{"canonical_url":"https://engineer.company/glossary/linux/","content_markdown":"# Linux\n\n## What it is\n\nLinux is a free, Unix-like operating system kernel: the program that starts\nfirst, drives the hardware, shares processor time and memory among every other\nprogram, and decides which of them may touch which file or network port. In\neveryday speech the word stretches to cover the whole operating system assembled\naround that kernel, and some people write GNU/Linux for the combination, to\ncredit the GNU tools that usually come with it.\n\nLinus Torvalds started it in 1991 as a Unix clone written from scratch, and a\nloose network of developers working over the internet has built it ever since.\nIt aims at the POSIX and Single UNIX Specification standards, so software\nwritten for one Unix-like system usually builds on it with little change. It\nbegan on the 32-bit Intel 386 and now runs on a long list of 32-bit and 64-bit\narchitectures. The code is licensed under the GPL, version 2 only, with an\nexplicit note that programs which merely call into the kernel are not covered by\nit. That note is why proprietary software runs on Linux without a licensing\nquestion.\n\nThe kernel alone is nothing you can log in to. A working system also needs a C\nlibrary, a shell, a service manager, a package manager and thousands of\nprograms, and the projects that assemble all of that into something installable\nare called distributions. Debian, Ubuntu, Red Hat Enterprise Linux and its\nrebuilds, SUSE, Fedora and Alpine are distributions. Each ships its own build of\nthe kernel, its own defaults and its own promise about how long it will keep\nfixing security problems.\n\nIts rivals depend on the job. On servers it is weighed against Windows Server\nand against the BSD family, FreeBSD and OpenBSD, which are complete Unix-like\nsystems developed as one piece rather than assembled from parts. On laptops it\nmeets Windows and macOS.\n\n## Why it matters\n\nFor a company that runs software on its own servers or in a cloud, the operating\nsystem has mostly been chosen before anyone asks. Web servers, databases,\ncontainer engines and monitoring tools are built for Linux first and tested\nthere hardest. The decision that remains is narrower, and it is worth getting\nright.\n\nLinux versus Windows Server comes down to the software, not the system. A\nbusiness built on Active Directory, IIS or older .NET applications has real\nreasons to stay on Windows. Nearly everything else in a modern web stack, from\nPostgreSQL and Nginx to Kubernetes, assumes Linux, and even Microsoft SQL Server\nhas a Linux edition.\n\nThe cost case is plain. The kernel and community distributions such as Debian\nand Ubuntu cost nothing to install on as many machines as you like, and there is\nno licence count to audit; Windows Server is sold under a commercial licence.\nWith Linux the money goes on people and, if you want one, on a support\nsubscription from a vendor such as Canonical, Red Hat or SUSE.\n\nLinux containers tie the knot tighter. Each image packages a Linux user space\ninto a file, and running it means running ordinary processes on a Linux kernel.\nOn a Mac or a Windows laptop, tools running these containers therefore start a\nsmall Linux virtual machine first; Apple's own `container` tool runs each one\nin a lightweight virtual machine of its own. Choosing Linux containers brings\nLinux into the stack, whatever the developers' laptops run.\n\nWhat actually separates two Linux servers is the distribution and how long it is\nmaintained. The kernel project releases a new mainline version every nine to ten\nweeks and keeps a handful of longterm branches for years: 6.12, for instance, is\nprojected to receive fixes until December 2028. Distributions build their own\nmaintenance on top, and a distribution's kernel is supported by the\ndistribution, not by the kernel developers. Pick the distribution for its\nsupport period and for the skills the team already has.\n\nOne rule of the kernel project makes upgrades less risky than that pace\nsuggests: a change that breaks software which used to work counts as a\nregression, and regressions get fixed, often by reverting the change. Programs\nbuilt long ago keep running on new kernels because of it.\n\nThe price of all this choice is that “we run Linux” describes very little. Two\ndistributions differ in package manager, firewall tool, network configuration\nand release rhythm, and a team that knows one has to relearn parts of the other.\n\n## What it looks like in practice\n\nThe closest view is a single 464 MB host running a forge and a web server, where the alarm for running out of memory or disk is a [dead-man's switch on a timer](/portfolio/built-dead-mans-switch-host-monitoring/). Every fifteen minutes, with jitter, a script measures available memory and free disk and pings an outside service only while both clear their thresholds, so silence is the alert. The script had always read the right figure, available memory, while six variable names around it said “free”. Because the kernel spends idle memory on cache, free memory on a healthy Linux box sits near zero: against a ten per cent threshold a reader would have seen twelve megabytes free on the 464 MB machine, and a one-character “fix” would have turned a working alarm into one that breaches permanently. The names were corrected. A received signal confirms that the script ran and its available-memory and free-disk thresholds passed; it does not certify general host health or that the site is serving. Monitoring can expose warning signs, although the timing depends on the workload and the conditions observed.\n\nIn an earlier role, the founder built an organisation's\n[first observability dashboard](/portfolio/built-the-companys-first-observability-dashboard/), which\ncollected CPU, RAM, HDD, temperature and other metrics from remote Linux servers\nover SSH, and watched Docker containers the same way. A second version moved to\nGrafana and Prometheus. It ran on a large TV in the office, with alerts on\nthreshold breaches, and teams detected and resolved incidents 40% faster.\n\nAcross a whole estate, access became something written down rather than\nremembered.\nThe Linux servers were one of the systems that\n[1,000 RBAC rules](/portfolio/wrote-1000-rbac-rules-to-secure-linux-servers/)\ngoverned, alongside developer environments, application instances and the\nPostgreSQL and MS SQL databases, and Ansible applied the rules consistently\nacross environments. Developers could reach only their designated servers.\nThe delivered result was a documented role-based rule set with a repeatable\nway to apply and review it.\n\n## Where it goes wrong\n\n**The dashboard shows memory almost full on a server that is perfectly\nhealthy.** Linux treats unused memory as wasted and fills it with cached file\ndata, which it hands back the moment a program asks. The “free” figure counts\nonly memory doing nothing at all, so it drifts towards zero on any busy machine.\nThe figure that answers “can this server take more work?” is MemAvailable, the\nkernel's own estimate of what new programs could use without swapping. Alert on\nthat one:\n\n```sh\ngrep -E '^(MemFree|MemAvailable):' /proc/meminfo\n```\n\n**A service disappears, and the explanation sits in the kernel log rather than\nits own.** By default Linux lets programs reserve more memory than the machine\nhas, on the bet that much of it will never be touched. When the bet loses and\nmemory truly runs out, the out-of-memory killer chooses a process, usually a\nlarge one, and ends it. The application gets no chance to write anything; the\nrecord is in the kernel log, which `journalctl -k` shows. Give each service a\nmemory ceiling of its own, through systemd's `MemoryMax=` or a container limit,\nso a runaway program is stopped before the kernel has to choose, and watch\navailable memory instead of waiting for the kill.\n\n**A kernel bug report sent upstream goes unanswered.** The machine runs its\ndistribution's kernel, which carries the distribution's own patches, and the\nkernel developers support only their own releases. `uname -r` shows the version,\nand anything after the dash, as in `6.8.0-45-generic`, marks a distribution\nbuild. Report the problem to the distribution first, or reproduce it on a\ncurrent kernel from kernel.org before taking it upstream.\n\n**Security fixes stop arriving while the server carries on as before.** Every\ndistribution release and every longterm kernel has an end date, and the machine\nitself does not announce it. The longterm 5.10 and 5.15 branches, for example,\nare projected to reach theirs in December 2026. Keep the end-of-support date in\nthe same inventory as the server's name, and plan the upgrade one release ahead\nrather than at the deadline.\n\n## Further reading\n\n- [About Linux Kernel](https://www.kernel.org/linux.html)\n  — the kernel project's own short account of what Linux is and why a newcomer\n  wants a distribution.\n- [Releases](https://www.kernel.org/category/releases.html)\n  — mainline, stable and longterm kernels, with the projected end of every\n  longterm branch.\n- [The /proc Filesystem](https://docs.kernel.org/filesystems/proc.html)\n  — where MemFree, MemAvailable and the other memory counters are defined.\n- [Overcommit Accounting](https://docs.kernel.org/mm/overcommit-accounting.html)\n  — the three overcommit modes and what each one promises a program.\n- [Reporting regressions](https://docs.kernel.org/admin-guide/reporting-regressions.html)\n  — the “no regressions” rule, and how to report a breakage so it counts.\n- [Linux kernel licensing rules](https://docs.kernel.org/process/license-rules.html)\n  — GPL-2.0 only, and the syscall note that leaves programs on top of it free.\n\n\u003chttps://engineer.company/glossary/linux/\u003e\n","date_modified":"2026-11-12T05:34:23Z","description":"Linux for people choosing servers: kernel versus distribution, what it costs, how long support lasts, and the memory figures that mislead its users.","formats":{"gemtext":"https://engineer.company/glossary/linux/index.gmi","gopher":"https://engineer.company/glossary/linux/gophermap","html":"https://engineer.company/glossary/linux/","json":"https://engineer.company/glossary/linux/index.json","markdown":"https://engineer.company/glossary/linux/index.md","pdf":"https://engineer.company/glossary/linux/index.pdf"},"kind":"page","language":"en","title":"What is Linux?","url":"https://engineer.company/glossary/linux/","version":1}