<?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 Farrel Aditya on Medium]]></title>
        <description><![CDATA[Stories by Farrel Aditya on Medium]]></description>
        <link>https://medium.com/@farreladityah?source=rss-4e1521e096db------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*V74F3Iy7ejDrD07OBa0pPg.jpeg</url>
            <title>Stories by Farrel Aditya on Medium</title>
            <link>https://medium.com/@farreladityah?source=rss-4e1521e096db------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Fri, 20 Nov 2026 00:12:06 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@farreladityah/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[Setting up Multi Availability Zone on OpenStack]]></title>
            <link>https://farreladityah.medium.com/setting-up-multi-availability-zone-on-openstack-0e57a2dc4e0d?source=rss-4e1521e096db------2</link>
            <guid isPermaLink="false">https://medium.com/p/0e57a2dc4e0d</guid>
            <category><![CDATA[high-availability]]></category>
            <category><![CDATA[openstack]]></category>
            <category><![CDATA[fault-tolerance]]></category>
            <category><![CDATA[cloud-computing]]></category>
            <dc:creator><![CDATA[Farrel Aditya]]></dc:creator>
            <pubDate>Wed, 24 Jun 2026 10:12:03 GMT</pubDate>
            <atom:updated>2026-06-24T10:12:03.725Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*TWTT5AQGAS3DLVvR" /></figure><h3>Introduction</h3><p>When running workloads on OpenStack, one of the most critical aspects of cloud infrastructure design in <strong>fault tolerant</strong>, ensuring your instance remain accessible even when part of your infrastructure goes down.</p><p><strong>Availability Zones (AZ)</strong> is one of the answers for this problem. By distributing compute resources accross logicaly isolated zones, you give your users and application the ability to survive partial infrastructure failures without full downtime.</p><p>In this article, I’ll walk you through the complete process of setting up Multi Availability Zone on a nested OpenStack cluster with the following architecture:</p><ul><li><strong>3 Controller Nodes: controller-01, controller-02, controller-03</strong></li><li><strong>3 Compute Nodes: compute-01, compute-02, compute-03</strong></li></ul><h3>Understanding The Building Blocks</h3><p>Before dive in, let’s clarify two concepts that are often confused.</p><h4>Host Aggregate</h4><p>A <strong>Host Aggregate</strong> is a logical grouping of compute nodes (hypervisors) defined by the admin. It’s the internal mechanism that tells Nova Scheduler where to place instances. Users never see aggregates directly, they’re an admin-only construct.</p><p>Think of it like a warehouse with labeled racks. The admin assigns each rack a label (metadata), and the scheduler picks the right rack based on what’s needed.</p><p><strong>Availability Zone</strong></p><p>An <strong>Availability Zone</strong> is the user facing representation of a Host Aggregate. When launching an instance, user can explicitly choose which AZ to deploy to. Each Host Aggregate maps to exactly one AZ, but one AZ can span multiple aggregates.</p><pre>Host Aggregate<br>├── Availability Zone (1 per aggregate)<br>├── Compute Node 1<br>├── Compute Node 2<br>└── Metadata/Properties</pre><p><strong>Why Multi-AZ Matters?</strong></p><p>Without AZ separation, all three compute nodes are treated equally by the scheduler. A user deploying theree instance might end up with all of them on a single node, creating a single point of failure.</p><pre>Without Multi-AZ:                 With Multi-AZ:<br>app-server-1 ──┐                  app-server-1 → az-primary<br>app-server-2 ──┼→ all on          app-server-2 → az-primary<br>app-server-3 ──┘   compute-01     app-server-3 → az-secondary</pre><p><strong>Architecture Design</strong></p><p>For a 3 node compute setup, the recommended split is:</p><pre>AZ: az-primary    → compute-01, compute-02<br>AZ: az-secondary  → compute-03</pre><p>This give you redundancy within az-primary (2 nodes) and an isolated zone in az-secondary for secondary workloads or failover targets.</p><h3>Prerequisites</h3><p>Before starting, verify all nodes are healty:</p><pre># Check all compute services are up<br>openstack compute service list<br><br># Verify all hypervisors are reachable<br>openstack hypervisor list<br><br># Confirm no existing aggregates<br>openstack aggregate list</pre><p><strong>STEP 1 — Create Host Aggregates with AZ Mapping</strong></p><pre># Create aggregate for AZ primary<br>openstack aggregate create --zone az-primary aggregate-az1<br><br># Create aggregate for AZ secondary<br>openstack aggregate create --zone az-secondary aggregate-az2</pre><p><strong>STEP 2 — Assign Compute Nodes to Aggregates</strong></p><pre># Add compute-01 and compute-02 to az-primary<br>openstack aggregate add host aggregate-az1 compute-01<br>openstack aggregate add host aggregate-az1 compute-02<br><br># Add compute-03 to az-secondary (ensure no instances are running on compute-03 first)<br>openstack server list --all-projects --host compute-03<br><br>openstack aggregate add host aggregate-az2 compute-03</pre><p><strong>STEP 3 — Verify AZ Configuration</strong></p><pre># Check AZ list<br>openstack availability zone list --long<br><br># Verify hosts in each aggregate<br>openstack aggregate show aggregate-az1<br>openstack aggregate show aggregate-az2</pre><p><strong>STEP 4 — Configure Cinder for AZ-Aware Storage</strong></p><p>By default, Cinder volumes are created in the nova AZ. This causes a mismatch when deploying instances to custom AZs. To fix this, assign each cinder-volume service to the correct AZ.</p><p>Edit /etc/cinder/cinder.conf on each compute node.</p><pre># On compute-01 and compute-02<br>[DEFAULT]<br>storage_availability_zone = az-primary<br>default_availability_zone = az-primary<br><br># On compute-03<br>[DEFAULT]<br>storage_availability_zone = az-secondary<br>default_availability_zone = az-secondary<br><br># Restart the service on each node after editing<br>sudo systemtctl restart cinder-volume.service</pre><p><strong>Creating Volumes Types per AZ</strong></p><p>To make storage placement fully automatic, create dedicated volume types:</p><pre># Volume type for az-primary<br>openstack volume type create az-primary<br>openstack volume type set az-primary --property RESKEY:availability_zones=az-primary<br><br># Volume type for az-secondary<br>openstack volume type create az-secondary<br>openstack volume type set az-secondary --property RESKEY:availability_zones=az-secondary</pre><p>With this in place, Cinder Scheduler will automatically route volume creation to the correct backend based on the AZ.</p><p><strong>STEP 5 — Test the Setup</strong></p><p>Deploy instance explicitly to each AZ and verify placement:</p><pre># Deploy to az-primary (should land on compute-01 or compute-02)<br>openstack server create \<br>  --flavor m1.small \<br>  --image ubuntu-24.04 \<br>  --network &lt;network-id&gt; \<br>  --availability-zone az-primary \<br>  test-vm-az-primary<br><br># Deploy to az-secondary (should land on compute-03)<br>openstack server create \<br>  --flavor m1.small \<br>  --image ubuntu-24.04 \<br>  --network &lt;network-id&gt; \<br>  --availability-zone az-secondary \<br>  test-vm-az-secondary</pre><p>Verify placement:</p><pre>openstack server show test-vm-az-primary | grep -E &quot;availability_zone|host&quot;<br>openstack server show test-vm-az-secondary | grep -E &quot;availability_zone|host&quot;</pre><p><strong>High Availability Pattern</strong></p><p>For applications requiring high availability, deploy across both AZs:</p><pre># Primary instance in az-primary<br>openstack server create \<br>  --availability-zone az-primary \<br>  --flavor m1.small --image ubuntu-24.04 --network &lt;net-id&gt; \<br>  app-server-primary<br><br># Standby instance in az-secondary<br>openstack server create \<br>  --availability-zone az-secondary \<br>  --flavor m1.small --image ubuntu-24.04 --network &lt;net-id&gt; \<br>  app-server-standby</pre><p>If az-primary goes down, app-server-standby in az-secondary continues serving traffic. This is the core value proposition of Multi-AZ.</p><p><strong>Maintenance Without Downtime</strong></p><p>One of the practical benefits of Multi-AZ is rolling maintenance. To take az-primary offline for maintenance:</p><pre># Disable compute nodes in az-primary<br>openstack compute service set --disable compute-01 nova-compute<br>openstack compute service set --disable compute-02 nova-compute<br><br># az-secondary continues accepting new instance request, perform maintenance on compute-01 and compute-02...<br><br># Re-enable after maintenance<br>openstack compute service set --enable compute-01 nova-compute<br>openstack compute service set --enable compute-02 nova-compute</pre><h3>Summary</h3><p>Multi-AZ is not just configuration exercise, it’s a fundamental shift in how you design for resillience. With this setup, your OpenStack cluster can tolerate partial failures, support rolling maintenance, and give users explicit control over workload placement.</p><p>Hopefully this article can provide benefits to readers. If you guys find mistakes in this article, please contact me on <a href="http://farreladitya375@gmail.com/">Email </a>so that I can fix it immediately.</p><p>Follow me on <a href="https://www.linkedin.com/in/farreladityah/">LinkedIn</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=0e57a2dc4e0d" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Setup Grafana and Prometheus on Ubuntu 24.04 LTS]]></title>
            <link>https://farreladityah.medium.com/setup-grafana-and-prometheus-on-ubuntu-24-04-lts-dd30904a1aa3?source=rss-4e1521e096db------2</link>
            <guid isPermaLink="false">https://medium.com/p/dd30904a1aa3</guid>
            <category><![CDATA[monitoring]]></category>
            <category><![CDATA[grafana]]></category>
            <category><![CDATA[prometheus]]></category>
            <dc:creator><![CDATA[Farrel Aditya]]></dc:creator>
            <pubDate>Wed, 24 Jun 2026 09:56:35 GMT</pubDate>
            <atom:updated>2026-06-24T09:56:35.809Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/600/0*5SlI3ByiXYIGGm9C.png" /></figure><h3>Introduction</h3><p><strong>Prometheus</strong> is a time-series database that continously scrapes metrics from your systems. <strong>Grafana </strong>sits on top of it, turning raw numbers into dashboards you can actually read. Together, they form the backbone of modern infrastructure monitoring.</p><p>In this article, we’ll go from zero to a fully running monitoring stack on Ubuntu. By the end, you’ll have:</p><ul><li>Prometheus collecting system metrics every 15 seconds</li><li>Node Exporter exposing CPU, RAM, disk, and network data</li><li>Grafana rendering it all into beautiful dashboard</li></ul><h4>STEP 1 — Install Prometheus</h4><pre># Downloading Prometheus (Check up for new versions)<br>wget https://github.com/prometheus/prometheus/releases/download/v2.52.0/prometheus-2.52.0.linux-amd64.tar.gz<br>tar xvf prometheus-2.52.0.linux-amd64.tar.gz<br>sudo mv prometheus-2.52.0.linux-amd64 /opt/prometheus<br><br># Creating dedicated user<br>sudo useradd --no-create-home --shell /bin/false prometheus<br><br># Creating Prometheus directory<br>sudo mkdir /etc/prometheus /var/lib/prometheus<br>sudo cp /opt/prometheus/{prometheus,promtool} /usr/local/bin/<br>sudo cp -r /opt/prometheus/{consoles,console_libraries} /etc/prometheus/<br>sudo chown -R prometheus:prometheus /etc/prometheus /var/lib/prometheus</pre><h4>STEP 2 — Configure Prometheus</h4><pre># This is a minimum configuration<br>global:<br>  scrape_interval: 15s<br>  evaluation_interval: 15s<br><br>scrape_configs:<br>  - job_name: &quot;prometheus&quot;<br>    static_configs:<br>      - targets: [&quot;localhost:9090&quot;]<br><br>  - job_name: &quot;node_exporter&quot;<br>    static_configs:<br>      - targets: [&quot;localhost:9100&quot;]</pre><h4>STEP 3 — Configure systemd service for Prometheus</h4><pre># /etc/systemd/system/prometheus.service<br><br>[Unit]<br>Description=Prometheus<br>After=network.target<br><br>[Service]<br>User=prometheus<br>Group=prometheus<br>ExecStart=/usr/local/bin/prometheus \<br>  --config.file=/etc/prometheus/prometheus.yml \<br>  --storage.tsdb.path=/var/lib/prometheus \<br>  --web.console.templates=/etc/prometheus/consoles \<br>  --web.console.libraries=/etc/prometheus/console_libraries \<br>  --storage.tsdb.retention.time=15d<br><br>[Install]<br>WantedBy=multi-user.target</pre><pre>sudo systemctl daemon-reload<br>sudo systemctl enable --now prometheus</pre><h4>STEP 4 — Installing Node Exporter for metrics server</h4><p>Node Exporter is the standard way to expose Linux host metrics, CPU usage, memory, disk I/O, network, and more.</p><pre>wget https://github.com/prometheus/node_exporter/releases/download/v1.8.0/node_exporter-1.8.0.linux-amd64.tar.gz<br>tar xvf node_exporter-1.8.0.linux-amd64.tar.gz<br>sudo cp node_exporter-1.8.0.linux-amd64/node_exporter /usr/local/bin/<br><br>sudo useradd --no-create-home --shell /bin/false node_exporter</pre><h4>STEP 5 — Configure systemd service for Node Exporter</h4><pre>[Unit]<br>Description=Node Exporter<br>After=network.target<br><br>[Service]<br>User=node_exporter<br>Group=node_exporter<br>ExecStart=/usr/local/bin/node_exporter<br><br>[Install]<br>WantedBy=multi-user.target</pre><pre>sudo systemctl enable --now node_exporter</pre><h4>STEP 6 — Installing Grafana</h4><p>Grafana has an official APT repository for Ubuntu/Debian systems, which makes installation and future upgrades clean.</p><pre>sudo apt-get install -y apt-transport-https software-properties-common<br>wget -q -O - https://apt.grafana.com/gpg.key | sudo apt-key add -<br>echo &quot;deb https://apt.grafana.com stable main&quot; | sudo tee /etc/apt/sources.list.d/grafana.list<br>sudo apt-get update &amp;&amp; sudo apt-get install grafana<br><br>sudo systemctl enable --now grafana-server</pre><p>Grafana runs an internal database migration on first boot. This can take several minutes. Monitor it with:</p><pre>sudo journalctl -u grafana-server -f</pre><p>Then verify the port is open:</p><pre>sudo ss -lntp | grep :3000</pre><p>Access Grafana on your web browser, open <a href="http://localhost:3000/">http://localhost:3000</a> here are the default credentials:</p><ul><li>Username: admin</li><li>Password: admin</li></ul><p>You’ll be prompted to change the password on first login.</p><h4>STEP 7 — Connect Prometheus as a Data Source</h4><ol><li>In Grafana, go to <strong>Connections</strong> then <strong>Data Sources</strong></li><li>Click <strong>Add data Source</strong></li><li>Select <strong>Prometheus</strong></li><li>Set the URL to: <a href="http://localhost:9090/"><strong>http://localhost:9090</strong></a></li><li>Scroll down and click <strong>Save &amp; Test</strong></li></ol><p>If everything is working, you’ll see a green banner: “<strong>Successfully queried the Prometheus API.</strong>”</p><h4>STEP 8 — Import a Pre-built Dashboard</h4><p>You don’t need to build dashboards from scratch. The Grafana community has thousands of ready made dashboards you can import in seconds.</p><ol><li>Go to <strong>Dashboards</strong> then<strong> Import</strong></li><li>Enter a dashboard ID from the list below for example</li><li>Select your Prometheus data source</li><li>Click <strong>Import</strong></li></ol><pre>Dashboard                          ID<br>Node Exporter Full                1860<br>Prometheus Stats                  3662<br>MySQL Overview                    7362<br>PostgreSQL Database               9628</pre><h3>Conclusion</h3><p>In about 15 minutes, you’ve gone from a bare server to a full obeservability stack. Prometheus is quietly pulling metrics every 15 seconds, Node Exporter is translating Linux internals into something Prometheus can read, and Grafana is ready to run all of that data into dashboards that actually make sense.</p><p>Hopefully this article can provide benefits to readers. If you guys find mistakes in this article, please contact me on <a href="http://farreladitya375@gmail.com/">Email </a>so that I can fix it immediately.</p><p>Follow me on <a href="https://www.linkedin.com/in/farreladityah/">LinkedIn</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=dd30904a1aa3" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[End-to-end Observability on OpenStack using Self-Hosted SigNoz]]></title>
            <link>https://farreladityah.medium.com/end-to-end-observability-on-openstack-using-self-hosted-signoz-01e41997f38c?source=rss-4e1521e096db------2</link>
            <guid isPermaLink="false">https://medium.com/p/01e41997f38c</guid>
            <category><![CDATA[observability]]></category>
            <category><![CDATA[cloud-computing]]></category>
            <category><![CDATA[signoz]]></category>
            <category><![CDATA[openstack]]></category>
            <dc:creator><![CDATA[Farrel Aditya]]></dc:creator>
            <pubDate>Tue, 24 Feb 2026 06:38:23 GMT</pubDate>
            <atom:updated>2026-02-24T06:38:23.870Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/832/1*WaPJGvBEb0UCJjlx4qyY3A.png" /><figcaption>Copyright Farrel Aditya Hermawan — @2026</figcaption></figure><p>SigNoz is an open-source observability platform designed to help engineers monitor and understand modern distributed systems. Built natively on OpenTelemetry, SigNoz provides unified visibility into metrics, traces, and logs through a single interface, making it easier to troubleshoot performance issues and analyze system behaviour end-to-end.</p><h3>SigNoz Architecture Overview:</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*H9qwA1QbtmuGblXE9v-aWw.png" /><figcaption><a href="https://signoz.io/blog/launching-signoz-single-binary/">https://signoz.io/blog/launching-signoz-single-binary/</a></figcaption></figure><p>At a high level, SigNoz is built on a modular and scalable architecture that follows the OpenTelemetry data pipeline. Each component has a clear responsibility, making SigNoz suitable for self-hosted and distributed environments.</p><ol><li><strong>Telemetry Ingestion (OpenTelemetry Collector):</strong> All metrics, traces, and logs, are received via the OpenTelemetry Collector (otelcol-contrib), which handles receiving, processing, and exporting telemetry data to SigNoz backend services. This component can run independently such as Docker container.</li><li><strong>Query &amp; API Layer:</strong> Backend services expose APIs by the UI to query telemetry data, enabling service-level metrics, distributed tracing, dashboards, and alerts.</li><li><strong>Storage Backend (ClickHouse):</strong> SigNoz uses ClickHouse as its primary database for unique metrics, traces, and logs. In production, ClickHouse usually deployed as a cluster coordinated by Zookeeper for scalability and high availability.</li><li><strong>Visualization Layer: </strong>The SigNoz UI provides a single interface for metrics dashboards, trace exploration, and log analysis, with full correlation across all telemetry signals.</li></ol><h3>Architecture Used in This Setup:</h3><p>This setup runs on a OpenStack environment consisting 3 controller nodes, and 3 compute nodes, with all SigNoz components centralized on the controller layer.</p><p>The controller nodes host entire observability stack, consisting:</p><ul><li>SigNoz backend services deployed in binary mode.</li><li>ClickHouse cluster as the primary storage backend.</li><li>Zookeeper for ClickHouse cluster coordination.</li><li>OpenTelemetry Collector (otelcol-contrib) running as a Docker container.</li></ul><p>The compute nodes are solely to running workloads and OpenStack compute services.</p><p>In this architecture, telemetry data from OpenStack services and infrastructure is collected at the controller level and processed centrally, reducing operational complexity and keeping the observability stack tightly integrated with the control plane.</p><h3>Preparing ClickHouse:</h3><h4>STEP 1 — Installing ClickHouse</h4><p>To install ClickHouse run the following commands:</p><pre># Install prerequisite packages<br>sudo apt-get install -y apt-transport-https ca-certificates curl gnupg<br><br># Download the ClickHouse GPG key and store it in the keyring<br>curl -fsSL &#39;https://packages.clickhouse.com/rpm/lts/repodata/repomd.xml.key&#39; | sudo gpg --dearmor -o /usr/share/keyrings/clickhouse-keyring.gpg<br><br># Get the system architecture<br>ARCH=$(dpkg --print-architecture)<br><br># Add the ClickHouse repository to apt sources<br>echo &quot;deb [signed-by=/usr/share/keyrings/clickhouse-keyring.gpg arch=${ARCH}] https://packages.clickhouse.com/deb stable main&quot; | sudo tee /etc/apt/sources.list.d/clickhouse.list<br><br># Update apt package lists<br>sudo apt-get update</pre><p>Notes: You can replace stable with lts to use different release kinds based on your need.</p><h4>STEP 2 — Install ClickHouse Server and Client</h4><pre>sudo apt-get install -y clickhouse-server clickhouse-client</pre><h4>STEP 3 — Start ClickHouse</h4><p>To start the ClickHouse server, run:</p><pre>sudo service clickhouse-server start</pre><p>To start ClickHouse client, run:</p><pre>clickhouse-client<br><br># If you set up a password for your server, then you will need to run:<br>clickhouse-client --password</pre><h3>Preparing Zookeeper:</h3><h4>STEP 1 — Install Java:</h4><p>Install Java as a prerequisite for ZooKeeper:</p><pre>sudo apt update<br>sudo apt upgrade -y<br>sudo apt install default-jdk -y</pre><h4>STEP 2 — Get and extract the latest release of Zookeeper</h4><p>This step retrieves and unpacks ZooKeeper binaries for installation:</p><pre>curl -L https://dlcdn.apache.org/zookeeper/zookeeper-3.8.5/apache-zookeeper-3.8.5-bin.tar.gz -o zookeeper.tar.gz<br>tar -xzf zookeeper.tar.gz</pre><h4>STEP 3 — Create the necessary Zookeeper directories and copy the extracted files to /opt/zookeeper</h4><p>This step prepares dedicated directories for ZooKeeper binaries, data, and logs:</p><pre>sudo mkdir -p /opt/zookeeper<br>sudo mkdir -p /var/lib/zookeeper<br>sudo mkdir -p /var/log/zookeeper<br>sudo cp -r apache-zookeeper-3.8.5-bin/* /opt/zookeeper</pre><h4>STEP 4 — Create the Zookeeper config file</h4><p>This step configures the ZooKeeper cluster settings:</p><pre>sudo bash -c &#39;cat &lt;&lt;EOF &gt; /opt/zookeeper/conf/zoo.cfg<br>tickTime=2000<br>dataDir=/var/lib/zookeeper<br>clientPort=2181<br>admin.serverPort=3181<br>initLimit=10<br>syncLimit=5<br>server.1=controller-01:2888:3888<br>server.2=controller-02:2888:3888<br>server.3=controller-03:2888:3888<br>4lw.commands.whitelist=ruok,stat,mntr,conf,srvr<br>autopurge.snapRetainCount=3<br>autopurge.purgeInterval=24<br>EOF&#39;</pre><h4>STEP 5 — Create the Zookeeper systemd environment file</h4><p>This step sets ZooKeeper environment variables for logging:</p><pre>sudo bash -c &#39;cat &lt;&lt;EOF &gt; /opt/zookeeper/conf/zoo.env<br>ZOO_LOG_DIR=/var/log/zookeeper<br>EOF&#39;</pre><h4>STEP 6 — Create a user and group to run Zookeeper and give it ownership of the Zookeeper directories</h4><p>This step creates a dedicated system user for ZooKeeper and assigns proper ownership to ensure secure and isolated service execution:</p><pre>sudo getent passwd zookeeper &gt;/dev/null || sudo useradd --system --home /opt/zookeeper --no-create-home --user-group --shell /sbin/nologin zookeeper<br>sudo chown -R zookeeper:zookeeper /opt/zookeeper<br>sudo chown -R zookeeper:zookeeper /var/lib/zookeeper<br>sudo chown -R zookeeper:zookeeper /var/log/zookeeper</pre><h4>STEP 7 — Create the Zookeeper systemd service</h4><p>This step enables ZooKeeper to run and be managed as a systemd service:</p><pre>sudo bash -c &#39;cat &lt;&lt;EOF &gt; /etc/systemd/system/zookeeper.service<br>[Unit]<br>Description=Zookeeper<br>Documentation=http://zookeeper.apache.org<br><br>[Service]<br>EnvironmentFile=/opt/zookeeper/conf/zoo.env<br>Type=forking<br>WorkingDirectory=/opt/zookeeper<br>User=zookeeper<br>Group=zookeeper<br>ExecStart=/opt/zookeeper/bin/zkServer.sh start /opt/zookeeper/conf/zoo.cfg<br>ExecStop=/opt/zookeeper/bin/zkServer.sh stop /opt/zookeeper/conf/zoo.cfg<br>ExecReload=/opt/zookeeper/bin/zkServer.sh restart /opt/zookeeper/conf/zoo.cfg<br>TimeoutSec=30<br>Restart=on-failure<br><br>[Install]<br>WantedBy=multi-user.target<br>EOF&#39;</pre><h4>STEP 8 — Reload the systemd daemon and start the Zookeeper service</h4><pre>sudo systemctl daemon-reload<br>sudo systemctl start zookeeper.service</pre><h4>STEP 9 — Configure ClickHouse to use Zookeeper and start the ClickHouse service</h4><p>This step configures ClickHouse to use ZooKeeper for cluster coordination and defines a multi-replica cluster across the controller nodes:</p><pre>sudo nano /etc/clickhouse-server/config.d/cluster.xml<br><br>&lt;clickhouse&gt;<br>    &lt;remote_servers replace=&quot;true&quot;&gt;<br>        &lt;cluster&gt;<br>            &lt;shard&gt;<br>                &lt;replica&gt;<br>                    &lt;host&gt;controller-01&lt;/host&gt;<br>                    &lt;port&gt;9000&lt;/port&gt;<br>                    &lt;user&gt;signoz&lt;/user&gt;<br>                    &lt;password&gt;admin&lt;/password&gt;<br>                &lt;/replica&gt;<br>                &lt;replica&gt;<br>                    &lt;host&gt;controller-02&lt;/host&gt;<br>                    &lt;port&gt;9000&lt;/port&gt;<br>                    &lt;user&gt;signoz&lt;/user&gt;<br>                    &lt;password&gt;admin&lt;/password&gt;<br>                &lt;/replica&gt;<br>                &lt;replica&gt;<br>                    &lt;host&gt;controller-03&lt;/host&gt;<br>                    &lt;port&gt;9000&lt;/port&gt;<br>                    &lt;user&gt;signoz&lt;/user&gt;<br>                    &lt;password&gt;admin&lt;/password&gt;<br>                &lt;/replica&gt;<br>            &lt;/shard&gt;<br>        &lt;/cluster&gt;<br>    &lt;/remote_servers&gt;<br><br>    &lt;zookeeper replace=&quot;true&quot;&gt;<br>        &lt;node&gt;<br>            &lt;host&gt;controller-01&lt;/host&gt;<br>            &lt;port&gt;2181&lt;/port&gt;<br>        &lt;/node&gt;<br>        &lt;node&gt;<br>            &lt;host&gt;controller-02&lt;/host&gt;<br>            &lt;port&gt;2181&lt;/port&gt;<br>        &lt;/node&gt;<br>        &lt;node&gt;<br>            &lt;host&gt;controller-03&lt;/host&gt;<br>            &lt;port&gt;2181&lt;/port&gt;<br>        &lt;/node&gt;<br>    &lt;/zookeeper&gt;<br><br>    &lt;macros replace=&quot;true&quot;&gt;<br>        &lt;cluster&gt;cluster&lt;/cluster&gt;<br>        &lt;shard&gt;1&lt;/shard&gt;<br>        &lt;replica&gt;controller-01&lt;/replica&gt; # In this section, set the replica according to the configured server<br>    &lt;/macros&gt;<br>&lt;/clickhouse&gt;<br><br>sudo chown clickhouse:clickhouse /etc/clickhouse-server/config.d/cluster.xml<br>sudo systemctl start clickhouse-server.service</pre><h3>Verify the installation:</h3><p>Check the status of the ClickHouse and Zookeeper services:</p><pre>sudo systemctl status clickhouse-server.service<br>sudo systemctl status zookeeper.service</pre><h3>Run ClickHouse migrations:</h3><p>Get and extract the latest release of SigNoz Schema Migrator:</p><pre>curl -L https://github.com/SigNoz/signoz-otel-collector/releases/latest/download/signoz-schema-migrator_linux_$(uname -m | sed &#39;s/x86_64/amd64/g&#39; | sed &#39;s/aarch64/arm64/g&#39;).tar.gz -o signoz-schema-migrator.tar.gz<br>tar -xzf signoz-schema-migrator.tar.gz</pre><p>Run the migrations:</p><pre>./signoz-schema-migrator_linux_amd64/bin/signoz-schema-migrator sync   --dsn=&quot;tcp://controller-01:9000?username=default&amp;password=admin&quot;   --replication=true   --up=<br>./signoz-schema-migrator sync   --dsn &quot;clickhouse://signoz:admin@controller-01:9000&quot;   --cluster-name cluster   --replication</pre><h3>Configure Load Balancing:</h3><h4>STEP 1 — Sets up HAProxy listeners to distribute traffic</h4><p>Make sure you already have HAProxy:</p><pre>sudo nano /etc/haproxy/haproxy.cfg<br><br># Load-balances access to the SigNoz Web UI across controller nodes using source-based balancing to maintain session consistency<br>listen signoz_dashboard<br>    bind 172.31.249.100:8088<br>    mode http<br>    option httplog<br>    option tcplog<br>    balance source<br>    timeout connect 30m<br>    timeout server 30m<br>    timeout queue 30m<br>    server controller-01 172.31.80.201:8080 check inter 2s fall 3 rise 2<br>    server controller-02 172.31.80.244:8080 check inter 2s fall 3 rise 2<br>    server controller-03 172.31.80.231:8080 check inter 2s fall 3 rise 2<br><br># Distributes ClickHouse HTTP queries using least connections for efficient query handling<br>listen clickhouse_http<br>    bind 172.31.249.100:8123<br>    mode http<br>    balance leastconn<br>    timeout connect 30m<br>    timeout server 30m<br>    timeout queue 30m<br>    server controller-01 172.31.80.201:8123 check inter 2s fall 3 rise 2<br>    server controller-02 172.31.80.244:8123 check inter 2s fall 3 rise 2<br>    server controller-03 172.31.80.231:8123 check inter 2s fall 3 rise 2<br><br># Provides high availability access to ClickHouse via the native TCP protocol<br>listen clickhouse_tcp<br>    bind 172.31.249.100:9000<br>    mode tcp<br>    timeout connect 30m<br>    timeout server 30m<br>    timeout queue 30m<br>    server controller-01 172.31.80.201:9000 check inter 2s fall 3 rise 2<br>    server controller-02 172.31.80.244:9000 check inter 2s fall 3 rise 2<br>    server controller-03 172.31.80.231:9000 check inter 2s fall 3 rise 2<br><br># Balances OpenTelemetry OTLP gRPC ingestion traffic evenly across controllers<br>listen otel_otlp_grpc<br>    bind 172.31.249.100:4317<br>    balance leastconn<br>    mode tcp<br>    timeout connect 30m<br>    timeout server 30m<br>    timeout queue 30m<br>    server controller-01 172.31.80.201:4317 check inter 2s fall 3 rise 2<br>    server controller-02 172.31.80.244:4317 check inter 2s fall 3 rise 2<br>    server controller-03 172.31.80.231:4317 check inter 2s fall 3 rise 2<br><br># Handles OTLP over HTTP traffic for telemetry exporters<br>listen otel_otlp_http<br>    bind 172.31.249.100:4318<br>    balance leastconn<br>    mode http<br>    timeout connect 30m<br>    timeout server 30m<br>    timeout queue 30m<br>    server controller-01 172.31.80.201:4318 check inter 2s fall 3 rise 2<br>    server controller-02 172.31.80.244:4318 check inter 2s fall 3 rise 2<br>    server controller-03 172.31.80.231:4318 check inter 2s fall 3 rise 2<br><br># Load balances Syslog log ingestion across all controller nodes<br>listen otel_syslog<br>    bind 172.31.249.100:54527<br>    balance leastconn<br>    mode tcp<br>    timeout connect 30m<br>    timeout server 30m<br>    timeout queue 30m<br>    server controller-01 172.31.80.201:54527 check inter 2s fall 3 rise 2<br>    server controller-02 172.31.80.244:54527 check inter 2s fall 3 rise 2<br>    server controller-03 172.31.80.231:54527 check inter 2s fall 3 rise 2</pre><h3>Install SigNoz:</h3><h4>STEP 1 — Get and extract the latest release of SigNoz</h4><pre>curl -L https://github.com/SigNoz/signoz/releases/latest/download/signoz_linux_$(uname -m | sed &#39;s/x86_64/amd64/g&#39; | sed &#39;s/aarch64/arm64/g&#39;).tar.gz -o signoz.tar.gz<br>tar -xzf signoz.tar.gz</pre><h4>STEP 2 — Create the necessary directories and copy the extracted files to /opt/signoz</h4><pre>sudo mkdir -p /opt/signoz<br>sudo mkdir -p /var/lib/signoz<br>sudo cp -r signoz_linux_$(uname -m | sed &#39;s/x86_64/amd64/g&#39; | sed &#39;s/aarch64/arm64/g&#39;)/* /opt/signoz</pre><h4>STEP 3 — Create the SigNoz systemd environment file</h4><pre>sudo bash -c &#39;cat &lt;&lt;EOF &gt; /opt/signoz/conf/systemd.env<br>SIGNOZ_INSTRUMENTATION_LOGS_LEVEL=info<br>INVITE_EMAIL_TEMPLATE=/opt/signoz/templates/invitation_email_template.html<br>SIGNOZ_SQLSTORE_SQLITE_PATH=/var/lib/signoz/signoz.db<br>SIGNOZ_WEB_ENABLED=true<br>SIGNOZ_WEB_DIRECTORY=/opt/signoz/web<br>SIGNOZ_JWT_SECRET=secret<br>SIGNOZ_ALERTMANAGER_PROVIDER=signoz<br>SIGNOZ_TELEMETRYSTORE_PROVIDER=clickhouse<br>SIGNOZ_TELEMETRYSTORE_CLICKHOUSE_DSN=tcp://localhost:9000?password=password<br>DOT_METRICS_ENABLED=true<br>EOF&#39;</pre><h4>STEP 4 — Create a user and group to run SigNoz and give it ownership of the signoz directories</h4><pre>sudo getent passwd signoz &gt;/dev/null || sudo useradd --system --home /opt/signoz --no-create-home --user-group --shell /sbin/nologin signoz<br>sudo chown -R signoz:signoz /var/lib/signoz<br>sudo chown -R signoz:signoz /opt/signoz</pre><h4>STEP 5 — Create the SigNoz systemd service</h4><pre>sudo bash -c &#39;cat &lt;&lt;EOF &gt; /etc/systemd/system/signoz.service<br>[Unit]<br>Description=SigNoz<br>Documentation=https://signoz.io/docs<br>After=clickhouse-server.service<br><br>[Service]<br>User=signoz<br>Group=signoz<br>Type=simple<br>KillMode=mixed<br>Restart=on-failure<br>WorkingDirectory=/opt/signoz<br>EnvironmentFile=/opt/signoz/conf/systemd.env<br>ExecStart=/opt/signoz/bin/signoz server<br><br>[Install]<br>WantedBy=multi-user.target<br>EOF&#39;</pre><h4>STEP 6 — Reload the systemd daemon and start the SigNoz service</h4><pre>sudo systemctl daemon-reload<br>sudo systemctl start signoz.service</pre><h3>Verify the installation:</h3><p>Check the status of the signoz service:</p><pre>sudo systemctl status signoz.service</pre><h3>Preparing SigNoz OpenTelemetry Collector Contrib:</h3><h4>STEP 1 — Get the latest OpenTelemetry Collector Contrib</h4><pre>wget https://github.com/open-telemetry/opentelemetry-collector-releases/releases/download/v0.143.1/otelcol-contrib_0.143.1_linux_amd64.tar.gz</pre><h4>STEP 2 — Create directory for otelcol-contrib and extract</h4><pre>sudo mkdir -p /opt/otelcol-contrib<br>sudo tar -xzf otelcol-contrib_0.143.1_linux_amd64.tar.gz -C /opt/otelcol-contrib</pre><p>This allows the OpenTelemetry Collector to be executed from any terminal location and prepares it for configuration as a system service:</p><pre>sudo cp /opt/otelcol-contrib/otelcol-contrib /usr/local/bin/<br>sudo chmod +x /usr/local/bin/otelcol-contrib</pre><h4>STEP 3 — Configure the OpenTelemetry Collector systemd service</h4><pre>sudo bash -c &#39;cat &lt;&lt;EOF &gt; /etc/systemd/system/otelcol-contrib.service<br><br>[Unit]<br>Description=OpenTelemetry Collector Contrib<br>After=network-online.target<br>Wants=network-online.target<br><br>[Service]<br>Type=simple<br>ExecStart=/usr/local/bin/otelcol-contrib --config /etc/otelcol-contrib/config.yaml<br>Restart=always<br>RestartSec=3<br>LimitNOFILE=65536<br><br>[Install]<br>WantedBy=multi-user.target<br>EOF&#39;</pre><h4>STEP 4 — Reload the systemd daemon and start the OpenTelemetry Collector service</h4><pre>sudo systemctl daemon-reload<br>sudo systemctl start otelcol-contrib.service</pre><h4>STEP 5— Configure the OpenTelemetry Collector</h4><pre># Create the otel-collector directory and get the config file<br>sudo mkdir -p /opt/signoz/otel-collector<br>cd /opt/signoz/otel-collector<br>sudo curl -L   https://raw.githubusercontent.com/SigNoz/signoz/main/deploy/docker/otel-collector-config.yaml   -o config.yaml<br><br># Configure the otel-collector based on your architecture<br>sudo nano /opt/signoz/otel-collector/config.yaml<br><br>connectors:<br>  signozmeter:<br>    metrics_flush_interval: 1h<br>    dimensions:<br>      - name: service.name<br>      - name: deployment.environment<br>      - name: host.name<br>receivers:<br>  otlp:<br>    protocols:<br>      grpc:<br>        endpoint: 172.31.80.244:4317 # Node internal IP<br>      http:<br>        endpoint: 172.31.80.244:4318 # Node internal IP<br><br>  syslog:<br>    tcp:<br>      listen_address: 172.31.80.244:54527 # Node internal IP<br>    protocol: rfc3164<br><br>  hostmetrics:<br>    collection_interval: 60s<br>    scrapers:<br>      cpu: {}<br>      memory: {}<br>      disk: {}<br>      filesystem: {}<br>      network: {}<br>      load: {}<br><br>  prometheus:<br>    config:<br>      global:<br>        scrape_interval: 60s<br>      scrape_configs:<br>        - job_name: otel-collector<br>          static_configs:<br>          - targets:<br>              - localhost:8888<br>            labels:<br>              job_name: otel-collector<br>        - job_name: openstack-exporter<br>          static_configs:<br>          - targets:<br>              - localhost:9180<br>            labels:<br>              job_name: openstack-exporter<br>              service.name: openstack<br>              deployment.environment: staging<br>processors:<br>  batch:<br>    send_batch_size: 10000<br>    send_batch_max_size: 11000<br>    timeout: 10s<br>  batch/meter:<br>    send_batch_max_size: 25000<br>    send_batch_size: 20000<br>    timeout: 1s<br>  resourcedetection:<br>    # Using OTEL_RESOURCE_ATTRIBUTES envvar, env detector adds custom labels.<br>    detectors: [env, system]<br>    timeout: 2s<br>  signozspanmetrics/delta:<br>    metrics_exporter: signozclickhousemetrics<br>    metrics_flush_interval: 60s<br>    latency_histogram_buckets: [100us, 1ms, 2ms, 6ms, 10ms, 50ms, 100ms, 250ms, 500ms, 1000ms, 1400ms, 2000ms, 5s, 10s, 20s, 40s, 60s ]<br>    dimensions_cache_size: 100000<br>    aggregation_temporality: AGGREGATION_TEMPORALITY_DELTA<br>    enable_exp_histogram: true<br>    dimensions:<br>      - name: service.namespace<br>        default: default<br>      - name: deployment.environment<br>        default: default<br>      # This is added to ensure the uniqueness of the timeseries<br>      # Otherwise, identical timeseries produced by multiple replicas of<br>      # collectors result in incorrect APM metrics<br>      - name: signoz.collector.id<br>      - name: service.version<br>      - name: browser.platform<br>      - name: browser.mobile<br>      - name: k8s.cluster.name<br>      - name: k8s.node.name<br>      - name: k8s.namespace.name<br>      - name: host.name<br>      - name: host.type<br>      - name: container.name<br>extensions:<br>  health_check:<br>    endpoint: 0.0.0.0:13133<br>  pprof:<br>    endpoint: 0.0.0.0:1777<br>exporters:<br>  clickhousetraces:<br>    datasource: tcp://signoz:admin@172.31.249.100:9000/signoz_traces<br>    use_new_schema: true<br><br>  signozclickhousemetrics:<br>    dsn: tcp://signoz:admin@172.31.249.100:9000/signoz_metrics<br><br>  clickhouselogsexporter:<br>    dsn: tcp://signoz:admin@172.31.249.100:9000/signoz_logs<br>    timeout: 10s<br>    use_new_schema: true<br><br>  signozclickhousemeter:<br>    dsn: tcp://signoz:admin@172.31.249.100:9000/signoz_meter<br>    timeout: 45s<br>    sending_queue:<br>      enabled: false<br>service:<br>  telemetry:<br>    logs:<br>      encoding: json<br>  extensions:<br>    - health_check<br>    - pprof<br>  pipelines:<br>    traces:<br>      receivers: [otlp]<br>      processors: [resourcedetection, signozspanmetrics/delta, batch]<br>      exporters: [clickhousetraces, signozmeter]<br>    metrics:<br>      receivers: [otlp, hostmetrics]<br>      processors: [resourcedetection, batch]<br>      exporters: [signozclickhousemetrics, signozmeter]<br>    metrics/prometheus:<br>      receivers: [prometheus]<br>      processors: [resourcedetection, batch]<br>      exporters: [signozclickhousemetrics, signozmeter]<br>    logs:<br>      receivers: [otlp, syslog]<br>      processors: [resourcedetection, batch]<br>      exporters: [clickhouselogsexporter, signozmeter]<br>    metrics/meter:<br>      receivers: [signozmeter]<br>      processors: [batch/meter]<br>      exporters: [signozclickhousemeter]</pre><h4>STEP 6 — Configure OpenStack credentials for OpenTelemetry Collector</h4><pre>sudo nano/opt/otelcol-contrib/clouds.yaml<br><br>clouds:<br>  openstack-prod:<br>    region_name: RegionOne<br>    auth:<br>      auth_url: http://vip.internal:5000/v3<br>      username: admin<br>      password: password<br>      project_name: admin<br>      user_domain_name: Default<br>      project_domain_name: Default</pre><h4>STEP 7 — Get and extract OpenStack Exporter binary</h4><pre>wget https://github.com/openstack-exporter/openstack-exporter/releases/download/v1.8.0-alpha/openstack-exporter_1.8.0-alpha_linux_amd64.tar.gz<br>tar xvf openstack-exporter_1.8.0-alpha_linux_amd64.tar.gz</pre><h4>STEP 8 — Install and verify OpenStack Exporter binary</h4><pre>sudo mv openstack-exporter /usr/local/bin/<br>sudo chmod +x /usr/local/bin/openstack-exporter<br>openstack-exporter --version</pre><h4>STEP 9 — Configure the OpenStack Exporter systemd service</h4><pre>sudo bash -c &#39;cat &lt;&lt;EOF &gt; /etc/systemd/system/openstack-exporter.service<br>[Unit]<br>Description=OpenStack Prometheus Exporter<br>After=network-online.target<br><br>[Service]<br>User=root<br>Group=root<br>ExecStart=/usr/local/bin/openstack-exporter \<br>  --os-client-config /opt/otelcol-contrib/clouds.yaml \<br>  --cache \<br>  --cache-ttl=60s \<br>  --disable-slow-metrics \<br>  --disable-service.object-store \<br>  --disable-service.load-balancer \<br>  --disable-service.container-infra \<br>  --disable-service.dns \<br>  --disable-service.baremetal \<br>  --disable-service.gnocchi \<br>  --disable-service.database \<br>  --disable-service.sharev2 \<br>  openstack-prod<br>Restart=on-failure<br><br>[Install]<br>WantedBy=multi-user.target<br>EOF&#39;<br></pre><h4>STEP 10 — Reload the systemd daemon and start the OpenStack Exporter service</h4><pre>sudo systemctl daemon-reload<br>sudo systemctl start openstack-exporter.service</pre><h4>STEP 11 — Deploy OpenTelemetry Collector as a Docker container</h4><pre># Make sure you already have Docker<br>sudo docker pull signoz/signoz-otel-collector:latest<br>sudo docker run -d   --name signoz-otel-collector   --restart unless-stopped   --network host   -v /opt/signoz/otel-collector/config.yaml:/etc/otelcol-contrib/config.yaml   signoz/signoz-otel-collector:latest   --config /etc/otelcol-contrib/config.yaml<br>sudo docker ps</pre><h3>Verify the installation:</h3><p>Accessing SigNoz by Open Your Browser:</p><pre>http://YOURSERVERIP:8088</pre><h4>Final Result</h4><p>With this setup, SigNoz is successfully deployed in a highly available, self-hosted architecture.</p><ul><li>Highly Available Observability Stack</li><li>Scalable Storage Backend</li><li>Centralized Telemetry Ingestion</li><li>OpenStack Metrics Visibility</li><li>Clean Separation of Responsibilities</li></ul><h4>Conclusion</h4><p>This deployment demonstrates how SigNoz can be effectively self-hosted in production for OpenStack without relying on managed SaaS observability platforms. By combining OpenTelemetry, ClickHouse, ZooKeeper, and HAProxy, the result is a robust, scalable, and fully controlled observability solution suitable for private cloud environments.</p><p>Hopefully this article can provide benefits to readers. If you guys find mistakes in this article, please contact me on <a href="http://farreladitya375@gmail.com/">Email </a>so that I can fix it immediately.</p><p>Follow me on <a href="https://www.linkedin.com/in/farreladityah/">LinkedIn</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=01e41997f38c" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Installing Odoo 19 on Ubuntu 24.04]]></title>
            <link>https://farreladityah.medium.com/installing-odoo-19-on-ubuntu-24-04-4cd040708ed1?source=rss-4e1521e096db------2</link>
            <guid isPermaLink="false">https://medium.com/p/4cd040708ed1</guid>
            <category><![CDATA[odoo]]></category>
            <category><![CDATA[erp-software]]></category>
            <category><![CDATA[odoo-19]]></category>
            <dc:creator><![CDATA[Farrel Aditya]]></dc:creator>
            <pubDate>Tue, 06 Jan 2026 08:23:41 GMT</pubDate>
            <atom:updated>2026-01-06T08:24:12.901Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*KtqURbz9nr8f9tsstRYqeA.png" /><figcaption>Copyright Farrel Aditya Hermawan — @2025</figcaption></figure><p>Odoo is a strong, flexible ERP system that works for businesses of all sizes. With the new Ubuntu 24.04 (Noble) and Python 3.12 as the default programming language, Odoo 19 is the best and most reliable version to use.</p><p>This article guides you through setting up a clean and ready-to-use Odoo 19 installation on Ubuntu 24.04.</p><h4>Prerequisites</h4><ul><li>Ubuntu <strong>24.04 LTS</strong></li><li>A user with sudo privileges</li><li>At least <strong>2 GB RAM</strong> (4 GB recommended)</li><li>Internet access</li></ul><h4>STEP 1 — Updating the System</h4><pre>sudo apt update &amp;&amp; sudo apt upgrade -y</pre><h4>STEP 2 — Install PostgreSQL</h4><pre>sudo apt install postgresql postgresql-contrib -y</pre><p>Create a PostgreSQL user for Odoo:</p><pre>sudo su - postgres<br>createuser --createdb --username postgres --no-createrole --no-superuser odoo<br>exit</pre><h4>STEP 3 — Install System Dependencies</h4><pre>sudo apt install -y \<br>  git python3-pip python3-dev python3-venv \<br>  build-essential libxslt-dev libzip-dev libldap2-dev \<br>  libsasl2-dev libssl-dev libjpeg-dev libpq-dev \<br>  node-less npm wkhtmltopdf</pre><h4>STEP 4— Create the Odoo System User</h4><pre>sudo useradd -m -d /opt/odoo -U -r -s /bin/bash odoo</pre><p>If you want the user to run sudo commands:</p><pre>sudo passwd odoo<br>sudo usermod -aG sudo odoo</pre><h4>STEP 5 — Clone Odoo 19 Source Code</h4><pre># Odoo 19 is available on master branch<br><br>sudo su - odoo<br>git clone https://github.com/odoo/odoo \<br>  --depth 1 \<br>  --branch master \<br>  /opt/odoo/odoo-server</pre><h4>STEP 6 — Create a Python Virtual Environment</h4><pre># Ubuntu 24.04 ships with Python 3.12, which is fully supported by Odoo 19.<br><br>python3 -m venv /opt/odoo/venv<br>source /opt/odoo/venv/bin/activate</pre><p>Install Python dependencies:</p><pre>pip install -U pip setuptools wheel<br>pip install -r /opt/odoo/odoo-server/requirements.txt<br>deactivate<br>exit</pre><h4>STEP 7 — Configure Odoo</h4><p>Create the configuration directory and file:</p><pre>sudo mkdir /etc/odoo<br>sudo nano /etc/odoo/odoo.conf</pre><p>Use the following minimal and correct configuration for Odoo 19:</p><pre>[options]<br>http_interface = 0.0.0.0<br>http_port = 8069<br><br>admin_passwd = strong_master_password<br><br>db_host = False<br>db_port = False<br>db_user = odoo<br>db_password = False<br><br>addons_path = /opt/odoo/odoo-server/addons<br>logfile = /var/log/odoo/odoo.log<br><br>workers = 0<br>limit_time_real = 1200<br>limit_time_cpu = 1200<br>max_cron_threads = 1</pre><p>Set proper permissions:</p><pre>sudo chown odoo:odoo /etc/odoo/odoo.conf<br>sudo chmod 640 /etc/odoo/odoo.conf</pre><h4>STEP 8 — Create Log Directory</h4><pre>sudo mkdir -p /var/log/odoo<br>sudo chown odoo:odoo /var/log/odoo<br>sudo chmod 755 /var/log/odoo</pre><h4>STEP 9 — Create Systemd Service</h4><pre>sudo nano /etc/systemd/system/odoo.service<br><br>[Unit]<br>Description=Odoo 19<br>After=network.target postgresql.service<br><br>[Service]<br>Type=simple<br>User=odoo<br>Group=odoo<br>ExecStart=/opt/odoo/venv/bin/python3 /opt/odoo/odoo-server/odoo-bin -c /etc/odoo/odoo.conf<br>StandardOutput=journal<br>Restart=always<br><br>[Install]<br>WantedBy=multi-user.target</pre><p>Enable and start the service:</p><pre>sudo systemctl status odoo<br>sudo ss -lntp | grep 8069</pre><h4>STEP 10 — Verify Installation</h4><pre>sudo systemctl status odoo<br>sudo ss -lntp | grep 8069</pre><p>Expected output:</p><pre>LISTEN 0.0.0.0:8069</pre><p>Accessing Odoo by Open Your Browser:</p><pre>http://YOURSERVERIP:8069</pre><h4>Important Notes</h4><ul><li>First database creation always takes longer</li><li>Odoo 19 uses http_interface, not xmlrpc_interface</li><li>Python 3.12 is fully supported</li><li>Do not install all modules at once — start small</li></ul><h4>Final Result</h4><ul><li>Odoo 19 running on Ubuntu 24.04</li><li>Stable Python 3.12 environment</li><li>Modular ERP ready for customization</li><li>Suitable for development or production</li></ul><p>Hopefully this article can provide benefits to readers. If you guys find mistakes in this article, please contact me on <a href="http://farreladitya375@gmail.com/">Email </a>so that I can fix it immediately.</p><p>Follow me on <a href="https://www.linkedin.com/in/farreladityah/">LinkedIn</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4cd040708ed1" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Migrating VMware Virtual Machines to OpenStack]]></title>
            <link>https://farreladityah.medium.com/migrating-vmware-virtual-machines-to-openstack-376c5b80fc08?source=rss-4e1521e096db------2</link>
            <guid isPermaLink="false">https://medium.com/p/376c5b80fc08</guid>
            <category><![CDATA[cloud-computing]]></category>
            <category><![CDATA[vmware]]></category>
            <category><![CDATA[openstack]]></category>
            <dc:creator><![CDATA[Farrel Aditya]]></dc:creator>
            <pubDate>Wed, 10 Sep 2025 10:33:58 GMT</pubDate>
            <atom:updated>2025-09-10T10:33:58.795Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/486/1*9WWXzPIG_5hiD44AXPH-vA.png" /><figcaption>Copyright Farrel Aditya Hermawan — @2025</figcaption></figure><p>Migrating workloads from <strong>VMware</strong> to <strong>OpenStack</strong> is a common scenario for organizations adopting open-source cloud infrastructure. The process can look intimidating at first, but in practice it breaks down into a few manageable steps: exporting the VM from VMware, transferring the disk files, converting them into a format OpenStack understands, and finally uploading the image to Glance.</p><p>In this article, I’ll walk you through the process I used to migrate a VMware virtual machine into my OpenStack cloud environment.</p><h4>STEP 1 — Export the Virtual Machine from VMware</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1002/1*6-xoATZ1FJCBbJclVJVNWQ.png" /><figcaption>Copyright Farrel Aditya Hermawan — @2025</figcaption></figure><p>The first step is to export your VM from VMware in <strong>OVF</strong> format.</p><ul><li>Go to the top menu in VMware and choose <strong>File → Export → Export OVF Template</strong>.</li><li>This will generate two files:</li><li>An <strong>.ovf</strong> descriptor (metadata about the VM)</li><li>One or more <strong>.vmdk</strong> virtual disk files</li></ul><p>For migration, the important file is the <strong>.vmdk</strong> (the actual virtual disk).</p><h4>STEP 2 — Move the .vmdk Files to Your OpenStack Node</h4><p>Once exported, transfer the .vmdk file to your OpenStack environment. I used scp from my Windows host:</p><pre>PS C:\Users\Farrel Aditya&gt; scp &quot;D:\Software\ISO\Ubuntu_64-bit-disk1.vmdk&quot; init5@172.31.249.18:/home/init5/&quot;Ubuntu 64-bit.vmdk&quot;<br>init5@172.31.249.18&#39;s password:<br>Ubuntu_64-bit-disk1.vmdk                                            100% 4130MB   2.1MB/s   33:35<br>PS C:\Users\Farrel Aditya&gt;</pre><p>The command securely copies the exported disk file into the /home/init5 directory of my OpenStack controller node.</p><h4>STEP 3 — Verify the Uploaded .vmdk</h4><pre>init5@controller1:~$ sudo ls<br>&#39;Ubuntu 64-bit.ovf&#39;    &#39;Ubuntu 64-bit.vmdk&#39;   cirros-0.6.2-x86_64-disk.img<br>noble-server-cloudimg-amd64.img   proxysql.db.bak   admin-openrc</pre><p>At this point, the VMware virtual disk (Ubuntu 64-bit.vmdk) is ready for conversion.</p><h4>STEP 4 — Convert VMDK to QCOW2</h4><p>OpenStack (via Glance and Nova) works best with the <strong>QCOW2</strong> format. Use qemu-img to convert the .vmdk file into .qcow2:</p><pre>init5@controller1:~$ sudo qemu-img convert -f vmdk -O qcow2 &quot;/home/init5/Ubuntu 64-bit.vmdk&quot; &quot;/home/init5/Ubuntu 64-bit.qcow2&quot;</pre><p>This step may take some time depending on the size of your virtual disk. Once finished, you’ll have a .qcow2 image suitable for OpenStack.</p><h4>STEP 5 — Upload the QCOW2 Image to OpenStack</h4><p>Finally, upload the converted image into <strong>Glance</strong>, the OpenStack Image service:</p><pre>init5@controller1:~$ openstack image create &quot;Ubuntu-VMware&quot; \<br>  --disk-format qcow2 --container-format bare \<br>  --file &quot;/home/init5/Ubuntu 64-bit.qcow2&quot; \<br>  --public<br><br># Success indicator<br>+------------------+---------------------------------------------------------------------------------+<br>| Field            | Value                                                                           |<br>+------------------+---------------------------------------------------------------------------------+<br>| container_format | bare                                                                            |<br>| created_at       | 2025-09-10T07:57:27Z                                                            |<br>| disk_format      | qcow2                                                                           |<br>| file             | /v2/images/1e70d36b-a75b-465f-9c9d-07f80de1c788/file                            |<br>| id               | 1e70d36b-a75b-465f-9c9d-07f80de1c788                                            |<br>| min_disk         | 0                                                                               |<br>| min_ram          | 0                                                                               |<br>| name             | Ubuntu-VMware                                                                   |<br>| owner            | bae1b5b66ae345169940d29f56d14c35                                                |<br>| properties       | os_hidden=&#39;False&#39;, owner_specified.openstack.md5=&#39;&#39;,                            |<br>|                  | owner_specified.openstack.object=&#39;images/Ubuntu-VMware&#39;,                        |<br>|                  | owner_specified.openstack.sha256=&#39;&#39;                                             |<br>| protected        | False                                                                           |<br>| schema           | /v2/schemas/image                                                               |<br>| status           | queued                                                                          |<br>| tags             |                                                                                 |<br>| updated_at       | 2025-09-10T07:57:27Z                                                            |<br>| visibility       | public                                                                          |<br>+------------------+---------------------------------------------------------------------------------+<br><br># Checking the available images<br>init5@controller1:~$ openstack image list<br>+--------------------------------------+--------------------------+--------+<br>| ID                                   | Name                     | Status |<br>+--------------------------------------+--------------------------+--------+<br>| 1e70d36b-a75b-465f-9c9d-07f80de1c788 | Ubuntu-VMware            | active |<br>| 2df9ce3d-b03e-497b-be1b-48f83ef13aed | cirros-0.6.2             | active |<br>| c6d65c0b-7c7b-4447-9b6a-705965756ac1 | ubuntu-24.04-noble-cloud | active |<br>+--------------------------------------+--------------------------+--------+<br>init5@controller1:~$</pre><ul><li>--disk-format qcow2 tells Glance the image format.</li><li>--container-format bare specifies no additional container.</li><li>--public makes the image available to all projects (optional; you could keep it private if needed).</li></ul><p>After the upload completes, your VMware VM is now an <strong>OpenStack image</strong>. You can use it to launch instances just like any other cloud image.</p><p>Hopefully this article can provide benefits to readers. If you guys find mistakes in this article, please contact me on <a href="http://farreladitya375@gmail.com/">Email </a>so that I can fix it immediately.</p><p>Follow me on <a href="https://www.linkedin.com/in/farreladityah/">LinkedIn</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=376c5b80fc08" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[a Minimal OpenStack Core Components Installation and Configuration in Ubuntu 22.04]]></title>
            <link>https://farreladityah.medium.com/a-minimal-openstack-core-components-installation-and-configuration-in-ubuntu-22-04-897389e1a793?source=rss-4e1521e096db------2</link>
            <guid isPermaLink="false">https://medium.com/p/897389e1a793</guid>
            <category><![CDATA[cloud-infrastructure]]></category>
            <category><![CDATA[devops]]></category>
            <category><![CDATA[open-source]]></category>
            <category><![CDATA[openstack]]></category>
            <dc:creator><![CDATA[Farrel Aditya]]></dc:creator>
            <pubDate>Sat, 09 Aug 2025 06:34:22 GMT</pubDate>
            <atom:updated>2025-08-09T06:34:22.427Z</atom:updated>
            <content:encoded><![CDATA[<h3>a Minimal OpenStack Core Components Installation and Configuration in Ubuntu 22.04 without Automation</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/443/0*VsJDe_WnnhXHru9A" /><figcaption><a href="https://th.bing.com/th/id/OIP.6fXeFtGY_5CHdcv0hOAYFQHaDl?w=341&amp;h=169&amp;c=7&amp;r=0&amp;o=7&amp;dpr=1.3&amp;pid=1.7&amp;rm=3">https://th.bing.com/th/id/OIP.6fXeFtGY_5CHdcv0hOAYFQHaDl?w=341&amp;h=169&amp;c=7&amp;r=0&amp;o=7&amp;dpr=1.3&amp;pid=1.7&amp;rm=3</a></figcaption></figure><h4>Introduction</h4><p>OpenStack is a open source software for creating IAAS private, public, hybrid, or community cloud services. OpenStack allows users to manage and provision virtual resources such as compute, storage, and networking through a web-based dashboard, command-line tools, or RESTful web services.</p><p>Here are the key components of OpenStack:</p><ul><li>Nova (Compute Service)</li><li>Neutron (Networking Service)</li><li>Swift (Object Storage)</li><li>Cinder (Block Storage)</li><li>Keystone (Identity Service)</li><li>Glance (Image)</li><li>Horizon (Dashboard)</li></ul><p>In this tutorial, we will walk through building a minimal OpenStack environment manually on a single Ubuntu 22.04 VM. This setup is tailored for limited resources and skips Cinder and Swift f, focusing on the essential OpenStack services: <strong>Keystone, Glance, Placement, Nova, Neutron, and Horizon</strong>.</p><h4>Environmet Setup</h4><p>VM Specs:</p><ul><li>OS: Ubuntu 22.04</li><li>Hostname: openstack</li><li>RAM: 5GB</li><li>Storage: 50GB</li><li>vCPU: 4</li><li>Network Adapters: enp0s3 (NAT), enp0s8 (Host-only)</li></ul><p>Before starting the configuration:</p><pre>sudo apt update &amp;&amp; sudo apt upgrade -y<br>sudo apt install python3-openstackclient mariadb-server rabbitmq-server memcached etcd -y</pre><h4>STEP 1 — Database Configuration (MariaDB)</h4><pre>sudo nano /etc/mysql/mariadb.conf.d/99-openstack.cnf<br><br># Add<br>[mysqld]<br>bind-address = 0.0.0.0<br>default-storage-engine = innodb<br>innodb_file_per_table = on<br>max_connections = 4096<br>collation-server = utf8_general_ci<br>character-set-server = utf8</pre><pre># Applying the configuration<br><br>sudo systemctl restart mariadb</pre><h4>STEP 2 — RabbitMQ Configuration</h4><p>RabbitMQ is used for inter-service communication.</p><pre>sudo rabbitmqctl add_user openstack RABBIT_PASS<br>sudo rabbitmqctl set_permissions openstack &quot;.*&quot; &quot;.*&quot; &quot;.*&quot;</pre><h4>STEP 3 — Memcached Configuration</h4><pre>sudo nano /etc/memcached.conf<br><br># Change the bind address to your management IP:<br>-l 172.0.255.254</pre><pre>sudo systemctl restart memcached</pre><h4>STEP 4 — Keystone Installation (Identity Services)</h4><pre>sudo apt install keystone -y</pre><pre>sudo nano /etc/keystone/keystone.conf<br><br># Edit this line<br>[database]<br>connection = mysql+pymysql://keystone:KEYSTONE_DBPASS@172.0.255.254/keystone<br><br>[token]<br>provider = fernet</pre><pre># Sync database<br>sudo su -s /bin/sh -c &quot;keystone-manage db_sync&quot; keystone</pre><pre># Initalize fernet keys<br>keystone-manage fernet_setup --keystone-user keystone --keystone-group keystone<br>keystone-manage credential_setup --keystone-user keystone --keystone-group keystone</pre><pre># Bootstrap keystone<br>keystone-manage bootstrap --bootstrap-password ADMIN_PASS \<br>  --bootstrap-admin-url http://172.0.255.254:5000/v3/ \<br>  --bootstrap-internal-url http://172.0.255.254:5000/v3/ \<br>  --bootstrap-public-url http://172.0.255.254:5000/v3/ \<br>  --bootstrap-region-id RegionOne</pre><pre># Configure Apache<br>sudo nano /etc/apache2/apache2.conf<br><br>ServerName openstack</pre><pre>sudo systemctl restart apache2</pre><h4>STEP 5 — Glance Installation (Image Service)</h4><pre>sudo nano /etc/glance/glance-api.conf<br><br># Edit this<br>[database]<br>connection = mysql+pymysql://glance:GLANCE_DBPASS@172.0.255.254/glance<br><br>[keystone_authtoken]<br>www_authenticate_uri = http://172.0.255.254:5000<br>auth_url = http://172.0.255.254:5000<br>memcached_servers = 172.0.255.254:11211<br>auth_type = password<br>project_domain_name = Default<br>user_domain_name = Default<br>project_name = service<br>username = glance<br>password = GLANCE_PASS</pre><pre># Sync database<br>sudo su -s /bin/sh -c &quot;glance-manage db_sync&quot; glance</pre><pre>sudo systemctl restart glance-api</pre><h4>STEP 6 — Placement Service Installation</h4><pre>sudo apt install placement-api -y</pre><pre>sudo nano /etc/placement/placement.conf<br><br># Edit this<br>[placement_database]<br>connection = mysql+pymysql://placement:PLACEMENT_DBPASS@172.0.255.254/placement<br><br>[keystone_authtoken]<br>auth_url = http://172.0.255.254:5000/v3<br>memcached_servers = 172.0.255.254:11211<br>auth_type = password<br>project_domain_name = Default<br>user_domain_name = Default<br>project_name = service<br>username = placement<br>password = PLACEMENT_PASS</pre><pre># Sync database<br>sudo su -s /bin/sh -c &quot;placement-manage db sync&quot; placement</pre><h4>STEP 7 — Nova Installation (Compute Service)</h4><pre>sudo apt install nova-api nova-conductor nova-novncproxy nova-scheduler -y</pre><pre>sudo nano /etc/nova/nova.conf<br><br># Edit this<br>[api_database]<br>connection = mysql+pymysql://nova:NOVA_DBPASS@172.0.255.254/nova_api<br><br>[database]<br>connection = mysql+pymysql://nova:NOVA_DBPASS@172.0.255.254/nova<br><br>[DEFAULT]<br>transport_url = rabbit://openstack:RABBIT_PASS@172.0.255.254:5672/<br>my_ip = 172.0.255.254<br>use_neutron = False<br>firewall_driver = nova.virt.firewall.NoopFirewallDriver</pre><pre># Sync database<br>sudo su -s /bin/sh -c &quot;nova-manage api_db sync&quot; nova<br>sudo su -s /bin/sh -c &quot;nova-manage db sync&quot; nova</pre><pre>sudo systemctl restart nova-api nova-scheduler nova-conductor nova-novncproxy</pre><h4>STEP 8— Horizon Installation (Dashboard)</h4><pre>sudo apt install openstack-dashboard -y</pre><pre>sudo nano /etc/openstack-dashboard/local_settings.py<br><br># Edit this<br>OPENSTACK_HOST = &quot;172.0.255.254&quot;<br>ALLOWED_HOSTS = [&#39;*&#39;]</pre><pre>sudo systemctl restart apache2</pre><h4>STEP 9 — Neutron Installation using Linux Bridge (Networking Services)</h4><pre>sudo nano /etc/neutron/neutron.conf<br><br># Edit this<br>[database]<br>connection = mysql+pymysql://neutron:NEUTRON_DBPASS@controller/neutron<br><br>[DEFAULT]<br>core_plugin = ml2<br>service_plugins =<br>transport_url = rabbit://openstack:RABBIT_PASS@controller<br>auth_strategy = keystone<br><br>[keystone_authtoken]<br>auth_url = http://controller:5000<br>memcached_servers = controller:11211<br>auth_type = password<br>project_domain_name = Default<br>user_domain_name = Default<br>project_name = service<br>username = neutron<br>password = NEUTRON_PASS</pre><pre>sudo nano /etc/plugins/ml2/ml2_conf.ini<br><br># Edit this<br>[ml2]<br>type_drivers = flat,vlan<br>tenant_network_types =<br>mechanism_drivers = linuxbridge<br><br>[ml2_type_flat]<br>flat_networks = provider<br><br>[securitygroup]<br>enable_ipset = true</pre><pre>sudo nano /etc/plugins/ml2/linuxbridge_agent.ini<br><br># Edit this<br>[linux_bridge]<br>physical_interface_mappings = provider:enp0s8<br><br>[vxlan]<br>enable_vxlan = false<br><br>[securitygroup]<br>enable_security_group = true<br>firewall_driver = iptables</pre><pre>sudo systemctl restart neutron-server neutron-linuxbridge-agent neutron-dhcp-agent neutron-metadata-agent nova-api</pre><h4>STEP 10 — Configuring user file</h4><pre>sudo nano ~/admin-openrc<br><br># Fill with this<br>export OS_PROJECT_DOMAIN_NAME=Default<br>export OS_USER_DOMAIN_NAME=Default<br>export OS_PROJECT_NAME=admin<br>export OS_USERNAME=admin<br>export OS_PASSWORD=ADMIN_PASS<br>export OS_AUTH_URL=http://controller:5000/v3<br>export OS_IDENTITY_API_VERSION=3<br>export OS_IMAGE_API_VERSION=2</pre><pre># Applying the configuration<br>source ~/admin-openrc</pre><h4>Checking the services</h4><pre>openstack service list<br><br>openstack compute service list<br><br>openstack networking list<br><br>openstack endpoint list<br><br>openstack hypervisor list</pre><h3>Conclusion</h3><p>This tutorial demonstrates a minimal OpenStack deployment, focusing on the core services to make it lightweight and suitable for low-resource environments like single VMs.</p><p>You can expand this setup later by adding Cinder, or Swift for a more complete OpenStack environment.</p><p>Hopefully this article can provide benefits to readers. If you guys find mistakes in this article, please contact me on <a href="http://farreladitya375@gmail.com/">Email </a>so that I can fix it immediately.</p><p>Follow me on <a href="https://www.linkedin.com/in/farreladityah/">LinkedIn</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=897389e1a793" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Building a Proxmox Cluster with Ceph and ZFS: Setup, Failover, and Live Migration Benchmark]]></title>
            <link>https://farreladityah.medium.com/building-a-proxmox-cluster-with-ceph-and-zfs-setup-failover-and-live-migration-benchmark-d8db2ee11950?source=rss-4e1521e096db------2</link>
            <guid isPermaLink="false">https://medium.com/p/d8db2ee11950</guid>
            <category><![CDATA[benchmark]]></category>
            <category><![CDATA[clustering]]></category>
            <category><![CDATA[proxmox]]></category>
            <category><![CDATA[zfs]]></category>
            <category><![CDATA[ceph]]></category>
            <dc:creator><![CDATA[Farrel Aditya]]></dc:creator>
            <pubDate>Mon, 04 Aug 2025 06:17:02 GMT</pubDate>
            <atom:updated>2025-08-04T06:17:02.947Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*VngSF1VQNBs4vdSBGcTn_A.png" /><figcaption>Copyright Farrel Aditya Hermawam — @2025</figcaption></figure><h3>Introduction</h3><p>This blog post walks through the real-world setup of a 3-node Proxmox cluster with Ceph as a shared storage backend, followed by benchmarking live migration performance using both Ceph and ZFS. All steps are based on a hands-on implementation in a virtualized environment using VirtualBox and Proxmox VE.</p><h3>Environment Overview</h3><ul><li><strong>3 Proxmox nodes</strong>: proxmox1, proxmox2, proxmox3</li><li>Each node is assigned:</li><li>enp0s3: NAT interface (external access)</li><li>enp0s8: Bridge port for vmbr0</li><li>enp0s9: Internal cluster communication (172.0.100.0/24)</li><li>Extra disks for Ceph and ZFS storage</li><li><strong>Shared storage</strong>: Ceph (primary), ZFS (for benchmarking)</li><li><strong>Test VM</strong>: ID 100, MikroTik CHR ISO</li></ul><h4>STEP 1 — Proxmox Cluster Configuration</h4><pre># Configure /etc/network/interfaces (example: proxmox1)<br><br>auto vmbr0<br>iface vmbr0 inet static<br>    address 192.168.255.254/24<br>    gateway 192.168.255.1<br>    bridge-ports enp0s8<br>    bridge-stp off<br>    bridge-fd 0</pre><pre>auto enp0s9<br>iface enp0s9 inet static<br>    address 172.0.100.101/24</pre><p>Adjust accordingly for other nodes:</p><ul><li>proxmox2: 172.0.100.102</li><li>proxmox3: 172.0.100.103</li></ul><pre># Create Cluster on proxmox1<br><br>pvecm create cluster-lab</pre><pre># Join Cluster from proxmox2 and proxmox3<br><br>pvecm add 172.0.100.101</pre><pre># Checking Cluster status<br><br>pvecm status</pre><h4>STEP 2 — Ceph Storage Setup</h4><pre># Install Ceph Components<br><br>pveceph install</pre><pre># Initialize Monitor and Manager (on proxmox1)<br><br>pveceph init --network 172.0.100.0/24<br>pveceph mon create<br>pveceph mgr create</pre><pre># Add OSDs (on each node with /dev/sdb)<br><br>pveceph osd create /dev/sdb</pre><pre># Create and Register Pool<br><br>pveceph pool create cephcluster<br>pvesm add rbd cephcluster --content images,rootdir --pool cephcluster</pre><h4>STEP 3 — Create and Configure VM</h4><pre># Create MikroTik VM (ID 100)<br><br>qm create 100 --name test --memory 2048 --net0 virtio,bridge=vmbr0<br>qm set 100 --scsihw virtio-scsi-single --scsi0 cephcluster:4<br>qm set 100 --ide2 local:iso/mikrotik-7.15.3.iso,media=cdrom<br>qm set 100 --boot order=scsi0</pre><p>Start the VM and ensure it’s reachable via the vmbr0 network (192.168.255.0/24).</p><h4>STEP 4 — Failover and Manual Migration Testing</h4><pre># Live Migration with Ceph (proxmox2 -&gt; proxmox1)<br><br>time qm migrate 100 proxmox1 --online</pre><p><strong>Result:</strong></p><pre>real    0m2.308s<br>user    0m1.309s<br>sys     0m0.399s</pre><h4>STEP 5 — ZFS Benchmark Preparation</h4><pre># Attach Extra Disk on All Nodes<br><br>zpool create cluster /dev/sdc # Create ZFS Pool</pre><pre># Register ZFS Storage on All Nodes<br><br>pvesm add zfspool zfs-test --pool cluster --content images</pre><p>Confirm with:</p><pre>pvesm status</pre><h4>STEP 6 — Migrate Disk from Ceph to ZFS</h4><pre># Move VM Disk to ZFS<br><br>qm move_disk 100 scsi0 zfs-test --delete</pre><p>Verify disk now uses zfs-test:</p><pre>qm config 100</pre><h4>STEP 7 — Live Migration Benchmark (ZFS)</h4><pre># Live Migration with ZFS (proxmox1 -&gt; proxmox2)<br><br>time qm migrate 100 proxmox2 --online</pre><p><strong>Result:</strong></p><pre>real    0m2.458s<br>user    0m1.440s<br>sys     0m0.553s</pre><h3>Conclusion</h3><p>This walkthrough demonstrates:</p><ul><li>Creating a 3-node Proxmox cluster</li><li>Configuring Ceph as a distributed storage backend</li><li>Moving VMs across nodes with live migration</li><li>Comparing migration speeds between Ceph and ZFS</li></ul><p>For labs and proof-of-concept deployments, both Ceph and ZFS offer flexibility. Ceph shines in HA and scalability, while ZFS provides speed and simplicity for small setups.</p><p>Hopefully this article can provide benefits to readers. If you guys find mistakes in this article, please contact me on <a href="http://farreladitya375@gmail.com/">Email </a>so that I can fix it immediately.</p><p>Follow me on <a href="https://www.linkedin.com/in/farreladityah/">LinkedIn</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d8db2ee11950" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Understanding Active vs Passive FTP: A Practical Analysis with Wireshark]]></title>
            <link>https://farreladityah.medium.com/understanding-active-vs-passive-ftp-a-practical-analysis-with-wireshark-e19bfe96c4c2?source=rss-4e1521e096db------2</link>
            <guid isPermaLink="false">https://medium.com/p/e19bfe96c4c2</guid>
            <category><![CDATA[wireshark]]></category>
            <category><![CDATA[ftp]]></category>
            <category><![CDATA[analysis]]></category>
            <dc:creator><![CDATA[Farrel Aditya]]></dc:creator>
            <pubDate>Thu, 31 Jul 2025 06:17:04 GMT</pubDate>
            <atom:updated>2025-07-31T06:17:04.258Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*yDe8lLZNFWeluibVz1uZ5w.png" /><figcaption>Copyright Farrel Aditya Hermawan — @<a href="https://medium.com/u/ee9640cbfc64">2025</a></figcaption></figure><h3>FTP Introduction</h3><p>File Transfer Protocol (FTP) is one of the oldest yet still relevant methods of transferring files between systems. When diagnosing connectivity or troubleshooting uploads/downloads, one crucial aspect to understand is the difference between <strong>active</strong> and <strong>passive</strong> FTP modes. In this blog, we’ll walk through two practical FTP sessions I captured and analyzed using Wireshark.</p><h3>What’s the Difference?</h3><h4>Active FTP</h4><ul><li><strong>Client opens a command connection</strong> (TCP port 21) to the server.</li><li><strong>Server initiates a data connection</strong> back to the client using <strong>port 20</strong> (FTP-data).</li><li>Client tells the server which IP and port it’s listening on using the PORT command.</li><li>Often blocked by NAT/firewalls because the server initiates the data connection.</li></ul><h4>Passive FTP</h4><ul><li>Client opens both the <strong>command (port 21)</strong> and <strong>data connections</strong>.</li><li>Client tells the server to switch to passive mode with PASV.</li><li>Server responds with an IP and port for data, and the client connects to it.</li><li>More firewall/NAT friendly since client initiates all connections.</li></ul><h3>Case Study 1: Active FTP (Captured from Windows CMD)</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*pLH1XOlynddNp81MmYfemw.png" /><figcaption>Copyright Farrel Aditya Hermawan —@ <a href="https://medium.com/u/ee9640cbfc64">2025</a></figcaption></figure><h4>Indicators of Active FTP:</h4><p><strong>1 — Client sent </strong><strong>PORT command:</strong></p><ul><li>PORT 192,168,255,100,231,120</li></ul><p>This means the client is requesting the server to connect to IP 192.168.255.100 at port 231*256 + 120 = 59256.</p><p><strong>2 — Server responds:</strong></p><ul><li>200 PORT command successful. Consider using PASV.</li></ul><p><strong>3 — Data connection initiated from server to client (Src Port = 20):</strong></p><ul><li>TCP 20 → 59256</li></ul><p><strong>4 — Data transfer occurred (LIST, NLST commands).</strong></p><h4>Conclusion:</h4><p>The Windows Command Prompt FTP client by default uses <strong>active mode</strong>, explaining why this behavior was observed.</p><h3>Case Study 2: Passive FTP (Captured from FileZilla)</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*D5BKQxBBwkQW5hH4KgymTQ.png" /><figcaption>Copyright Farrel Aditya Hermawan — @<a href="https://medium.com/u/ee9640cbfc64">2025</a></figcaption></figure><h4>Indicators of Passive FTP:</h4><p><strong>1 — Client sent </strong><strong>PASV command:</strong></p><ul><li>PASV</li></ul><p><strong>2 — Server responded with IP and port for data connection:</strong></p><ul><li>227 Entering Passive Mode (192,168,255,254,156,144)</li></ul><p>Which decodes to IP 192.168.255.254 and port 156*256 + 144 = 40080.</p><p><strong>3 — Client initiated data connection to port 40080.</strong></p><p><strong>4 — Data commands like </strong><strong>LIST, </strong><strong>STOR, and </strong><strong>MDTM occurred.</strong></p><h4>Conclusion:</h4><p>FileZilla, like most modern GUI FTP clients, defaults to <strong>passive mode</strong> because it’s more compatible with firewalls and NAT.</p><h3>Summary Table</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/665/1*oYZrVbhrkHzCCNxq4aC_kg.png" /><figcaption>Copyright Farrel Aditya Hermawan — @2025</figcaption></figure><p>Hopefully this article can provide benefits to readers. If you guys find mistakes in this article, please contact me on <a href="http://farreladitya375@gmail.com/">Email </a>so that I can fix it immediately.</p><p>Follow me on <a href="https://www.linkedin.com/in/farreladityah/">LinkedIn</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e19bfe96c4c2" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Building a ZFS RAIDZ1 Storage Pool on Ubuntu 22.04: Step-by-Step Guide]]></title>
            <link>https://farreladityah.medium.com/building-a-zfs-raidz1-storage-pool-on-ubuntu-22-04-step-by-step-guide-69c63ffd4379?source=rss-4e1521e096db------2</link>
            <guid isPermaLink="false">https://medium.com/p/69c63ffd4379</guid>
            <category><![CDATA[guides-and-tutorials]]></category>
            <category><![CDATA[linux]]></category>
            <category><![CDATA[storage]]></category>
            <category><![CDATA[virtualization]]></category>
            <dc:creator><![CDATA[Farrel Aditya]]></dc:creator>
            <pubDate>Wed, 30 Jul 2025 05:01:26 GMT</pubDate>
            <atom:updated>2025-07-30T09:27:22.763Z</atom:updated>
            <content:encoded><![CDATA[<h3>Building a ZFS RAIDZ1 Storage Pool on Ubuntu 22.04</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*9XLh7EbLvhO-hbC9yJhF7w.png" /><figcaption>Copyright Farrel Aditya Hermawan — @2025</figcaption></figure><h4>Introduction</h4><p>When managing disk storage for critical systems, two priorities always stand out: <strong>data integrity</strong> and <strong>fault tolerance</strong>. One of the most powerful tools for this job is <strong>ZFS</strong>, an advanced filesystem and volume manager with built-in RAID capabilities.</p><p>In this guide, I’ll show you how I created a <strong>RAIDZ1 ZFS pool</strong> on a single Ubuntu 22.04 VM in VirtualBox, using <strong>three 1GB virtual disks</strong>. We’ll also dive into how ZFS works, the role of RAID, and why <strong>RAIDZ1</strong> is a practical choice for small setups.</p><h4>What is ZFS?</h4><p>ZFS (Zettabyte File System) is a <strong>next-generation file system and volume manager</strong> originally developed by Sun Microsystems. It is designed for high availability and data protection.</p><h4>Key features:</h4><ul><li><strong>Pooled storage</strong> (no partitions or LVM required)</li><li><strong>Data integrity checks</strong> and self-healing</li><li><strong>Snapshots and rollback</strong></li><li><strong>Native RAID support</strong> (RAIDZ)</li><li><strong>Compression</strong>, deduplication, and more</li></ul><h4>What is RAID?</h4><p>RAID stands for <strong>Redundant Array of Independent Disks</strong> — a method of combining multiple physical disks into a single logical volume for:</p><ul><li><strong>Performance</strong> (via striping)</li><li><strong>Redundancy</strong> (via mirroring or parity)</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/631/1*_GsE2y_7UjD6MqkBllJN-g.png" /><figcaption>Copyright Farrel Aditya Hermawan — @2025</figcaption></figure><h4>What is Striping, Mirroring, and Parity?</h4><ul><li><strong>Striping:</strong></li></ul><p>Striping is the process of <strong>splitting data into small blocks</strong> and writing them <strong>across multiple disks in a rotating sequence</strong>.</p><p><strong>Main goal:</strong> Improve performance, especially <strong>read/write speed</strong>.</p><ul><li><strong>Mirroring:</strong></li></ul><p>Mirroring means <strong>duplicating the same data to two or more disks</strong>, so each disk contains an <strong>exact copy</strong>.</p><p><strong>Main goal:</strong> Ensure <strong>data availability and fault tolerance</strong>.</p><ul><li><strong>Parity:</strong></li></ul><p>Parity is <strong>extra information</strong> calculated from real data using logical operations (like XOR). It allows you to <strong>rebuild lost data if one disk fails</strong>.</p><p><strong>Main goal:</strong> Provide <strong>redundancy while saving space</strong>.</p><h4>What is RAIDZ?</h4><p>ZFS implements its own RAID system, called <strong>RAIDZ</strong>, which is similar to traditional RAID 5 but avoids issues like the “write hole.”</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/413/1*IKyQyk2wYsffhpAsr_kCcQ.png" /><figcaption>Copyright Farrel Aditya Hermawan — <a href="https://medium.com/u/ee9640cbfc64">2025</a></figcaption></figure><h4>Lab Setup</h4><p>To test RAIDZ1, I created a virtual lab using:</p><ul><li><strong>Ubuntu 22.04 LTS</strong> on <strong>VirtualBox</strong></li><li><strong>3 x 1GB virtual disks</strong> (added via VirtualBox)</li><li><strong>ZFS utilities installed via APT</strong></li></ul><h3>RAIDZ1 Installation &amp; Configuration Steps</h3><h4>STEP 1 — Add 3 virtual disks (1GB each) in VirtualBox</h4><p>In your VM settings:</p><ul><li>Go to Settings &gt; Storage</li><li>Add 3 new VDI disks under the SATA controller</li><li>Ensure your main disk (/dev/sda) is untouched</li></ul><h4>STEP 2 — Start the VM and update packages</h4><pre>sudo apt update<br>sudo apt install zfsutils-linux -y<br><br># If any mirrors fail, replace id.archive.ubuntu.com with archive.ubuntu.com in /etc/apt/sources.list.</pre><h4>STEP 3 — Check available disks</h4><pre>lsblk<br><br># For example, you should see like this<br><br>init5@zfsraid:~$ lsblk<br>NAME   MAJ:MIN RM   SIZE RO TYPE MOUNTPOINTS<br>loop0    7:0    0 111.9M  1 loop /snap/lxd/24322<br>loop1    7:1    0  63.4M  1 loop /snap/core20/1974<br>loop2    7:2    0  53.3M  1 loop /snap/snapd/19457<br>sda      8:0    0    20G  0 disk<br>├─sda1   8:1    0     1M  0 part<br>├─sda2   8:2    0    12G  0 part /<br>├─sda3   8:3    0     4G  0 part /home<br>├─sda4   8:4    0     2G  0 part [SWAP]<br>└─sda5   8:5    0     2G  0 part /boot<br>sdb      8:16   0     1G  0 disk<br>├─sdb1   8:17   0  1014M  0 part<br>└─sdb9   8:25   0     8M  0 part<br>sdc      8:32   0     1G  0 disk<br>├─sdc1   8:33   0  1014M  0 part<br>└─sdc9   8:41   0     8M  0 part<br>sdd      8:48   0     1G  0 disk<br>├─sdd1   8:49   0  1014M  0 part<br>└─sdd9   8:57   0     8M  0 part<br>sr0     11:0    1  1024M  0 rom</pre><h4>STEP 4 — Wipe any existing data (Optional)</h4><pre>sudo wipefs -a /dev/sdb /dev/sdc /dev/sdd</pre><h4>STEP 5 — Create RAIDZ1 pool</h4><pre>sudo zpool create test raidz1 /dev/sdb /dev/sdc /dev/sdd</pre><h4>STEP 6 — Verify pool status</h4><pre>zpool status<br>zpool list<br>df -h | grep test<br><br># For example, you should see like this<br>init5@zfsraid:~$ zpool status<br>  pool: test<br> state: ONLINE<br>  scan: resilvered 0B in 00:00:01 with 0 errors on Wed Jul 30 04:20:24 2025<br>config:<br><br>        NAME        STATE     READ WRITE CKSUM<br>        test        ONLINE       0     0     0<br>          raidz1-0  ONLINE       0     0     0<br>            sdb     ONLINE       0     0     0<br>            sdc     ONLINE       0     0     0<br>            sdd     ONLINE       0     0     0<br>init5@zfsraid:~$ zpool list<br>NAME   SIZE  ALLOC   FREE  CKPOINT  EXPANDSZ   FRAG    CAP  DEDUP    HEALTH  ALTROOT<br>test  2.75G   240K  2.75G        -         -     0%     0%  1.00x    ONLINE  -<br>init5@zfsraid:~$ df -h | grep test<br>test            1.8G  128K  1.8G   1% /test</pre><h3>Pool Testing</h3><h4>Creating and Reading files:</h4><pre>init5@zfsraid:~$ echo &quot;hello zfs&quot; | sudo tee /test/hello.txt<br>hello zfs<br><br>init5@zfsraid:~$ sudo cat /test/hello.txt<br>hello zfs</pre><h4>Snapshot and Rollback:</h4><pre>init5@zfsraid:~$ sudo echo &quot;before snapshoot&quot; | sudo tee /test/snapdemo.txt<br>before snapshoot<br><br>init5@zfsraid:~$ sudo zfs snapshot test@satu<br><br>init5@zfsraid:~$ sudo echo &quot;after snapshoot&quot; | sudo tee /test/snapdemo.txt<br>after snapshoot<br><br>init5@zfsraid:~$ sudo cat /test/snapdemo.txt<br>after snapshoot<br><br>init5@zfsraid:~$ sudo zfs rollback test@satu<br><br>init5@zfsraid:~$ sudo cat /test/snapdemo.txt<br>before snapshoot</pre><h4>Disk Failure simulation:</h4><pre>init5@zfsraid:~$ sudo zpool offline test sdc<br><br>init5@zfsraid:~$ sudo zpool status<br>  pool: test<br> state: DEGRADED<br>status: One or more devices has been taken offline by the administrator.<br>        Sufficient replicas exist for the pool to continue functioning in a<br>        degraded state.<br>action: Online the device using &#39;zpool online&#39; or replace the device with<br>        &#39;zpool replace&#39;.<br>config:<br><br>        NAME        STATE     READ WRITE CKSUM<br>        test        DEGRADED     0     0     0<br>          raidz1-0  DEGRADED     0     0     0<br>            sdb     ONLINE       0     0     0<br>            sdc     OFFLINE      0     0     0<br>            sdd     ONLINE       0     0     0<br><br>init5@zfsraid:~$ sudo zpool online test sdc<br><br>init5@zfsraid:~$ sudo zpool status<br>  pool: test<br> state: ONLINE<br>config:<br><br>        NAME        STATE     READ WRITE CKSUM<br>        test        ONLINE       0     0     0<br>          raidz1-0  ONLINE       0     0     0<br>            sdb     ONLINE       0     0     0<br>            sdc     ONLINE       0     0     0<br>            sdd     ONLINE       0     0     0</pre><h3>RAID5 Installation &amp; Configuration Steps</h3><h4>STEP 1 — Update Package Lists &amp; Install mdadm</h4><pre>sudo apt update &amp;&amp; sudo apt install mdadm -y</pre><h4>STEP 2 — List all disks</h4><pre>lsblk</pre><h4>STEP 3 — Wipe Existing Filesystem Signatures</h4><pre>sudo wipefs -a /dev/sdc /dev/sdd /dev/sde</pre><h4>STEP 4 — Create RAID5 Array</h4><pre>sudo mdadm --create --verbose /dev/md0 --level=5 --raid-devices=3 /dev/sdc /dev/sdd /dev/sde</pre><h4>STEP 5 — Check RAID Synchronization Progress</h4><pre>cat /proc/mdstat</pre><h4>STEP 6 — Format the RAID Array with ext4 Filesystem</h4><pre>sudo mkfs.ext4 /dev/md0</pre><h4>STEP 7 — <strong>Create Mount Directory</strong></h4><pre>sudo mkdir /mnt/raid5</pre><h4>STEP 8 — Mount the RAID Array</h4><pre>sudo mount /dev/md0 /mnt/raid5</pre><h4>STEP 9 — Save RAID Configuration for Boot-Time Assembly</h4><pre>sudo mdadm --detail --scan | sudo tee -a /etc/mdadm/mdadm.conf</pre><h4>STEP 10 — Update Initial RAM Filesystem</h4><pre>sudo update-initramfs -u</pre><h3>Pool Testing</h3><h4>Checking RAID Status</h4><pre>sudo cat /proc/mdstat</pre><h4>Creating and Reading Files</h4><pre>sudo touch /mnt/raid5/testfile.txt<br><br>echo &quot;RAID5 test successful!&quot; | sudo tee /mnt/raid5/testfile.txt<br>RAID5 test successful!<br><br>cat /mnt/raid5/testfile.txt<br>RAID5 test successful!</pre><h4>Disk Failure Simulation</h4><pre>sudo mdadm --fail /dev/md0 /dev/sdd<br>mdadm: set /dev/sdd faulty in /dev/md0 # Output<br><br>sudo mdadm --remove /dev/md0 /dev/sdd<br>mdadm: hot removed /dev/sdd from /dev/md0<br><br>cat /proc/mdstat # Checking RAID status, as you can see [U_U] means a disk has failed<br>Personalities : [linear] [multipath] [raid0] [raid1] [raid6] [raid5] [raid4] [raid10]<br>md0 : active raid5 sde[3] sdc[0]<br>      2093056 blocks super 1.2 level 5, 512k chunk, algorithm 2 [3/2] [U_U]<br><br>sudo wipefs -a /dev/sdd # Removing the cache in the previous disk to add it back<br>/dev/sdd: 4 bytes were erased at offset 0x00001000 (linux_raid_member): fc 4e 2b a9<br><br>sudo mdadm --add /dev/md0 /dev/sdd # Adding disk<br>mdadm: added /dev/sdd<br><br>cat /proc/mdstat # Checking RAID status again, and array status is normal [UUU]<br>Personalities : [linear] [multipath] [raid0] [raid1] [raid6] [raid5] [raid4] [raid10]<br>md0 : active raid5 sdd[4] sde[3] sdc[0]<br>      2093056 blocks super 1.2 level 5, 512k chunk, algorithm 2 [3/3] [UUU]</pre><h3>Key Factors When Choosing RAID/RAIDZ</h3><ol><li><strong>Desired Redundancy Level</strong></li></ol><ul><li>How many disks can fail without causing data loss?</li></ul><p><strong>2. Required Read/Write Performance</strong></p><ul><li>Is your workload read-heavy, write-heavy, or mixed?</li></ul><p><strong>3. Disk Capacity Efficiency</strong></p><ul><li>What percentage of total disk space do you want to utilize?</li></ul><p><strong>4. Number of Available Disks</strong></p><ul><li>Some RAID/RAIDZ levels require a minimum of 3, 4, or more disks.</li></ul><p><strong>5. Type of Workload</strong></p><ul><li>Examples include OLTP databases, file servers, backups/archives, VM storage, etc.</li></ul><h4>Results &amp; Observations</h4><p>The RAIDZ1 pool was:</p><ul><li>Created successfully</li><li>Mountable and writable at /test</li><li>Capable of snapshot/rollback</li><li>Automatically resilient to one disk failure (tested via zpool offline)</li></ul><p>This makes RAIDZ1 a great choice for minimal fault tolerance when using <strong>3 disks or more</strong>.</p><h4>Recommended Use Cases for RAID &amp; RAIDZ:</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/791/1*pkqOM4q2lOA1NCzHrzJLQg.png" /><figcaption>Copyright Farrel Aditya Hermawan — <a href="https://medium.com/u/ee9640cbfc64">2025</a></figcaption></figure><h4>Conclusion</h4><p>ZFS is a <strong>powerful tool for storage management</strong>, especially when used with RAIDZ. Even in a small lab setup with limited RAM and virtual disks, its performance, snapshot capabilities, and fault tolerance are impressive.</p><p>If you’re exploring ZFS for the first time or looking to simulate enterprise-grade storage systems in a home lab — <strong>Ubuntu 22.04 + ZFS RAIDZ1 is a great place to start</strong>.</p><p>Hopefully this article can provide benefits to readers. If you guys find mistakes in this article, please contact me on <a href="http://farreladitya375@gmail.com/">Email </a>so that I can fix it immediately.</p><p>Follow me on <a href="https://www.linkedin.com/in/farreladityah/">LinkedIn</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=69c63ffd4379" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Build a Minimal Ceph Cluster Using ceph-deploy]]></title>
            <link>https://farreladityah.medium.com/build-a-minimal-ceph-cluster-using-ceph-deploy-5f136eb9434f?source=rss-4e1521e096db------2</link>
            <guid isPermaLink="false">https://medium.com/p/5f136eb9434f</guid>
            <category><![CDATA[storage]]></category>
            <category><![CDATA[high-availability]]></category>
            <category><![CDATA[ceph]]></category>
            <dc:creator><![CDATA[Farrel Aditya]]></dc:creator>
            <pubDate>Tue, 29 Jul 2025 10:54:45 GMT</pubDate>
            <atom:updated>2025-07-30T04:50:11.552Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/575/1*Qo3iDrpng4ymGduIt0o2Cg.png" /><figcaption>Copyright Farrel Aditya Hermawan — @2025</figcaption></figure><p>Setting up a Ceph cluster may sound intimidating, but with the right structure and verified steps, it becomes highly manageable even when done manually. In this guide, I walk through my real-life simulation of building a Ceph storage cluster using 3 Ubuntu VMs on VirtualBox: one admin node and two monitor nodes.</p><h3>Overview</h3><ul><li><strong>Tooling Used</strong>: ceph-deploy, Ceph Quincy (17.2.9)</li><li><strong>OS</strong>: Ubuntu 20.04 (admin node), Ubuntu 22.04 (monitor nodes)</li><li><strong>Cluster Layout</strong>:</li><li>ceph-deploy (admin node)</li><li>ceph-monitor1 (monitor + OSD + MGR)</li><li>ceph-monitor2 (monitor + OSD + MGR standby)</li></ul><h3>Prerequisites</h3><p>Ensure all nodes:</p><ul><li>Have hostname resolution between each other (via /etc/hosts or DNS)</li><li>Are accessible via passwordless SSH (ssh-copy-id)</li><li>Are time-synchronized (e.g., using chrony)</li></ul><p>Update and install dependencies on all nodes:</p><pre>sudo apt update &amp;&amp; sudo apt upgrade -y<br>sudo apt install ceph-deploy -y  # Only on admin node</pre><h3>Step-by-Step Guide</h3><h4>STEP 1 — Create Working Directory on Admin Node</h4><pre>mkdir ~/ceph-cluster<br>cd ~/ceph-cluster</pre><h4>STEP 2 — Create the Initial Cluster</h4><pre>ceph-deploy new ceph-monitor1 ceph-monitor2</pre><p>This generates ceph.conf and ceph.mon.keyring.</p><h4>STEP 3 — Install Ceph on All Nodes</h4><pre>ceph-deploy install ceph-monitor1 ceph-monitor2</pre><h4>STEP 4 — Deploy Monitors</h4><pre>ceph-deploy mon create ceph-monitor1 ceph-monitor2</pre><p>Check status:</p><pre>ceph-deploy mon list ceph-monitor1</pre><h4>STEP 5 — Copy Configuration and Admin Keyring</h4><pre>ceph-deploy admin ceph-monitor1 ceph-monitor2</pre><p>Ensure correct permissions:</p><pre>sudo chown ceph:ceph /etc/ceph/ceph.client.admin.keyring<br>sudo chmod 640 /etc/ceph/ceph.client.admin.keyring<br>sudo chown root:root /etc/ceph/ceph.conf<br>sudo chmod 644 /etc/ceph/ceph.conf</pre><p>Create symbolic keyring (required for ceph CLI to work normally):</p><pre>sudo ln -s /etc/ceph/ceph.client.admin.keyring /etc/ceph/keyring<br>sudo chown ceph:ceph /etc/ceph/*keyring<br>sudo chmod 640 /etc/ceph/*keyring</pre><h4>STEP 6 — Prepare and Activate OSDs</h4><p>Example for /dev/sdb:</p><pre>ceph-deploy osd prepare ceph-monitor1:/dev/sdb<br>ceph-deploy osd activate ceph-monitor1:/dev/sdb1</pre><pre>ceph-deploy osd prepare ceph-monitor2:/dev/sdb<br>ceph-deploy osd activate ceph-monitor2:/dev/sdb1</pre><h4>STEP 7 — Deploy Manager Daemon</h4><pre>ceph-deploy mgr create ceph-monitor1<br>ceph-deploy mgr create ceph-monitor2</pre><h4>Final Verification</h4><p>Check cluster health:</p><pre>ceph -s</pre><p>Example output:</p><pre>cluster:<br>  id:     41bfdafd-3802-4e90-b92b-105e6b9aa557<br>  health: HEALTH_WARN<br>          mon is allowing insecure global_id reclaim<br>services:<br>  mon: 1 daemons, quorum ceph-monitor1<br>  mgr: ceph-monitor1(active), standbys: ceph-monitor2<br>  osd: 2 osds: 2 up, 2 in<br>data:<br>  pools:   1 pools, 1 pgs<br>  objects: 2 objects, 577 KiB<br>  usage:   583 MiB used, 19 GiB / 20 GiB avail<br>  pgs:     1 active+clean</pre><h4>Conclusion</h4><p>This setup is now functional, with:</p><ul><li>Dual monitors (for redundancy)</li><li>Active and standby managers</li><li>2 OSDs storing replicated objects</li></ul><p>This tutorial reflects <strong>tested and verified commands</strong> from my actual environment using ceph-deploy and is suitable as a learning lab or prototype foundation.</p><p>Hopefully this article can provide benefits to readers. If you guys find mistakes in this article, please contact me on <a href="http://farreladitya375@gmail.com/">Email </a>so that I can fix it immediately.</p><p>Follow me on <a href="https://www.linkedin.com/in/farreladityah/">LinkedIn</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5f136eb9434f" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>