Linux Server Admin with Fexingo: Sysadmin, Bash, and Server Engineering

When your server talks to clients halfway around the world, default TCP settings can choke performance. In this episode, Lucas and Luna explain how to tune Linux server TCP parameters for high-latency links. They walk through a real-world case: a video transcoding service in Singapore serving users in South America, where default TCP windowing caused throughput to stall at 12 Mbps. Learn how adjusting net.core.rmem_max, net.core.wmem_max, net.ipv4.tcp_rmem, and net.ipv4.tcp_wmem, plus enabling BBR congestion control, boosted throughput to 85 Mbps. Lucas breaks down the math: bandwidth-delay product, Socket buffer sizing, and why 'one size fits all' doesn't work for global traffic. Luna shares a tip on persistent configuration with sysctl and when to use tc qdisc for shaping. Practical sysadmin advice you can apply today.

#LinuxServerAdmin #TCPTuning #HighLatency #BBR #Sysadmin #NetworkPerformance #BandwidthDelayProduct #SocketBuffers #Sysctl #LinuxNetworking #ServerOptimization #VideoTranscoding #TechTips #OpenSource #FexingoBusiness #TechnologyPodcast #LinuxAdmin #NetworkTuning

Keep every episode free: buymeacoffee.com/fexingo

In this episode, Lucas and Luna dive deep into Linux network namespaces as a practical tool for multi-tenant server isolation. They walk through a real-world scenario: setting up two isolated network stacks on a single host using 'ip netns', configuring virtual Ethernet pairs, and assigning IP addresses so each namespace can talk to the outside world independently. Lucas explains the difference between network namespaces and VLANs, and why namespaces are lighter weight for container-like workloads. Luna asks about troubleshooting common pitfalls like missing routes or firewall rules that leak between namespaces. The hosts also touch on how cloud providers use network namespaces under the hood for virtual networks. If you manage servers where multiple clients or services need separate network environments without full virtualization, this episode gives you the concrete steps to implement it today.

#Linux #NetworkNamespaces #ServerAdmin #Sysadmin #Bash #ContainerNetworking #MultiTenant #ipNetns #VirtualEthernet #NetworkIsolation #Technology #FexingoBusiness #BusinessPodcast #Fexingo #LinuxNetworking #ServerEngineering #SysadminTips #Networking

Keep every episode free: buymeacoffee.com/fexingo

In this episode of Linux Server Admin with Fexingo, Lucas and Luna dive into the practical side of container runtimes—specifically, how to choose between Docker, containerd, and Podman for production server workloads. They break down the architecture of each runtime, explain how the Open Container Initiative (OCI) standards ensure interoperability, and walk through a real-world scenario: deploying a Python web app on a bare-metal server using containerd directly (without Docker). Lucas shares a surprising benchmark: containerd can reduce memory overhead by 30-40% compared to a full Docker daemon on lightweight servers. Luna asks about security differences, especially around daemonless architectures and rootless containers. By the end, you'll know when to reach for which runtime and how to test them yourself. Perfect for sysadmins tired of Docker defaults or curious about lower-level container orchestration. Produced by Fexingo Business Podcast Network.

#ContainerRuntimes #Docker #Containerd #Podman #OCI #LinuxServerAdmin #Sysadmin #DevOps #Containers #ServerDeployment #PythonApp #RootlessContainers #Daemonless #OpenContainerInitiative #Technology #Fexingo #FexingoBusiness #BusinessPodcast

Keep every episode free: buymeacoffee.com/fexingo

In this episode of Linux Server Admin with Fexingo, Lucas and Luna dive into a powerful but often overlooked feature of Logical Volume Manager (LVM): snapshot-based backups for fast recovery. They walk through a real-world scenario where a failed software update on a production database server requires rolling back in under 10 minutes. Lucas explains how LVM snapshots work at the block level, the key difference between thin and thick snapshots, and the gotchas around space allocation and performance. Listeners will learn the exact commands to create, mount, and merge a snapshot, plus a production-ready backup script pattern. Perfect for sysadmins who want to recover quickly without restoring from full disk images. Episode 56 of Linux Server Admin with Fexingo – ad-free and listener-supported.

#Linux #Sysadmin #LVM #Snapshots #Backup #Recovery #ServerEngineering #Bash #BlockLevel #ThinSnapshots #DatabaseBackup #Production #Rollback #Technology #LinuxServerAdmin #Fexingo #FexingoBusiness #BusinessPodcast

Keep every episode free: buymeacoffee.com/fexingo

Episode 55 of Linux Server Admin tackles a common but infuriating problem: DNS resolution failures that can cripple server connectivity. Lucas and Luna walk through a real-world scenario where an application suddenly can't reach external APIs. They cover the diagnostic toolkit — digging into /etc/resolv.conf, testing with dig and nslookup, checking systemd-resolved stubs, and tracing timeouts with tcpdump. Specific commands like 'dig +trace example.com' and 'resolvectl query' are demonstrated. They also discuss how stub resolvers can mask upstream failures and when to use fallback nameservers. By the end, you'll have a repeatable process for isolating whether the problem is a misconfigured resolver, a blocked port, or a slow upstream server. No fluff — just practical server troubleshooting for sysadmins.

#LinuxServerAdmin #DNSResolution #Sysadmin #Troubleshooting #Linux #Dig #Nslookup #SystemdResolved #Tcpdump #Resolvectl #Technology #ServerAdministration #NetworkDiagnostics #DnsDebugging #FexingoBusiness #BusinessPodcast #ServerEngineering #Bash

Keep every episode free: buymeacoffee.com/fexingo

Episode 54 of Linux Server Admin with Fexingo dives into Logical Volume Manager (LVM) for flexible server storage. Lucas and Luna walk through a real-world scenario: a web server running out of disk space on its root partition. They explain how LVM's physical volumes, volume groups, and logical volumes let you resize, snapshot, and migrate storage without downtime. Specific commands include pvcreate, vgcreate, vgextend, lvextend, resize2fs, and lvcreate with the --snapshot flag. The hosts also cover when LVM adds complexity versus when it's a lifesaver, and warn against thin provisioning traps. Perfect for sysadmins managing on-prem or cloud Linux servers.

#Linux #LVM #StorageManagement #Sysadmin #ServerEngineering #Bash #LinuxServerAdmin #VolumeManagement #DiskManagement #LogicalVolumeManager #Technology #FexingoBusiness #BusinessPodcast #Podcast #LinuxAdmin #ServerStorage #StoragePools #ResizePartition

Keep every episode free: buymeacoffee.com/fexingo

In this episode of Linux Server Admin, Lucas and Luna dive into Linux namespaces for user isolation — a key technique for securing multi-tenant servers and containers. They explain how user namespaces map unprivileged UIDs to root inside a container, preventing container breakouts. Lucas walks through a practical example: creating a user namespace with unshare, mapping UIDs, and running a shell as 'root' inside the namespace without real privileges. They discuss real-world use cases like running legacy applications or multi-user development environments without risking the host. The episode also touches on common pitfalls, such as improper UID mapping that can allow privilege escalation. Packed with actionable commands and insights, this is essential for any sysadmin hardening Linux servers.

#Linux #Sysadmin #Namespaces #UserIsolation #ContainerSecurity #Unshare #UIDMapping #ServerHardening #MultiTenant #PrivilegeEscalation #Technology #LinuxServerAdmin #FexingoBusiness #BusinessPodcast #Podcast #DevOps #Security #Bash

Keep every episode free: buymeacoffee.com/fexingo

In this episode, Lucas and Luna dive into the practical steps for hardening SSH on Linux servers using FIDO2/U2F security keys. They explain why password and even key-based authentication often fall short, then walk through the setup: generating a key backed by a hardware token, configuring sshd to accept only ed25519-sk keys, and disabling password login. The conversation covers real-world friction points like handling multiple servers, backup authentication methods, and the surprising lesson one sysadmin learned after locking themselves out. By the end, listeners have a concrete, implementable security upgrade for their own infrastructure.

#Linux #SSH #SecurityKeys #FIDO2 #U2F #Sysadmin #ServerHardening #Ed25519 #Cybersecurity #Authentication #OpenSSH #Technology #FexingoBusiness #BusinessPodcast #ServerAdmin #InfrastructureSecurity #SSHConfig #HardwareToken

Keep every episode free: buymeacoffee.com/fexingo

Episode 51 of Linux Server Admin with Fexingo dives into BPF (Berkeley Packet Filter) as a practical tool for real-time performance tracing on Linux servers. Hosts Lucas and Luna walk through a concrete scenario: diagnosing a mysterious latency spike in a web application without restarting or installing new packages. Lucas explains how bpftrace lets you attach one-liners to kernel probes to measure disk I/O, syscall latency, and scheduling delays, using real examples like tracing ext4 file operations and identifying a slow NFS server. The episode covers the difference between classic BPF and eBPF, common one-liners for sysadmins, and safety mechanisms that prevent crashes. Listeners learn how BPF can replace older tools like strace for production use by being safer and more efficient. The hosts also touch on how to check if your kernel supports BPF and where to find pre-built tools like BCC. Perfect for server admins who want deeper observability without overhead.

#Linux #BPF #eBPF #PerformanceTracing #Sysadmin #Bpftrace #Kernel #Observability #ServerManagement #LinuxPerformance #BCC #Troubleshooting #Technology #FexingoBusiness #BusinessPodcast #ServerEngineering #DevOps #RealTimeMonitoring

Keep every episode free: buymeacoffee.com/fexingo

In episode 50 of Linux Server Admin with Fexingo, Lucas and Luna dive into Linux namespaces for process isolation, focusing on the PID and network namespaces. They walk through real commands — unshare, nsenter, ip netns — to show how to create isolated environments without Docker. Lucas shares a story about using namespaces to run a legacy application in isolation on a production server, avoiding the overhead of full containerization. They also cover how to inspect existing namespaces and common pitfalls like namespace leaks. Perfect for sysadmins who want lightweight isolation without pulling in container runtimes. Tune in for practical, copy-paste-ready commands.

#Linux #Namespaces #ProcessIsolation #Sysadmin #Bash #ServerEngineering #ContainerSecurity #Unshare #Nsenter #PIDNamespace #NetNS #Tech #Podcast #FexingoBusiness #BusinessPodcast #DevOps #Infrastructure #SystemAdministration

Keep every episode free: buymeacoffee.com/fexingo