website·Indexed 30 days ago
https://blog.yasitha.dev/A collection of technical guides and tutorials focusing on infrastructure and system administration. Topics include multi-master Kubernetes cluster setups with Ceph and vSphere on CentOS 8, HashiCorp Vault HA clusters, etcd HA clusters, Confluent Kafka deployment, and MariaDB Galera Cluster configuration. It also provides instructions for AWS services like Athena and AMI creation, as well as OS-specific guides for CentOS, Ubuntu, macOS, and Windows.
/opt/nexus with a dedicated service account.# Install Prerequisites
yum install -y epel-release unzip vim wget
# Install OpenJDK 8
yum -y install java-1.8.0-openjdk java-1.8.0-openjdk-devel
# Create folder structure
mkdir -p /data/nexus-data /opt/nexus
# Download and extract Nexus
wget -O /tmp/nexus.tar.gz http://download.sonatype.com/nexus/3/latest-unix.tar.gz
tar xvfz /tmp/nexus.tar.gz -C /opt/nexus --strip-components 1
# Create service account and set permissions
sudo useradd --system --no-create-home nexus
chown -R nexus:nexus /opt/nexus
chown -R nexus:nexus /data/nexus-dataBefore installing Kafka, ensure each node has a unique MAC address and product_uuid. Use the following commands to verify:
ip link | grep link/ethercat /sys/class/dmi/id/product_uuidThen, perform the system preparation steps: cleaning YUM cache, updating packages, installing prerequisites, and configuring Java 8.
# Clean YUM repository cache and update packages
yum clean all -y
yum update -y
# Install prerequisites
yum install -y curl which ntp
# Install JDK 8
yum install -y java-1.8.0-openjdk java-1.8.0-openjdk-devel
# Set JAVA_HOME in /etc/bashrc
echo 'export JAVA_HOME=$(dirname $(dirname $(readlink $(readlink $(which javac)))))' >> /etc/bashrc
source /etc/bashrc
# Verify Java version
java -versionsudo nmcli connection modify ens192 ipv6.method ignoreFollow these steps to access the macOS Utilities menu when the standard boot process does not provide a recovery option:
recovery and press Enter.recovery option and follow the prompts until the macOS Utilities menu appears.autofs service. This process involves installing the necessary Samba and autofs packages, defining a master mount point, creating a map file for the specific share, and securing a credentials file for authentication.# 1. Install necessary packages
yum install samba-client samba-common autofs
# 2. Create the mount point
mkdir /mnt/cifs_share
# 3. Add the mount point and map file to /etc/auto.master
# Format: <mount-point> <map-file> --timeout=<seconds> --ghost
echo "/mnt/cifs_share /etc/auto.cifs --timeout=600 --ghost" >> /etc/auto.master
# 4. Create the map file /etc/auto.cifs
# Format: <mount-name> -fstype=cifs,rw,noperm,credentials=<path-to-creds> ://<Host Name>/<Folder Name>
echo "share_data -fstype=cifs,rw,noperm,credentials=/etc/credentials.txt ://<Host Name>/<Folder Name>" > /etc/auto.cifs
# 5. Create and secure the credentials file
echo "username=DOMAIN/User
password=Password" > /etc/credentials.txt
chmod 0700 /etc/credentials.txt
# 6. Restart the automounter service
/etc/init.d/autofs restartsystemd cgroup driver, which is required for Kubernetes stability. Set the storage driver to overlay2 and limit log sizes to 100MB.# Add repo and install
dnf config-manager --add-repo https://download.docker.com/linux/centos/docker-ce.repo
dnf install -y docker-ce-20.10.1 docker-ce-cli-20.10.1
# Configure daemon.json
mkdir -p /etc/docker
cat <<EOF | sudo tee /etc/docker/daemon.json > /dev/null
{
"exec-opts": ["native.cgroupdriver=systemd"],
"log-driver": "json-file",
"log-opts": {
"max-size": "100m"
},
"storage-driver": "overlay2",
"storage-opts": [
"overlay2.override_kernel_check=true"
]
}
EOF
# Start service
systemctl daemon-reload
systemctl enable --now dockerREADY and all pods in kube-system are running before joining new nodes. Use the join command generated during the first master's initialization.# Example Control Plane Join (requires token and cert-key from first master)
kubeadm join kube-api.example.local:6443 --token <token> \
--discovery-token-ca-cert-hash sha256:<hash> \
--control-plane --certificate-key <key> --v=5
# Example Worker Join
kubeadm join kube-api.example.local:6443 --token <token> \
--discovery-token-ca-cert-hash sha256:<hash> --v=5To deploy an etcd High Availability (HA) cluster, the following minimum specifications are required per node for a 3-node cluster:
# Create ETCD service user
sudo useradd --system --no-create-home etcd
# Create ETCD directory structure
sudo mkdir -p /etc/etcd/tls /var/lib/etcd
# Provide necessary permissions
sudo chown -R etcd:etcd /etc/etcd /var/lib/etcd
sudo chmod 0700 /var/lib/etcd
# Restore SELinux context
sudo restorecon -RvF /etc/etcd /var/lib/etcd