, and Azure NSG flow logs — and run directly into a hard Azure platform limit that blocked the entire connector configuration. The fix was simple once I understood it, but finding it took longer than it should have.</p><p><em>If you’d like to connect with me or discuss cloud security projects, you can find me on </em><a href="https://www.linkedin.com/in/florence-nwizugbe"><em>LinkedIn</em></a><em>.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=dca60c5e1e9d" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Essential Terms to Know Before Building a SIEM on Azure]]></title>
            <link>https://medium.com/@nwizugbeflorence/essential-terms-to-know-before-building-a-siem-on-azure-3492aa441df6?source=rss-aff89d81f065------2</link>
            <guid isPermaLink="false">https://medium.com/p/3492aa441df6</guid>
            <category><![CDATA[cloud-computing]]></category>
            <category><![CDATA[azure]]></category>
            <category><![CDATA[microsoft]]></category>
            <category><![CDATA[data-connectors]]></category>
            <category><![CDATA[cybersecurity]]></category>
            <dc:creator><![CDATA[Florence Nwizugbe]]></dc:creator>
            <pubDate>Fri, 06 Mar 2026 11:46:36 GMT</pubDate>
            <atom:updated>2026-03-06T11:46:36.901Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/800/1*P1sq5H9E9j0Kc5HtP8Fyhg.jpeg" /></figure><p>Before we dive into the hands-on process of building a SIEM lab, it’s important to first understand the key terms and concepts that will appear repeatedly throughout this series. Having a clear grasp of these fundamentals will make the rest of the journey much easier to follow and far less frustrating when things inevitably break along the way.</p><h3>1. SIEM (Security Information and Event Management)</h3><p>A SIEM is a system that collects, analyzes, and correlates security data from across an organization’s environment. It doesn’t just store logs — it helps detect suspicious activity, generate alerts, and support incident response.</p><p><strong>In our lab:</strong> Microsoft Sentinel is the SIEM platform we’re using.</p><h3>2. Microsoft Sentinel</h3><p>Microsoft Sentinel is a cloud-native SIEM that runs on Azure. It collects logs from multiple sources, applies analytics, and provides dashboards and alerts for security monitoring. It is highly scalable, integrates with other Microsoft and third-party services, and uses KQL (Kusto Query Language) for searching and analyzing data.</p><h3>3. Log Analytics Workspace</h3><p>A Log Analytics Workspace is a container in Azure where all log data is stored. It is the foundation of Microsoft Sentinel. Without it, Sentinel has no place to ingest, query, or analyze logs.</p><p><strong>Key Point:</strong> Think of it as the central “data lake” for your security logs.</p><h3>4. Microsoft Entra ID (Formerly Azure AD)</h3><p>This is Microsoft’s identity and access management system. It manages users, groups, and devices, and produces logs that tell you who accessed what, when, and from where. Integrating these logs into Sentinel is crucial for monitoring identity-based threats.</p><p><strong>Sign-in Logs</strong><br> These logs record authentication events in Microsoft Entra ID. They show when users or applications attempt to sign in, whether the attempt was successful or failed, the location of the sign-in, the device used, and any Conditional Access policies applied. They are important for detecting suspicious login activity.</p><p><strong>Audit Logs</strong><br> Audit logs track administrative and configuration changes within Microsoft Entra ID. This includes actions such as creating or deleting users, modifying group memberships, updating security settings, or assigning roles. They help monitor who made changes in the environment and when those changes occurred.</p><p><strong>Provisioning Logs</strong><br> Provisioning logs record events related to the creation, update, or removal of user accounts and groups as they are synchronized or provisioned between systems. These logs are useful for monitoring identity lifecycle processes and ensuring that accounts are being created and managed correctly across connected applications and directories.</p><h3>5. Microsoft Defender for Cloud</h3><p>Microsoft Defender for Cloud (formerly Azure Security Center) provides security posture management and threat protection for your Azure resources. Connecting its logs to Sentinel gives visibility into attacks or misconfigurations in your cloud environment.</p><h3>6. Network Security Group (NSG) Logs</h3><p>Network Security Groups act like firewalls for your Azure network, controlling inbound and outbound traffic. Enabling NSG logs allows Sentinel to see network-level events — for example, blocked connections or unusual traffic patterns.</p><h3>7. Kusto Query Language (KQL)</h3><p>KQL is the language Sentinel uses to query logs. Think of it as SQL for security logs, with additional tools for filtering, summarizing, and visualizing security events. Mastering KQL is essential for making a SIEM truly useful.</p><h3>8. Syslog</h3><p>Syslog is a standard protocol used for sending log messages from Unix and Linux systems. Integrating Syslog into Sentinel allows non-Windows systems to contribute security data to your monitoring environment.</p><h3>9. Data Collection Rules (DCRs)</h3><p>Data Collection Rules define what data gets collected from endpoints and where it is sent. They ensure that agents and data sources send the correct logs to the correct Log Analytics Workspace, helping prevent data loss or unnecessary noise.</p><h3>10. Agents</h3><p>Agents are small pieces of software installed on servers or endpoints that send logs to Sentinel. Without agents, many types of system-level data will never reach your SIEM.</p><h3>11. Validation &amp; Monitoring</h3><p>Collecting logs is only half the battle. Validating that data is complete, accurate, and useful is what turns a simple lab setup into a functional security environment. In our lab, we use KQL queries, dashboards, and simulated activity to test and verify that everything is working as expected.</p><h3>Why These Terms Matter</h3><p>As we build the lab, you’ll see these concepts in action. Configurations will fail, connections won’t always work, and limits will appear unexpectedly — but understanding what each component does makes troubleshooting far less painful. This isn’t just about theory; it’s about learning to think like an engineer rather than simply following steps.</p><h3>Wrapping Up</h3><p>With these key concepts in mind, the next step is to begin building the foundation of the lab. In the next part of the series, we’ll start by creating a <strong>Log Analytics Workspace</strong> and enabling <strong>Microsoft Sentinel</strong>, which together form the core of the SIEM environment. Once that foundation is in place, we’ll move on to integrating our data sources and verifying that logs are successfully flowing into the workspace in real time.</p><p><em>If you’d like to connect with me or discuss cloud security projects, you can find me on </em><a href="https://www.linkedin.com/in/florence-nwizugbe"><em>LinkedIn</em></a><em>.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3492aa441df6" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[I Built a Full SIEM Lab on Azure — Here’s Everything That Broke (and What I Learned)| Part 1]]></title>
            <link>https://medium.com/@nwizugbeflorence/i-built-a-full-siem-lab-on-azure-heres-everything-that-broke-and-what-i-learned-part-1-e60ec5088d61?source=rss-aff89d81f065------2</link>
            <guid isPermaLink="false">https://medium.com/p/e60ec5088d61</guid>
            <category><![CDATA[cybersecurity]]></category>
            <category><![CDATA[microsoft-sentinel]]></category>
            <category><![CDATA[cloud-security]]></category>
            <category><![CDATA[siem]]></category>
            <category><![CDATA[azure]]></category>
            <dc:creator><![CDATA[Florence Nwizugbe]]></dc:creator>
            <pubDate>Thu, 05 Mar 2026 14:32:34 GMT</pubDate>
            <atom:updated>2026-03-05T14:32:34.310Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*sAep_CdPQX5cVDQtxnQbAw.png" /></figure><p>It’s been a while since I last wrote on Medium. In that time, a lot changed. I spent more time building than talking, more time testing than posting, and somewhere in between, my understanding of cloud security started to shift from theory to practice.</p><p>Coming back now feels intentional.</p><p>Instead of waiting until I have a perfectly packaged project, I’m documenting this one as it happens.</p><p>I’m currently building a full SIEM lab using Microsoft Sentinel on Azure — not just to “complete a project,” but to understand what it really means to design, connect, and validate a working security monitoring environment. And very quickly, I realized something: things don’t always work the way the documentation suggests they should.</p><p>Configurations fail. Limits appear out of nowhere. Systems that are supposed to connect… don’t. And in those moments, you’re forced to move beyond steps and start thinking like an engineer.</p><p>So this isn’t a final report. It’s a live build.</p><p>I’ll be sharing this project in parts, writing each one after I’ve implemented and tested that stage. The goal is to capture the process as it actually unfolds — from the straightforward setups to the unexpected blockers, and the decisions that sit in between.</p><p>The journey starts with laying the foundation — creating a Log Analytics workspace and enabling Microsoft Sentinel in a way that is structured and intentional, not just functional. From there, I move into integrating identity logs through Microsoft Entra ID, which, despite looking simple, introduces one of the first real challenges in the setup.</p><p>As the lab expands, I bring in additional data sources — connecting Microsoft Defender for Cloud and enabling Network Security Group logs to build visibility beyond identity. At that point, the focus shifts from configuration to data flow: deploying agents, creating data collection rules, and ensuring that endpoints are actually communicating with the SIEM.</p><p>Once logs begin to arrive, I validate everything using KQL queries — because a SIEM is only as useful as the data you can see and interpret.</p><p>The environment doesn’t stop at Windows. I extend the setup by introducing a Linux system, which brings its own layer of complexity — domain integration issues, DNS dependencies, and configuration errors that require careful troubleshooting. From there, I configure syslog, structure log collection, and simulate continuous activity to make the environment feel more real.</p><p>Finally, everything comes together as I connect these different systems back into Microsoft Sentinel, ensuring that logs are not just collected, but properly ingested and usable for monitoring and detection.</p><p>Each part of this series will focus on a specific stage in that journey — written as I go, tested before I share, and grounded in what actually worked (and what didn’t).</p><p>If you’ve ever wondered what it really takes to build a SIEM environment from scratch — or why things break even when you follow the steps ; this series is for you.</p><h3>Wrapping Up &amp; What’s Next</h3><p>Before diving into the technical steps, it helps to define some key terms and concepts that will appear repeatedly throughout the series. From Log Analytics Workspaces and Sentinel itself, to KQL queries, Syslog, and Data Collection Rules — having a shared understanding will make it easier to follow each stage without getting lost in jargon.</p><p>In the next part, we’ll break down these terms, explain their roles in a SIEM environment, and show why they matter in practice — not just in theory. By the end of that section, you’ll have a solid vocabulary that will make the rest of the series much clearer and more actionable.</p><p><em>If you’d like to connect with me or discuss cloud security projects, you can find me on </em><a href="https://www.linkedin.com/in/florence-nwizugbe"><em>LinkedIn</em></a><em>.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e60ec5088d61" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How I Passed the AZ-500 Exam : Practical Tips for Aspiring Azure Security Engineers]]></title>
            <link>https://medium.com/@nwizugbeflorence/how-i-passed-the-az-500-exam-practical-tips-for-aspiring-azure-security-engineers-071bb65a34f8?source=rss-aff89d81f065------2</link>
            <guid isPermaLink="false">https://medium.com/p/071bb65a34f8</guid>
            <category><![CDATA[azure]]></category>
            <category><![CDATA[az-500]]></category>
            <category><![CDATA[security-engineer]]></category>
            <category><![CDATA[cloud-security]]></category>
            <category><![CDATA[cybersecurity]]></category>
            <dc:creator><![CDATA[Florence Nwizugbe]]></dc:creator>
            <pubDate>Wed, 31 Dec 2025 10:42:22 GMT</pubDate>
            <atom:updated>2025-12-31T10:42:22.063Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/521/0*8-Mm57kPuI9a7h0A" /><figcaption>My certificate!</figcaption></figure><p>Preparing for and passing the <strong>AZ-500 Microsoft Azure Security Engineer Associate</strong> exam was one of the most demanding and rewarding steps in my cloud security journey. Unlike foundational certifications, AZ-500 tests your ability to <em>think like a security engineer</em>, not just memorize concepts.</p><p>This post is for anyone preparing for AZ-500 or considering it, and who wants a realistic picture of what the exam demands, how to prepare effectively, and what actually helped me pass.</p><h4>What the AZ-500 Really Tests</h4><p>AZ-500 goes beyond theory. It evaluates how well you can <strong>design, implement, and manage security controls in Azure</strong>. Expect questions that require you to understand <em>why</em> a control is used, not just <em>what</em> it is.</p><p>The exam focuses heavily on:</p><ul><li>Identity and access management</li><li>Network security</li><li>Compute and storage security</li><li>Security operations, monitoring, and incident response</li></ul><p>If you’re coming from a hands-on background, you’ll have a major advantage.</p><h4>Tip 1: Hands-On Labs Are Non-Negotiable</h4><p>One of the biggest mistakes candidates make is relying solely on videos or reading materials. AZ-500 questions are often scenario-based, and without hands-on experience, they can feel overwhelming.</p><p>What helped me:</p><ul><li>Building and securing Azure resources myself</li><li>Configuring Conditional Access, RBAC, NSGs, and Defender services</li><li>Breaking things intentionally and fixing them</li></ul><p>If you can explain <em>why</em> a setting exists and <em>what happens if it’s misconfigured</em>, you’re on the right path.</p><blockquote><em>Pro Tip: If you don’t have access to a full Azure subscription, use the Microsoft Learn sandbox or watch practical videos online. Doing so helps concepts stick far more than passive reading.</em></blockquote><h4>Tip 2: Master Identity and Access Management</h4><p>Identity is at the heart of AZ-500.</p><p>You should be very comfortable with:</p><ul><li>Microsoft Entra ID roles vs Azure RBAC</li><li>Conditional Access policies and use cases</li><li>Privileged Identity Management (PIM)</li><li>MFA enforcement and exclusions</li></ul><p>Many questions test subtle differences — for example, <em>when to use Entra roles instead of subscription-level roles</em>. These details matter.</p><h4>Tip 3: Don’t Underestimate Defender and Security Operations</h4><p>A significant portion of the exam focuses on Microsoft Defender for Cloud, Defender for Endpoint, alerts, and security monitoring.</p><p>Make sure you understand:</p><ul><li>What Defender plans protect which resources</li><li>How alerts are generated and investigated</li><li>How security recommendations translate into real actions</li></ul><p>Think like a SOC analyst, not just an administrator.</p><h4>Tip 4: Learn How Azure Security Components Work Together</h4><p>AZ-500 isn’t about isolated services — it’s about integration.</p><p>You’ll see questions that combine:</p><ul><li>Identity + network security</li><li>Defender alerts + incident response</li><li>Logging + remediation</li></ul><p>Understanding how services interact is far more important than memorizing individual features.</p><h4>Tip 5: Practice Reading Questions Slowly</h4><p>This exam is notorious for <strong>wording traps</strong>.</p><p>My approach:</p><ul><li>Read each question carefully — twice if needed.</li><li>Identify what Microsoft is <em>really</em> asking (least privilege? cost? security best practice?)</li><li>Eliminate options that solve the problem but introduce unnecessary risk.</li></ul><p>Often, more than one answer looks “correct,” but only one aligns with Microsoft’s recommended approach.</p><h4>Tip 6: Leverage Past Questions</h4><ul><li>Go through <strong>past exam questions at least twice</strong></li><li>Verify your answers by always referencing <strong>Microsoft Learn</strong>, and take time to understand <em>why</em> each option is correct</li><li>This repetition reinforces critical concepts and builds confidence.</li></ul><h4>Tip 7: Use Microsoft Learn Strategically</h4><p>Microsoft Learn is excellent — but don’t try to read everything.</p><p>Focus on:</p><ul><li>Exam skills outline topics</li><li>Official docs linked within Learn modules</li><li>Architecture diagrams and examples</li></ul><p>If a concept shows up repeatedly in Learn, it’s likely important for the exam.</p><h4>Resources I Used</h4><p>Here are some of the <strong>key resources</strong> that helped me prepare effectively:</p><ul><li><a href="https://www.youtube.com/playlist?list=PLlVtbbG169nHw9T1L_CiLxC-DTwKu-BZG">John Savill AZ-500 Playlist</a> (YouTube)</li><li><a href="https://www.youtube.com/watch?v=6vISzj-z8k4">John Savill AZ-500 Exam Cram (YouTube)</a></li><li><a href="https://www.examtopics.com/exams/microsoft/az-500/view/">ExamTopics AZ-500 Practice Questions</a></li></ul><blockquote><em>Combining hands-on labs, repeated practice with past questions, and these resources gave me confidence on exam day.</em></blockquote><h4>Exam Day Advice</h4><ul><li>Stay calm — the exam is tough, but manageable</li><li>Don’t rush; time is usually sufficient</li><li>Flag difficult questions and return to them</li><li>Trust your hands-on experience</li></ul><p>If you’ve built and secured real environments, many questions will feel familiar.</p><h3>Final Thoughts</h3><p>Passing AZ-500 isn’t about being perfect, it’s about learning to <strong>think like a security engineer</strong>. The exam rewards practical experience, architectural understanding, and sound security judgment far more than rote memorization.</p><p>Reaching this milestone has been a meaningful moment in my cloud security journey. Preparing for AZ-500 taught me discipline, persistence, and the value of hands-on practice — lessons that extend well beyond the exam itself.</p><p>For anyone stepping into cloud security, this certification is absolutely worth the effort. More importantly, the skills you gain while preparing are directly applicable to real-world roles, helping you approach security challenges with clarity and confidence.</p><p>As the year comes to an end, remember this: every lab you built, every concept you struggled with, and every exam you challenged yourself to attempt has moved you forward. Growth in tech isn’t instant; it’s earned through consistency, curiosity, and showing up even when it’s uncomfortable.</p><p>Step into the new year with confidence. The skills you’re building today are shaping the professional you’re becoming tomorrow and you’re far more capable than you think.</p><p><strong>See you next year.</strong></p><p><em>If you’d like to connect with me or discuss cloud security projects, you can find me on </em><a href="https://www.linkedin.com/in/florence-nwizugbe"><em>LinkedIn</em></a><em>.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=071bb65a34f8" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Building My Hybrid Security Lab (Part 5): Implementing Hybrid Identity with Microsoft Entra Connect]]></title>
            <link>https://medium.com/@nwizugbeflorence/building-my-hybrid-security-lab-part-5-implementing-hybrid-identity-with-microsoft-entra-connect-f9730c36fdfe?source=rss-aff89d81f065------2</link>
            <guid isPermaLink="false">https://medium.com/p/f9730c36fdfe</guid>
            <category><![CDATA[cybersecurity]]></category>
            <category><![CDATA[cloud-security]]></category>
            <category><![CDATA[entra-id]]></category>
            <category><![CDATA[active-directory]]></category>
            <category><![CDATA[cloud-computing]]></category>
            <dc:creator><![CDATA[Florence Nwizugbe]]></dc:creator>
            <pubDate>Tue, 30 Dec 2025 09:04:49 GMT</pubDate>
            <atom:updated>2025-12-30T09:04:49.371Z</atom:updated>
            <content:encoded><![CDATA[<p>By this stage of the lab series, I had established a solid hybrid security foundation: an on-prem Active Directory environment, endpoint protection via Microsoft Defender for Endpoint, identity monitoring with Defender for Identity, and SaaS governance through Defender for Cloud Apps.</p><p>The next logical step was to bridge the gap between on-prem identities and the cloud. That’s where <strong>Microsoft Entra Connect</strong> comes in.</p><p>Microsoft Entra Connect enables organizations to synchronize identities from on-prem Active Directory to Microsoft Entra ID, creating a hybrid identity environment<strong>. </strong>This setup allows users to access cloud services using the same credentials they use on-prem — without managing separate identities.</p><p>In this part of the series, I focused on establishing <strong>hybrid identity synchronization</strong> and enabling <strong>seamless single sign-on (SSO)</strong> to complete the identity layer of my hybrid security lab.</p><h3>Why Entra Connect Matters in a Hybrid Environment</h3><p>Most enterprise environments don’t move to the cloud overnight. Instead, they operate in a hybrid model where on-prem infrastructure and cloud services coexist.</p><p>Microsoft Entra Connect solves a critical problem in this model by:</p><ul><li>Synchronizing users, groups, and credentials between on-prem AD and Entra ID.</li><li>Reducing identity sprawl and administrative overhead.</li><li>Enabling a consistent authentication experience across on-prem and cloud resources.</li><li>Laying the foundation for Conditional Access, MFA, and identity-driven security controls.</li></ul><p>Without Entra Connect, many of the advanced security features in the Defender XDR ecosystem would not work effectively in a hybrid setup.</p><p>At this stage of the lab, I had the infrastructure and security layers in place, but one key piece was still missing: <strong>users</strong>.</p><p>Since I built this environment from scratch, there were no real identities to synchronize yet. Before configuring Microsoft Entra Connect, I needed to create <strong>test users in Active Directory</strong> so I could properly validate synchronization, authentication, and single sign-on behavior.</p><p>In real-world environments, Entra Connect is deployed <em>because</em> users already exist on-prem. Recreating that scenario in the lab makes the setup more realistic and easier to test.</p><h3>Create Test Users in Active Directory</h3><p>To simulate a production-like environment, I created a small set of test users directly in Active Directory.</p><ol><li>Open <strong>Active Directory Users and Computers (ADUC)</strong>.</li><li>Create a dedicated OU, for example: <strong>Hybrid-Lab-Users </strong>(or you can add the users in one of the OUS we created during our initial AD setup)</li></ol><ul><li><strong>Create a User</strong>: Right-click the OU → <strong>New → User</strong>.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/757/1*h1GtVUuyBvzhQEwOWmgPdQ.png" /></figure><ul><li><strong>Enter Details</strong>: Fill in first name, last name, and a username . Click <strong>Next</strong>.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/443/1*AVlVQ_lcNDwtPXYmFif05A.png" /></figure><ul><li><strong>Set Password</strong>: Enter a strong password and check <strong>User must change password at next logon</strong>. Ensure <strong>User cannot change password</strong> and <strong>Password never expires</strong> are set if desired. Click <strong>Next</strong>.</li><li><strong>Finish</strong>: Review settings and click <strong>Finish</strong>. The user now appears in the OU.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/437/1*99-rPfM9h0azkcVa1gjvoQ.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/666/1*Cu4ttruitbicXNJ85F652w.png" /></figure><p>Repeat as needed for additional test accounts.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/1*QAdv1yTBvtpOlwNzjjPHHQ.png" /></figure><p>These users will later be synchronized to Microsoft Entra ID and used to test authentication and SSO.</p><p><strong>Why this matters:</strong></p><ul><li>Entra Connect only synchronizes <strong>existing directory objects</strong>.</li><li>Having users beforehand allows you to immediately confirm sync health.</li><li>It makes troubleshooting easier if something goes wrong.</li></ul><h3>Prepare for Microsoft Entra Connect</h3><p>Before installing Entra Connect, I verified the following:</p><ul><li>Active Directory health (replication and DNS).</li><li>Availability of <strong>Global Administrator or Hybrid identity admin </strong>credentials in Entra ID.</li><li>An account with <strong>Enterprise Admin</strong> permissions on-prem.</li><li>A dedicated <strong>server (</strong>or DC, for lab purposes<strong>)</strong> to install Entra Connect.</li></ul><p><strong>Best practice:</strong><br> In production environments, Entra Connect should be installed on a <strong>dedicated member server</strong>, not a domain controller. For lab scenarios, installing on a DC is acceptable.</p><h3>Step-by-step walkthrough of every configuration stage in the Entra Connect wizard:</h3><h4>Step 1 — Launch Microsoft Entra Connect</h4><p>On the domain-joined Windows Server in the lab, I downloaded the <strong>Microsoft Entra Connect</strong> installer.</p><figure><img alt="" src="https://cdn-images-1.medium.com/proxy/1*1QsIsqko7gAIRM_VgT5fAw.png" /></figure><p>After the download completed, I right-clicked the installer and selected <strong>Run as administrator</strong>, then accepted the license terms to proceed.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*00MtCjA9i-dtgY3ivbDi0w.png" /></figure><h4>Step 2 — Choose Installation Type</h4><p>When prompted to choose an installation type, I selected: <strong>Custom installation</strong></p><p>While Express settings work, Custom installation provides better visibility and control, especially in a learning or lab environment. This allowed me to explicitly configure:</p><ul><li>the authentication method</li><li>seamless single sign-on</li><li>directory and OU filtering</li><li>optional features</li></ul><p>I then clicked <strong>Next</strong>.</p><h4>Step 3 — Install Required Components</h4><p>On the <strong>Install required components</strong> screen, I intentionally left all options unchecked:</p><ul><li>no custom install location</li><li>no existing SQL Server</li><li>no existing service account</li><li>no custom sync groups</li><li>no imported synchronization settings</li></ul><p>This allows Entra Connect to install SQL Express automatically, create and manage its own service accounts, and apply Microsoft-recommended defaults — which is ideal for a lab setup.</p><p>I clicked <strong>Next</strong> to continue.</p><h4>Step 4 — User Sign-In Configuration</h4><p>On the <strong>User sign-in</strong> screen, I selected:</p><ul><li><strong>Password Hash Synchronization</strong></li><li><strong>Enable single sign-on</strong></li></ul><p>With this configuration:</p><ul><li>password hashes are securely synchronized from on-prem AD to Entra ID</li><li>domain-joined devices can authenticate to cloud resources without repeated sign-ins</li></ul><p>I clicked <strong>Next</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*OQ3TLhG5UTDue5ZgBCsJmg.png" /></figure><h4>Step 5 — Connect to Microsoft Entra ID</h4><p>Next, I signed in using a <strong>Hybrid Identity</strong> <strong>Administrator</strong> account in Microsoft Entra ID.</p><p>This account is only used to establish trust between Entra ID and the on-premises directory.</p><p>After successful authentication, I clicked <strong>Next</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*tAnlhd0nQUSDrLswBTfo8A.png" /></figure><h4>Step 6 — Connect to On-Premises Active Directory</h4><p>On the <strong>Connect Directories</strong> screen, the wizard automatically detected my on-prem Active Directory forest.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*tZdqo7Ijqp34y3Tp4Z0JZg.png" /></figure><p>I provided <strong>Domain Admin credentials</strong>, which are required to:</p><ul><li>read directory objects</li><li>create the <strong>AZUREADSSOACC</strong> computer account for seamless SSO</li></ul><p>I then clicked <strong>Next</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_tflDXM0uILrJPWSiG5wtA.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/673/1*T-yTLEGI2ZOH4o1q36bm4Q.png" /></figure><p>Directory added.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ibOYtKRnJAW6d5VYGFYmHA.png" /></figure><h4>Step 7 — Microsoft Entra Sign-In Configuration</h4><p>At this stage, I verified that:</p><ul><li>the on-prem domain was added and verified in Entra ID</li><li>the correct <strong>UPN suffix</strong> was selected</li></ul><p>This step is critical. Users must be able to sign in using a routable domain (for example, <a href="/service/mailto:user@company.com"><em>user@company.com</em></a> instead of <em>user@company.local</em>).</p><p>Once confirmed, I clicked <strong>Next</strong>.</p><h4>Step 8 — Domain and OU Filtering</h4><p>On the <strong>Domain and OU Filtering</strong> screen, I selected <strong>Sync all domains and OUs</strong>.</p><p>For lab environments, this ensures:</p><ul><li>all users, groups, and computers are synchronized</li><li>no objects are accidentally excluded during testing</li></ul><p>I clicked <strong>Next</strong>.</p><h4>Step 9— Optional Features</h4><p>On the <strong>Optional features</strong> screen, I enabled the features in the screenshot below.</p><p>I clicked <strong>Next</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/proxy/1*ga1hGRBFEwTcURbfR2oCJg.png" /></figure><h4>Step 10 — Seamless Single Sign-On Configuration</h4><p>To complete seamless SSO, I entered <strong>Domain Administrator credentials</strong> again.</p><p>Entra Connect uses these credentials to:</p><ul><li>create the <strong>AZUREADSSOACC</strong> computer account</li><li>configure Kerberos delegation</li></ul><p>I clicked <strong>Next</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*RaPPMr3lYYuNZkAaZ-ZU6A.png" /></figure><h4>Step 11 — Ready to Configure</h4><p>On the final configuration screen, I checked:</p><p><strong>Start the synchronization process when configuration completes</strong></p><p>This ensures that the initial synchronization begins immediately after setup.</p><p>I clicked <strong>Configure</strong>.</p><h4>Step 12— Configuration Completion</h4><p>Once configuration completed, I clicked <strong>Exit</strong>.</p><p>The synchronization service began running in the background, and the initial sync took a few minutes to complete.</p><h4><strong>Manually Triggering Microsoft Entra Connect Synchronization</strong></h4><p>After completing the Microsoft Entra Connect configuration, I manually triggered synchronization using PowerShell to immediately validate that directory synchronization was working as expected.</p><p>By default, Microsoft Entra Connect runs a scheduled sync every <strong>30 minutes</strong>. However, during testing or lab setups, it’s best practice to manually start a sync to avoid waiting for the next scheduled cycle.</p><h4>PowerShell Commands Used</h4><pre>Start-ADSyncSyncCycle-PolicyType Initial</pre><p>This command triggers a <strong>full (initial) synchronization</strong>. It performs the following actions:</p><ul><li>Reads all synchronized objects from on-premises Active Directory</li><li>Applies synchronization rules</li><li>Exports all eligible objects to Microsoft Entra ID</li></ul><p>This is typically used:</p><ul><li>Immediately after installing Microsoft Entra Connect</li><li>When validating a new configuration</li><li>After major sync rule or filtering changes</li></ul><pre>Start-ADSyncSyncCycle-PolicyType Delta</pre><p>This command triggers a <strong>delta (incremental) synchronization</strong>. It only syncs:</p><ul><li>Objects that have changed since the last sync cycle</li><li>Attribute updates (for example, password changes, group membership updates)</li></ul><p>Delta syncs are faster and are the default sync type used by Entra Connect after the initial sync.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/972/1*KOLyNXbmHqHH19CPznHseQ.png" /></figure><h3>Step 15 — Validation and Testing</h3><p>To confirm everything was working:</p><p><strong>Synchronization Service Manager</strong></p><ul><li>verified successful import and sync cycles</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/proxy/1*W9BcOoqN_jZmmdfrEq2j6g.png" /></figure><p><strong>Microsoft Entra Admin Center</strong></p><ul><li>confirmed users appeared under <strong>Users</strong></li><li>reviewed sign-in logs for synced accounts</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_UyfyBvynP9J5OfbH6zeNg.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*THW_lNFK2GJCk4GczfM9-g.png" /></figure><p><strong>Seamless SSO</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*J3TTdD2IYX7Q6IQUjjJKPQ.png" /></figure><h3>Key Takeaways</h3><p>Deploying Microsoft Entra Connect completes the identity bridge in a hybrid environment, allowing on-premises users to securely access cloud resources with a consistent authentication experience. Password Hash Synchronization provides a simple and secure hybrid authentication model, while Seamless Single Sign-On significantly improves the user experience on domain-joined devices by removing repeated sign-in prompts. Using the default Entra Connect settings is sufficient for labs and many production scenarios, but ensuring proper domain verification and correct UPN configuration is critical for success. With this final piece in place, the lab now mirrors a real-world hybrid enterprise environment, fully integrating on-premises systems, cloud identities, and modern security controls into a unified, end-to-end security architecture.</p><p><em>If you’d like to connect with me or discuss cloud security projects, you can find me on </em><a href="https://www.linkedin.com/in/florence-nwizugbe"><em>LinkedIn</em></a><em>.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f9730c36fdfe" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>