# syntax=docker/dockerfile:1.24.0@sha256:87999aa3d42bdc6bea60565083ee17e86d1f3339802f543c0d03998580f9cb89
# =============================================================================
# Multi-architecture build
# =============================================================================
# Supported MikroTik Container architectures (build with `docker buildx`):
#   linux/amd64        x86 / CHR
#   linux/arm64        RB5009, CCR2xxx, hAP ax3, L009, Chateau (most modern)
#   linux/arm/v7       ARMv7: hAP ac2, RB3011, RB4011, RB1100AHx4
#
# NOT supported here: ARMv5 (hEX Refresh / hAP ax S, EN7562CT CPU). ARMv5 has
# no Alpine/musl base image, so it cannot use the musl + scratch design below;
# it would need a glibc (Debian) base and produces a much larger image. See
# README for details if you need it.
#
# Both the Go (Tailscale) stage and the C (busybox) stage cross-compile: they
# always run NATIVELY on the build host ($BUILDPLATFORM) and produce binaries
# for $TARGETPLATFORM. This eliminates QEMU emulation entirely from the build,
# which is the main source of slowness in multi-arch builds. Only the final
# scratch stage pulls in the target-arch-specific layers (CA certs, busybox
# rootfs) which are just file copies with no emulated execution.
#
# Cross-compilation for C (busybox) is provided by tonistiigi/xx, which
# configures clang+lld as a cross-compiler and installs musl headers for the
# target arch via xx-apk.

# =============================================================================
# xx: Dockerfile cross-compilation helpers (provides xx-clang, xx-apk, etc.)
# =============================================================================
# renovate: datasource=docker depName=tonistiigi/xx versioning=docker
FROM --platform=$BUILDPLATFORM tonistiigi/xx:1.9.0@sha256:c64defb9ed5a91eacb37f96ccc3d4cd72521c4bd18d5442905b95e2226b0e707 AS xx

# =============================================================================
# Stage 1: Build Tailscale combined binary (cross-compiled, runs natively)
# =============================================================================
FROM --platform=$BUILDPLATFORM golang:1.26.4-alpine@sha256:f1ddd9fe14fffc091dd98cb4bfa999f32c5fc77d2f2305ea9f0e2595c5437c14 AS builder

# renovate: datasource=github-releases depName=tailscale packageName=tailscale/tailscale versioning=semver
ARG TAILSCALE_VERSION=v1.98.5

# Provided automatically by buildx for the target platform.
ARG TARGETOS
ARG TARGETARCH
ARG TARGETVARIANT

RUN apk add --no-cache \
    git \
    upx \
    ca-certificates

# Clone the exact release tag (no full history)
RUN git clone --depth 1 --branch ${TAILSCALE_VERSION} \
    https://github.com/tailscale/tailscale.git /src/tailscale

WORKDIR /src/tailscale

# Inject a stderr verbosity filter into the tailscaled package.
#
# With logtail compiled out (ts_omit_logtail), tailscaled never installs
# logpolicy (see `if buildfeatures.HasLogTail` in cmd/tailscaled/tailscaled.go),
# so log output goes raw to stderr: the [v1]/[v2] verbosity tags embedded in
# messages are neither parsed nor filtered, and --verbose has NO effect. The
# result is constant log spam in the RouterOS container log (filter
# "Accept: TCP" verdicts, "netcheck: [v1] report", "wg: [v2]" handshakes and
# keepalives) — see tailscale/tailscale#12158 and #1548.
#
# The injected file (build-tagged ts_omit_logtail, so it's a no-op if logtail
# is ever re-enabled) registers a log writer in init() that drops lines
# carrying a [v1]+ tag, restoring the equivalent of logtail's StderrLevel=0
# default. Setting TS_LOG_VERBOSITY=1 (or higher) in the container environment
# disables the filter at runtime for debugging — no rebuild needed.
COPY patches/stderr_verbosity_filter.go cmd/tailscaled/

# Build a minimal combined binary (tailscale CLI + tailscaled daemon in one file).
#
# Tag strategy — ALLOWLIST, not blocklist:
#   1. cmd/featuretags --min --add=osrouter generates the full ts_omit_* set
#      (identical to build_dist.sh --extra-small), omitting every optional feature.
#   2. We pipe that through sed to REMOVE the ts_omit_ tags for the features
#      we explicitly want, leaving everything else omitted.
#   3. We prepend ts_include_cli (combined daemon+CLI binary).
#
# This means any NEW ts_omit_* tag added in a future Tailscale release will
# automatically be omitted — we only get features we consciously opt into.
#
# Features opted in (removed from the omit list):
#   advertiseexitnode   — run as exit node for the tailnet
#   advertiseroutes     — advertise LAN subnets to the tailnet
#   useexitnode         — route router's own traffic via a remote exit node
#   useroutes           — accept routes advertised by other tailnet nodes
#   dns                 — MagicDNS; configure MikroTik DNS to forward
#                         *.ts.net → 100.100.100.100; use --no-dns daemon
#                         flag to skip writing /etc/resolv.conf
#   portmapper          — NAT-PMP / PCP / UPnP to punch through upstream NAT
#   listenrawdisco      — raw sockets for more robust disco/NAT-traversal
#   health              — health subsystem required by 'tailscale status'
#   iptables            — Linux iptables support for routing rules
#   unixsocketidentity  — REQUIRED for the CLI to talk to the daemon. Without it,
#                         the localapi can't verify a request arrived over the
#                         trusted unix socket, so PermitRead/PermitWrite are
#                         always false and EVERY CLI call (status, up, set, ...)
#                         returns "access denied" (tailscale/tailscale#17873).
#   ipnbus              — IPN bus watch. Without it, 'tailscale up' cannot wait
#                         for completion: it fires config at the daemon and
#                         returns immediately ("built with ts_omit_ipnbus; not
#                         waiting for completion") WITHOUT printing the auth URL
#                         or confirming success. Including it makes interactive
#                         'up' behave normally (blocks, prints login URL).
#   netstack            — gVisor userspace network stack. Counter-intuitively
#                         REQUIRED even though the router uses a real kernel TUN
#                         (NOT --tun=userspace-networking). In v1.98.5 the
#                         100.100.100.100:53 MagicDNS listener is served ONLY by
#                         netstack's handleLocalPackets, installed via
#                         PreFilterPacketOutboundToWireGuardNetstackIntercept.
#                         The non-netstack "engine" interceptor that the wrap.go
#                         comments claim handles quad-100 "if netstack is not
#                         installed" does NOT actually do so on Linux (its body
#                         only reflects loopback on darwin/ios/plan9, else
#                         Accept). So with ts_omit_netstack, NOTHING absorbs
#                         packets to 100.100.100.100: queries fall through to
#                         WireGuard, no peer owns that IP, and even tailnet-name
#                         resolution (and 'ping host.tailnet.ts.net') times out.
#                         The 'dns' tag links the resolver but nothing routes
#                         packets to it without netstack — the two tags are
#                         independent (dns has no Dep on netstack). Omitting
#                         netstack ALSO triggered a panic("unreachable") in
#                         net/tstun.invertGSOChecksum on the exit-node inject
#                         path (HasNetstack=const false made the guard always
#                         panic); enabling netstack makes that guard dead code,
#                         fixing the crash as a side effect. Cost (arm64, vs a
#                         netstack-omitted build): ~+0.5 MB extracted on flash
#                         and ~+2.3 MB resident RAM after UPX decompression —
#                         measured, acceptable for a 16 MB-flash router.
#   gro                 — Generic Receive Offload (perf). Depends on netstack;
#                         pulled in with it. Small, and improves throughput on
#                         the netstack DNS/inject path.
#   peerapiserver       — REQUIRED to be a functional exit node. In v1.98.5
#                         'advertiseexitnode' DECLARES a dependency on
#                         peerapiserver (featuretags.go Deps, "to run the ExitDNS
#                         server"), but this build's allowlist works by stripping
#                         individual ts_omit_ tags and does NOT re-resolve Deps —
#                         so featuretags --min still emitted ts_omit_peerapiserver
#                         and our advertiseexitnode opt-in alone left it omitted.
#                         peerapiserver gates the entire PeerAPI HTTP server,
#                         including the /dns-query DoH endpoint (peerapi.go,
#                         guarded by buildfeatures.HasPeerAPIServer). Without it
#                         initPeerAPIListenerLocked() returns early: the node
#                         never advertises the PeerAPIDNS service, so exit-node
#                         CLIENTS' exitNodeCanProxyDNS(thisNode) returns false.
#                         With no tailnet global nameserver configured, the
#                         client's resolver then has an empty Routes["."] and
#                         returns an INSTANT authoritative SERVFAIL locally
#                         (forwarder.go servfailResponse, aa=1, 0 ms, no I/O) —
#                         i.e. devices using this router as their exit node could
#                         not resolve PUBLIC names. Including peerapiserver makes
#                         the node serve the exit-node DoH DNS proxy, so clients
#                         get public DNS automatically (the normal exit-node
#                         behavior) with no tailnet DNS config required.
#                         peerapiserver has NO Deps and pulls in no large
#                         subsystems — a small addition. (outboundproxy is NOT
#                         needed for this and stays omitted.)
#
# Everything else remains omitted, including (rationale):
#   clientupdate  — DELIBERATELY removed. The built-in updater would download
#                   the FULL official upstream tailscale binary (tens of MB,
#                   with all features) directly onto the device, defeating the
#                   entire point of this minimal build and risking filling the
#                   16 MB flash. It also can't update a binary baked into a
#                   read-only container image. Updates are instead delivered by
#                   rebuilding/republishing this image (CI) and pulling the new
#                   image only when it actually changed (see the RouterOS
#                   update script). This keeps the on-device footprint minimal
#                   and the update path controlled, reproducible, and flash-safe.
#   cachenetmap   — DELIBERATELY omitted. It ONLY persists the netmap to disk so
#                   the node can come online from the last-known config after a
#                   COLD START while the control plane is simultaneously
#                   unreachable. The in-memory netmap is NOT gated by this tag:
#                   a running daemon that loses its control connection keeps its
#                   map and can still reach known peers (data path is direct
#                   WireGuard/DERP, not via control). The only loss is the narrow
#                   reboot-during-control-outage case. In exchange we avoid disk
#                   writes on every netmap delta (frequent on busy tailnets),
#                   which is exactly the flash wear we want to avoid.
#   logtail       — no persistent log writes to flash; also pass
#                   --no-logs-no-support at runtime
#   ssh           — not needed; access via MikroTik SSH + tailscale CLI
#   all GUI/desktop/cloud/k8s features — irrelevant for a headless router
#
# NOTE: netstack/gro are NOT in this omit list — see the opted-in section above
# for why MagicDNS quad-100 serving structurally requires them in v1.98.5.

RUN mkdir -p /out && \
    ALL_OMIT=$(GOOS= GOARCH= go run ./cmd/featuretags --min --add=osrouter) && \
    TAGS=$(echo "ts_include_cli,${ALL_OMIT}" | \
        sed \
          -e 's/ts_omit_advertiseexitnode,\{0,1\}//g' \
          -e 's/ts_omit_advertiseroutes,\{0,1\}//g' \
          -e 's/ts_omit_useexitnode,\{0,1\}//g' \
          -e 's/ts_omit_useroutes,\{0,1\}//g' \
          -e 's/ts_omit_dns,\{0,1\}//g' \
          -e 's/ts_omit_portmapper,\{0,1\}//g' \
          -e 's/ts_omit_listenrawdisco,\{0,1\}//g' \
          -e 's/ts_omit_health,\{0,1\}//g' \
          -e 's/ts_omit_iptables,\{0,1\}//g' \
          -e 's/ts_omit_unixsocketidentity,\{0,1\}//g' \
          -e 's/ts_omit_ipnbus,\{0,1\}//g' \
          -e 's/ts_omit_netstack,\{0,1\}//g' \
          -e 's/ts_omit_gro,\{0,1\}//g' \
          -e 's/ts_omit_peerapiserver,\{0,1\}//g' \
          -e 's/,$//' \
    ) && \
    echo "Build tags: ${TAGS}" && \
    # Map Docker's TARGETARCH/TARGETVARIANT to Go's GOARCH/GOARM.
    # For arm/v7 -> GOARM=7 (hardfloat). Other arches leave GOARM unset.
    GOARM="" && \
    if [ "${TARGETARCH}" = "arm" ]; then \
      case "${TARGETVARIANT}" in \
        v7) GOARM=7 ;; \
        v6) GOARM=6 ;; \
        v5) GOARM=5 ;; \
        *)  GOARM=7 ;; \
      esac; \
    fi && \
    echo "Cross-compiling: GOOS=${TARGETOS:-linux} GOARCH=${TARGETARCH} GOARM=${GOARM}" && \
    CGO_ENABLED=0 GOOS=${TARGETOS:-linux} GOARCH=${TARGETARCH} GOARM=${GOARM} \
    go build \
      -tags "${TAGS}" \
      -gcflags="all=-l" \
      -ldflags="-s -w" \
      -trimpath \
      -o /out/tailscale.combined \
      ./cmd/tailscaled

# Compress with UPX LZMA.
# Expected: ~14 MB raw → ~3.8 MB compressed (with -gcflags=all=-l)
RUN upx --lzma --best /out/tailscale.combined

# Lay out the final /usr/local/bin HERE (binary + argv[0] symlinks) so the final
# stage can bring it in with a SINGLE COPY layer. Creating the symlinks with a
# `RUN` in the final scratch stage instead would force overlayfs to copy-up the
# whole directory — duplicating the ~3 MB binary into another layer and roughly
# doubling the extracted on-disk size on RouterOS (overlay layers are extracted
# separately). Building it in one place keeps it to one copy.
RUN mkdir -p /out/usrlocalbin && \
    mv /out/tailscale.combined /out/usrlocalbin/tailscale.combined && \
    ln -s /usr/local/bin/tailscale.combined /out/usrlocalbin/tailscale && \
    ln -s /usr/local/bin/tailscale.combined /out/usrlocalbin/tailscaled

# Entrypoint wrapper: enable IP forwarding inside the container's network
# namespace, then exec tailscaled. tailscaled does NOT reliably enable IPv6
# forwarding itself in a container netns ("IPv6 forwarding is disabled" warning),
# which silently breaks advertised IPv6 subnet routes. The sysctls ARE writable
# from inside a RouterOS container, so we set both here. Written in the builder
# stage so it ships in the same single /usr/local/bin COPY layer (preserves the
# overlayfs single-copy property). `exec` keeps tailscaled as PID 1.
RUN printf '%s\n' \
      '#!/bin/sh' \
      '# Enable IPv4/IPv6 forwarding. Required for advertised subnet routes and' \
      '# exit-node functionality.' \
      'for f in /proc/sys/net/ipv4/ip_forward /proc/sys/net/ipv6/conf/all/forwarding; do' \
      '  if [ -w "$f" ]; then echo 1 > "$f" 2>/dev/null || echo "warn: could not write $f"; fi' \
      'done' \
      'exec /usr/local/bin/tailscaled "$@"' \
      > /out/usrlocalbin/entrypoint.sh && \
    chmod +x /out/usrlocalbin/entrypoint.sh

# =============================================================================
# Stage 2: Custom minimal busybox (cross-compiled, runs natively on build host)
# =============================================================================
# The official busybox:musl image ships all ~404 applets at ~1.24 MB. For a
# debug shell on a flash-constrained router we only need ~100 applets, so we
# build a static busybox from source with a curated applet set, then UPX it
# down to ~230 kB on disk.
#
# UPX is normally dangerous with busybox: the ash shell's standalone applet
# dispatch re-execs /proc/self/exe, which UPX breaks, so typed commands fail
# (https://github.com/upx/upx/issues/248, closed as "invalid"). We sidestep
# this by building WITHOUT the standalone/nofork features (see
# busybox-applets.config) and providing an explicit /bin/<applet> symlink
# farm. Commands then resolve via the ordinary PATH -> symlink -> argv[0]
# dispatch, which works fine under UPX. The cost is a fork+exec per command,
# acceptable for an occasional debug shell. RouterOS stores the EXTRACTED
# rootfs on disk (overlayfs), so the ~190 kB UPX saving is real on-disk space.
#
# This stage runs NATIVELY on the build host (--platform=$BUILDPLATFORM) and
# cross-compiles busybox for the target architecture using clang+lld via the
# tonistiigi/xx helpers. This eliminates QEMU emulation from the busybox build,
# which was the main source of slowness for arm64/arm/v7 targets.
#
# Cross-compilation setup:
#   - xx-apk installs musl-dev and linux-headers for the TARGET arch under
#     /<triple> (a secondary sysroot), while clang/lld/upx/make stay native.
#   - xx-clang --setup-target-triple creates <triple>-clang / <triple>-cc
#     aliases in PATH that busybox's Makefile picks up via CROSS_COMPILE.
#   - Busybox make receives:
#       CROSS_COMPILE=<triple>-   → picks up <triple>-clang (from xx aliases)
#       CC=clang                  → use clang (aliased target via CROSS_COMPILE)
#       HOSTCC=gcc                → compile host helper tools with native gcc
#   - upx (native x86_64 binary) can compress target-arch binaries since UPX
#     operates on the ELF file format regardless of the target ISA.
#
# Applet symlink probing: for native-arch builds the probe runs directly;
# for cross-compiled binaries we use QEMU user-mode emulation (from binfmt)
# only for this one lightweight probe step (busybox --help per applet), not
# for the compile itself. The probe can alternatively be skipped by using
# a pre-enumerated applet list, but the current approach is simpler.
FROM --platform=$BUILDPLATFORM alpine:3.24.1@sha256:28bd5fe8b56d1bd048e5babf5b10710ebe0bae67db86916198a6eec434943f8b AS busybox

# Copy xx cross-compilation helpers (xx-clang, xx-apk, xx-info, etc.)
COPY --from=xx / /

# renovate: datasource=docker depName=busybox versioning=docker
ARG BUSYBOX_VERSION=1.38.0

# Target platform ARGs (provided automatically by buildx).
ARG TARGETPLATFORM
ARG TARGETARCH
ARG TARGETVARIANT

# Native build tools (clang/lld for cross-compiling; gcc/make/upx run natively).
# xx-apk installs the target-arch sysroot: musl-dev (C library headers + CRT),
# gcc (provides crtbeginS.o/crtendS.o and libgcc needed by clang on Alpine),
# and linux-headers (required by busybox for <linux/*.h> / <net/*.h>).
RUN apk add --no-cache \
        clang \
        lld \
        llvm \
        gcc \
        make \
        wget \
        bzip2 \
        perl \
        upx && \
    xx-apk add --no-cache musl-dev gcc linux-headers

RUN wget -q https://busybox.net/downloads/busybox-${BUSYBOX_VERSION}.tar.bz2 \
 && tar xf busybox-${BUSYBOX_VERSION}.tar.bz2
WORKDIR /busybox-${BUSYBOX_VERSION}

# allnoconfig = every feature OFF; then enable only the curated applet set.
COPY busybox-applets.config /tmp/applets.config
# Set up xx cross-compiler aliases (<triple>-clang, <triple>-cc, etc.) and
# build busybox.
#
# Key make variables:
#   ARCH           — busybox ARCH; must match the cross-target, not the build
#                    host. busybox's Makefile would otherwise read SUBARCH from
#                    `uname -m` (the BUILD host's arch) which is wrong when
#                    cross-compiling. We map TARGETARCH to busybox's arch name.
#                    busybox uses -include arch/$(ARCH)/Makefile; missing arch
#                    dirs are silently ignored, so any value is safe.
#   CC             — busybox defaults to $(CROSS_COMPILE)gcc. We override CC to
#                    the full <triple>-clang path so it resolves to the xx alias
#                    (which sets --target and --sysroot for the cross-compiler).
#                    Setting CC= avoids needing a <triple>-gcc symlink.
#   HOSTCC         — native compiler for host-side build tools (scripts/kconfig,
#                    gen_build_files, etc.); must NOT be the cross-compiler.
#   SKIP_STRIP     — defer stripping to after symlink probing (we strip below
#                    with llvm-strip, which handles any target ELF arch).
RUN xx-clang --setup-target-triple && \
    CROSS=$(xx-info triple) && \
    # Map TARGETARCH to the busybox ARCH value.
    case "${TARGETARCH}" in \
      amd64)  BUSYBOX_ARCH=x86_64 ;; \
      arm64)  BUSYBOX_ARCH=aarch64 ;; \
      arm)    BUSYBOX_ARCH=arm ;; \
      *)      BUSYBOX_ARCH=${TARGETARCH} ;; \
    esac && \
    make allnoconfig ARCH="${BUSYBOX_ARCH}" && \
    while read -r sym; do \
      case "$sym" in ''|\#*) continue ;; esac; \
      if grep -q "^# CONFIG_${sym} is not set" .config; then \
        sed -i "s/^# CONFIG_${sym} is not set/CONFIG_${sym}=y/" .config; \
      elif ! grep -q "^CONFIG_${sym}=y" .config; then \
        echo "CONFIG_${sym}=y" >> .config; \
      fi; \
    done < /tmp/applets.config && \
    yes "" | make oldconfig ARCH="${BUSYBOX_ARCH}" >/dev/null 2>&1 && \
    make -j"$(nproc)" \
        ARCH="${BUSYBOX_ARCH}" \
        CROSS_COMPILE="${CROSS}-" \
        CC="${CROSS}-clang" \
        HOSTCC=gcc \
        SKIP_STRIP=y \
        >/dev/null 2>&1 && \
    llvm-strip busybox

# Lay out a minimal rootfs with busybox + an applet symlink per applet.
# Symlinks (argv[0] dispatch) are how busybox selects an applet and make the
# applets resolvable via $PATH from inside the shell. We derive the applet
# names from the build .config: a symbol is an applet if its lowercase name
# resolves to a runnable applet (busybox returns "applet not found" on stderr
# for non-applet symbols like FEATURE_* / STATIC, which we filter out).
# We generate symlinks from the UNCOMPRESSED binary (so the probe is reliable),
# then UPX-compress the binary in place afterwards.
#
# Note: probing cross-compiled binaries requires binfmt/QEMU user-mode. This
# is only a lightweight per-applet help-flag check, not a full emulated build.
# If QEMU is unavailable in CI, replace the probe with a static applet list.
RUN mkdir -p /rootfs/bin && \
    grep '^CONFIG_.*=y' .config \
      | sed -e 's/^CONFIG_//' -e 's/=y$//' \
      | tr 'A-Z' 'a-z' \
      | while read -r app; do \
          if ! ./busybox "$app" --help 2>&1 | grep -q "applet not found"; then \
            ln -sf /bin/busybox /rootfs/bin/"$app"; \
          fi; \
        done && \
    ln -sf /bin/busybox /rootfs/bin/sh && \
    echo "Applet symlinks created: $(ls /rootfs/bin | wc -l)" && \
    upx --lzma --best busybox && \
    cp busybox /rootfs/bin/busybox

# =============================================================================
# Stage 3: Final runtime image
# =============================================================================
FROM scratch

# Release version (the git tag, e.g. v1.98.3-mt.1), injected by CI at build
# time. This is the value the MikroTik update cronjob compares against the
# registry to decide whether to recreate the container: it changes ONLY on a
# meaningful release (Tailscale bump -> mt.1, or a manual mt.N), never on a
# build-system-only rebuild. Defaults to "dev" for local builds.
ARG OCI_VERSION=dev

# OCI image annotations. org.opencontainers.image.version is the canonical place
# for the release version and is what the router reads back from the registry.
LABEL org.opencontainers.image.title="mikrotik-tailscale" \
      org.opencontainers.image.description="Minimal Tailscale image for MikroTik RouterOS Container" \
      org.opencontainers.image.source="https://gitea.lumpiasty.xyz/lumpiasty/mikrotik-tailscale" \
      org.opencontainers.image.version="${OCI_VERSION}"

# Custom static busybox + applet symlinks (provides /bin/sh and utilities)
COPY --from=busybox /rootfs/ /

# CA certificates (needed to reach Tailscale coordination server)
COPY --from=builder /etc/ssl/certs/ca-certificates.crt /etc/ssl/certs/

# Combined Tailscale binary + its argv[0] symlinks, in a single layer (built in
# the builder stage to avoid overlayfs copy-up duplicating the binary — see the
# builder stage comment).
COPY --from=builder /out/usrlocalbin/ /usr/local/bin/

# Ensure /usr/local/bin and busybox dirs are on PATH for interactive shells
ENV PATH=/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin

# -----------------------------------------------------------------------------
# Volume layout (to be created by deploy script):
#
#   /var/lib/tailscale       — persistent state (authkey, node identity)
#                              → bind-mount to MikroTik disk storage
#                              → survives reboots, written infrequently (only on
#                                auth / key rotation / prefs change)
#                              → netmap is NOT cached to disk (cachenetmap is
#                                omitted), so this dir sees no per-netmap writes
#
#   /var/run/tailscale       — runtime socket dir
#                              → tmpfs, lost on reboot (expected)
# -----------------------------------------------------------------------------
VOLUME ["/var/lib/tailscale"]

# entrypoint.sh enables IP forwarding (incl. IPv6) in the container netns, then
# exec's tailscaled with the CMD flags below as its arguments.
ENTRYPOINT ["/usr/local/bin/entrypoint.sh"]

# Default flags:
#   --no-logs-no-support  disables logtail uploads (logtail binary code is
#                         omitted, but the flag also suppresses any remaining
#                         log buffering and prevents the daemon from trying
#                         to write log files)
#   --state               persistent node identity / authkey storage
#   --socket              CLI communication socket (on tmpfs)
#   --statedir            var root (derpmap cache, certs, etc.); no netmap
#                         disk cache here since cachenetmap is omitted
CMD ["--no-logs-no-support", \
     "--state=/var/lib/tailscale/tailscaled.state", \
     "--socket=/var/run/tailscale/tailscaled.sock", \
     "--statedir=/var/lib/tailscale"]
