Files
platform/scripts/setup/machine-setup/machine-setup.sh
T
pastilhasandClaude Opus 5 9b0e05bfd8 add three resource-pressure sections, each asked rather than assumed
Swap covers memory pressure. These are its neighbours:

  8.  Emergency disk ballast  the same valve, for disk
  9.  earlyoom                what happens when swap runs out too
  10. inotify watch limit     the silent one

All three follow the rule this script now works to: the role sets which way the
recommendation points, never whether the question is asked. A dev machine is
still offered the ballast, with the recommendation pointing the other way; a
server is still offered the inotify raise, because anything running `bun --watch`
or serving a file browser is a watcher too.

The ballast is section 22 of the original, moved up beside swap where it belongs
and moved out of the user's home. The original wrote the checker into
$USER_HOME/.local/bin and ran it from a root cron — a root cron executing a
script in a directory its owner can write is a privilege escalation waiting to be
noticed. Moot on a box where that user already has passwordless sudo, but wrong.
Both the checker and the file are in root-owned system paths now.

Two bugs found by running the generated checker rather than reading it:

  It df'd the ballast's own directory, which does not exist before the ballast is
  created — and with `set -euo pipefail` that meant cron mailing an error every
  ten minutes. It now walks up to a directory that exists, and the installer
  creates the directory itself rather than depending on the create step.

  The inotify text claimed a default of 8192. This host is at 29461: Ubuntu
  raised it, and stating a number the reader can see is wrong on their own screen
  undermines the rest of the explanation. It now describes the failure instead
  and prints the machine's actual value.

earlyoom is a distro package and a systemd unit, so it is checked with
`systemctl is-active` and reports honestly when it installs but fails to start.

Verified: checker --status and its no-op path both exit 0 with no directory
present, and the helpers report correctly against this host.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-08-12 17:54:48 +00:00

490 lines
19 KiB
Bash
Executable File

#!/bin/bash
set -e
# =============================================================================
# machine-setup — provisioning for a fresh machine
#
# Brings a blank box up to a usable state: users, SSH, networking, firewall,
# Docker, shell and editor tooling, language runtimes.
#
# Run as root: sudo scripts/setup/machine-setup/machine-setup.sh
# =============================================================================
SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
PROGRESS_FILE="$SCRIPT_DIR/.setup-progress"
# Shared state, output helpers, the step/resume machine and OS detection. Kept in
# lib/ so a step can eventually be read — or run — on its own without dragging the
# whole script in. Definitions only; nothing in there acts.
# shellcheck source=lib/base.sh
source "$SCRIPT_DIR/lib/base.sh"
# shellcheck source=lib/packages.sh
source "$SCRIPT_DIR/lib/packages.sh"
# shellcheck source=lib/tools.sh
source "$SCRIPT_DIR/lib/tools.sh"
# shellcheck source=lib/system.sh
source "$SCRIPT_DIR/lib/system.sh"
# Trap errors with context. Installed here rather than in lib/base.sh, because
# that file is definitions only and a trap is a side effect on whoever sources it.
trap 'echo ""; echo -e "${RED}╔══════════════════════════════════════════════════╗${NC}"; echo -e "${RED}║ SETUP FAILED${NC}"; echo -e "${RED}║ Step: ${CURRENT_STEP:-unknown}${NC}"; echo -e "${RED}║ Line: $LINENO${NC}"; echo -e "${RED}║ Command: $BASH_COMMAND${NC}"; echo -e "${RED}╚══════════════════════════════════════════════════╝${NC}"' ERR
# =============================================================================
# 1. Pre-flight
# =============================================================================
echo ""
echo -e "${BOLD}╔══════════════════════════════════════════════════╗${NC}"
echo -e "${BOLD}║ Machine Setup ║${NC}"
echo -e "${BOLD}╚══════════════════════════════════════════════════╝${NC}"
detect_os
echo ""
info "Machine: ${OS_NAME} (${ARCH})"
info "Packages: ${PM:-none detected}"
[[ "$IS_WSL" == true ]] && warn "WSL detected — the suspend, logind and boot-hang steps do not apply here"
# Everything below this line is written against apt and systemd. Detection above
# recognises pacman, dnf and brew so the branches have somewhere to hang, but
# nothing implements them yet — and running the apt path on Arch would half-build
# a machine and stop somewhere unhelpful. Refuse clearly instead, and relax this
# list one entry at a time as each package manager grows a real path.
case "$PM" in
apt) ;;
"") fail "Could not find a package manager for '${OS_NAME}'." ;;
*) fail "${OS_NAME} uses ${PM}, which this script does not implement yet — apt-based systems only, so far." ;;
esac
ask_machine_role
info "Role: ${MACHINE_ROLE}"
if [[ -f "$PROGRESS_FILE" ]]; then
DONE_COUNT=$(wc -l < "$PROGRESS_FILE")
echo -e "${YELLOW} Resuming — $DONE_COUNT step(s) already completed${NC}"
echo -e "${YELLOW} Progress file: $PROGRESS_FILE${NC}"
echo -e "${YELLOW} To start fresh: rm $PROGRESS_FILE${NC}"
fi
if [[ "$EUID" -ne 0 ]]; then
fail "Please run as root: sudo ./machine-setup.sh"
fi
prompt_value USERNAME "New admin username (or existing)" ""
if [[ -z "$USERNAME" ]]; then
fail "Username cannot be empty"
fi
USER_HOME="/home/$USERNAME"
# Always, and outside any step: everything below reads this index — core utils,
# the fastfetch PPA, the Docker repo — and `step` skips a step whose name is
# already in the progress file. With the refresh inside one of those, a resumed
# run installed against whatever the index happened to say hours or days ago.
echo ""
info "Refreshing the package index..."
pkg_refresh >/dev/null
# =============================================================================
# 2. System update
# =============================================================================
#
# Its own section because it is the only thing in the script that moves versions
# of software already on the machine. Everything else only ever adds what is
# absent, so this is the one that deserves to be refused on its own.
#
# The index refresh is NOT here — it runs in pre-flight, unconditionally, because
# every later step reads it and this one can be skipped.
step "System update"
if ! skip; then
mapfile -t UPGRADABLE < <(pkg_upgradable)
echo ""
info "System update — upgrades installed packages to their latest versions"
if ((${#UPGRADABLE[@]} == 0)); then
echo " to upgrade: nothing, everything is current"
SUMMARY+=("System update: already up to date")
else
echo " to upgrade: ${#UPGRADABLE[@]} package(s)"
# Capped, because a box that has not been touched in months lists hundreds
# and a wall of names is no more informative than a count.
printf ' %s\n' "${UPGRADABLE[@]:0:25}"
((${#UPGRADABLE[@]} > 25)) && echo " … and $((${#UPGRADABLE[@]} - 25)) more"
if confirm "Proceed?"; then
pkg_upgrade_all
ok "System upgraded"
SUMMARY+=("System upgraded: ${#UPGRADABLE[@]} package(s)")
else
warn "skipped by request"
SUMMARY+=("System update: SKIPPED by request — ${#UPGRADABLE[@]} package(s) left as they are")
fi
fi
step_ok
fi
# =============================================================================
# 3. Core utils
# =============================================================================
#
# What the distribution provides: the six this script would break without, and
# the command-line tools that make a machine worth sitting at.
step "Core utils"
if ! skip; then
# shellcheck disable=SC2046 # word splitting is how the list is passed
pkg_install "Core utils" $(pkgs_core)
summarise_last "Core utils"
step_ok
fi
# =============================================================================
# 4. Command-line tools
# =============================================================================
#
# A different thing from core utils, and kept apart from them: upstream binaries
# fetched from upstream, on their own release cadence, none of which the
# distribution ships. Lumping them in made a run look like it was installing
# system packages and then start pulling tarballs unannounced.
step "Command-line tools"
if ! skip; then
# shellcheck disable=SC2046 # word splitting is how the list is passed
tools_install "Command-line tools" $(tools_default)
summarise_last "Command-line tools"
step_ok
fi
# =============================================================================
# 5. Locale
# =============================================================================
#
# LOCALE in the environment overrides the default.
step "Locale"
if ! skip; then
LOCALE="${LOCALE:-en_US.UTF-8}"
CURRENT_LOCALE="$(locale_current)"
echo ""
info "Locale — the system language and character encoding"
echo " current: ${CURRENT_LOCALE:-none configured}"
echo " to set: ${LOCALE}"
if [[ "$CURRENT_LOCALE" == "$LOCALE" ]] && locale_is_generated "$LOCALE"; then
echo " already set and generated, nothing to do"
SUMMARY+=("Locale: already ${LOCALE}")
else
# Say which of the two is actually wrong, since they fail differently: a
# missing LANG means the C locale, a missing generation means every login
# prints a setlocale warning.
[[ "$CURRENT_LOCALE" != "$LOCALE" ]] && echo " LANG is not set to it"
locale_is_generated "$LOCALE" || echo " the locale has not been generated on this machine"
if confirm "Proceed?"; then
locale_set "$LOCALE"
ok "Locale set to ${LOCALE}"
SUMMARY+=("Locale: ${LOCALE}")
else
warn "skipped by request"
SUMMARY+=("Locale: SKIPPED by request — left at ${CURRENT_LOCALE:-unset}")
fi
fi
step_ok
fi
# =============================================================================
# 6. Timezone
# =============================================================================
#
# TIMEZONE in the environment answers the prompt ahead of time.
step "Timezone"
if ! skip; then
CURRENT_TZ="$(timezone_current)"
echo ""
info "Timezone — what logs, timers and every printed date are relative to"
echo " current: ${CURRENT_TZ:-unknown}"
if [[ -z "${TIMEZONE:-}" ]]; then
echo ""
for i in "${!TZ_OPTIONS[@]}"; do
printf ' [%d] %s\n' "$((i + 1))" "${TZ_OPTIONS[$i]}"
done
echo ""
while [[ -z "${TIMEZONE:-}" ]]; do
if ! read -rp " Pick a number, or type a zone name — Enter keeps ${CURRENT_TZ:-the current one}: " TZ_CHOICE; then
echo ""
fail "No answer. Set TIMEZONE=<zone> to answer this ahead of time."
fi
if [[ -z "$TZ_CHOICE" ]]; then
TIMEZONE="$CURRENT_TZ"
elif [[ "$TZ_CHOICE" =~ ^[0-9]+$ ]]; then
if ((TZ_CHOICE >= 1 && TZ_CHOICE <= ${#TZ_OPTIONS[@]})); then
TIMEZONE="${TZ_OPTIONS[$((TZ_CHOICE - 1))]}"
else
warn "There is no option ${TZ_CHOICE}."
fi
else
# Validated here rather than left to timedatectl, which fails on an
# unknown name and would take the whole run down over a typo.
if timezone_is_valid "$TZ_CHOICE"; then
TIMEZONE="$TZ_CHOICE"
else
warn "Not a zone this machine knows: '${TZ_CHOICE}' — try e.g. Europe/Berlin"
fi
fi
done
elif ! timezone_is_valid "$TIMEZONE"; then
fail "TIMEZONE='${TIMEZONE}' is not a zone this machine knows."
fi
if [[ "$TIMEZONE" == "$CURRENT_TZ" ]]; then
echo " keeping ${CURRENT_TZ}, nothing to do"
SUMMARY+=("Timezone: already ${CURRENT_TZ}")
else
echo " to set: ${TIMEZONE}"
if confirm "Proceed?"; then
timezone_set "$TIMEZONE"
ok "Timezone set to ${TIMEZONE}"
SUMMARY+=("Timezone: ${TIMEZONE}")
else
warn "skipped by request"
SUMMARY+=("Timezone: SKIPPED by request — left at ${CURRENT_TZ:-unknown}")
fi
fi
step_ok
fi
# =============================================================================
# 7. Swap
# =============================================================================
#
# Disk the kernel can park cold pages on when RAM fills, so a spike costs
# latency instead of a process. A `bun install` or a Docker build on a small
# machine is exactly the spike this is for.
step "Swap"
if ! skip; then
ACTIVE_SWAP_GB="$(swap_active_gb)"
WANT_SWAP_GB="$(swap_recommended_gb)"
SWAPPINESS="$(swappiness_for_role)"
CURRENT_SWAPPINESS="$(sysctl -n vm.swappiness 2>/dev/null || echo unknown)"
echo ""
info "Swap — overflow space so a memory spike costs speed rather than a process"
echo " RAM: $(ram_gb)G"
echo " active swap: ${ACTIVE_SWAP_GB}G"
echo " swappiness: ${CURRENT_SWAPPINESS} -> ${SWAPPINESS} (${MACHINE_ROLE})"
if [[ "$IS_WSL" == true ]]; then
# WSL2 runs its own managed swap inside the VM; a swapfile here is wasted
# disk and is not what the kernel would use anyway.
echo " WSL manages its own swap — leaving it alone"
SUMMARY+=("Swap: left to WSL")
elif ((ACTIVE_SWAP_GB > 0)); then
echo " already has ${ACTIVE_SWAP_GB}G of swap, leaving it alone"
if [[ "$CURRENT_SWAPPINESS" != "$SWAPPINESS" ]] && confirm "Set swappiness to ${SWAPPINESS}?"; then
swappiness_set "$SWAPPINESS"
ok "swappiness set to ${SWAPPINESS}"
SUMMARY+=("Swap: kept ${ACTIVE_SWAP_GB}G, swappiness ${SWAPPINESS}")
else
SUMMARY+=("Swap: kept ${ACTIVE_SWAP_GB}G")
fi
elif ((WANT_SWAP_GB == 0)); then
# Capped to nothing by the disk check rather than by choice.
warn "not enough free disk to add swap safely — $(disk_free_gb)G free"
SUMMARY+=("Swap: none added, disk too full")
else
echo " to create: ${WANT_SWAP_GB}G at ${SWAPFILE} ($(disk_free_gb)G free now)"
if confirm "Proceed?"; then
swap_create "$WANT_SWAP_GB"
swappiness_set "$SWAPPINESS"
ok "${WANT_SWAP_GB}G swap active, swappiness ${SWAPPINESS}"
SUMMARY+=("Swap: ${WANT_SWAP_GB}G created, swappiness ${SWAPPINESS}")
else
warn "skipped by request"
SUMMARY+=("Swap: SKIPPED by request")
fi
fi
step_ok
fi
# =============================================================================
# 8. Emergency disk ballast
# =============================================================================
#
# Always offered, whatever the role — the role only decides which way the
# recommendation points.
step "Emergency disk ballast"
if ! skip; then
echo ""
info "Emergency disk ballast — a reserve you can burn when the disk fills up"
echo " A junk file holding no data, sized at 10% of free disk. A root cron"
echo " checks every 10 minutes and deletes it if free space drops below"
echo " ${BALLAST_THRESHOLD}%, so you get room to log in and clean up instead of meeting a"
echo " wedged machine — Docker, journald and postgres all misbehave badly"
echo " at 100% full, and not all of them recover on their own."
echo " It is a one-shot valve: once spent, run this again to recreate it."
echo ""
if is_server; then
echo " recommended for ${MACHINE_ROLE} — a full disk on an unattended box is the bad case"
else
echo " less useful on ${MACHINE_ROLE} — you are sitting at this machine and will notice"
fi
if ballast_exists; then
echo " already present: $(ballast_size_human) at ${BALLAST_FILE}"
SUMMARY+=("Disk ballast: already present ($(ballast_size_human))")
else
BALLAST_MB="$(ballast_size_mb)"
echo " to create: ${BALLAST_MB}M at ${BALLAST_FILE}"
if confirm "Create it?"; then
ballast_create "$BALLAST_MB"
ballast_install_checker
ok "ballast $(ballast_size_human), checker at ${BALLAST_CHECKER}"
SUMMARY+=("Disk ballast: $(ballast_size_human), checked every 10 min")
else
warn "skipped by request"
SUMMARY+=("Disk ballast: SKIPPED by request")
fi
fi
step_ok
fi
# =============================================================================
# 9. earlyoom
# =============================================================================
step "earlyoom"
if ! skip; then
echo ""
info "earlyoom — keeps the machine reachable when it runs out of memory"
echo " The kernel's own OOM killer waits until an allocation actually"
echo " fails, and by then the machine has usually spent minutes thrashing:"
echo " unresponsive, ssh refusing to connect, nothing to do but reset it."
echo " earlyoom watches free memory and kills the biggest consumer while"
echo " there is still enough left to stay logged in."
echo ""
echo " This is what happens after swap runs out, so the two go together."
if earlyoom_is_active; then
echo " already installed and running"
SUMMARY+=("earlyoom: already running")
elif confirm "Install it?"; then
earlyoom_install
if earlyoom_is_active; then
ok "earlyoom running"
SUMMARY+=("earlyoom: installed and running")
else
warn "earlyoom installed but not running — check: systemctl status earlyoom"
SUMMARY+=("earlyoom: installed, not running")
fi
else
warn "skipped by request"
SUMMARY+=("earlyoom: SKIPPED by request")
fi
step_ok
fi
# =============================================================================
# 10. inotify watch limit
# =============================================================================
step "inotify watch limit"
if ! skip; then
CURRENT_WATCHES="$(inotify_current_watches)"
echo ""
info "inotify watch limit — how many files can be watched for changes at once"
echo " A single file watcher walking a project with node_modules in it can"
echo " exhaust the stock limit on its own, and every watcher on the machine"
echo " draws from the same pool. The failure is silent: nothing errors, the"
echo " watcher just stops noticing changes. Hot reload goes quiet, a build"
echo " stops rebuilding, and the reason is never on screen."
echo ""
echo " current: ${CURRENT_WATCHES}"
echo " to set: ${INOTIFY_WATCHES}"
if is_role dev; then
echo " recommended on dev — editors, bun --watch and vite are all watchers"
else
echo " less pressing on ${MACHINE_ROLE}, but anything running bun --watch or"
echo " serving a file browser is a watcher too"
fi
if ((CURRENT_WATCHES >= INOTIFY_WATCHES)); then
echo " already at or above that, nothing to do"
SUMMARY+=("inotify watches: already ${CURRENT_WATCHES}")
elif confirm "Raise it?"; then
inotify_raise
ok "inotify watches raised to $(inotify_current_watches)"
SUMMARY+=("inotify watches: raised to ${INOTIFY_WATCHES}")
else
warn "skipped by request"
SUMMARY+=("inotify watches: SKIPPED by request — left at ${CURRENT_WATCHES}")
fi
step_ok
fi
# =============================================================================
# NOT PORTED YET
# =============================================================================
#
# Sections still to move across from scripts/setup-old/setup-ubuntu.sh, in order:
#
# auto-suspend · boot-hang fix · user creation ·
# ssh keys · ssh hardening · dns · static ip · fail2ban · unattended-upgrades ·
# git config · docker · zsh + prompt · tailscale · neovim · js runtimes ·
# dev tools · ufw · zshrc
#
# Each arrives as its own commit. Delete this block when the list is empty.
# =============================================================================
# 23. Summary
# =============================================================================
echo ""
echo ""
if [[ ${#ERRORS[@]} -gt 0 ]]; then
echo -e "${YELLOW}╔══════════════════════════════════════════════════╗${NC}"
echo -e "${YELLOW}║ Setup Complete (with warnings) ║${NC}"
echo -e "${YELLOW}╚══════════════════════════════════════════════════╝${NC}"
else
echo -e "${GREEN}╔══════════════════════════════════════════════════╗${NC}"
echo -e "${GREEN}║ Setup Complete ║${NC}"
echo -e "${GREEN}╚══════════════════════════════════════════════════╝${NC}"
fi
echo ""
echo -e "${BOLD} What was done:${NC}"
for item in "${SUMMARY[@]}"; do
echo -e " ${GREEN}+${NC} $item"
done
if [[ ${#ERRORS[@]} -gt 0 ]]; then
echo ""
echo -e "${BOLD} Non-critical issues:${NC}"
for err in "${ERRORS[@]}"; do
echo -e " ${YELLOW}!${NC} $err"
done
fi
echo ""
echo -e "${BOLD} Machine:${NC}"
echo " System: $OS_NAME ($ARCH)"
echo " Role: $MACHINE_ROLE"
echo " User: $USERNAME"
echo " Home: $USER_HOME"
[[ -n "${TS_IP:-}" && "$TS_IP" != "unknown" ]] && echo " Tailscale: $TS_IP"
echo ""
# Clean up progress file on success
rm -f "$PROGRESS_FILE"