archy/scripts/container-doctor.sh

#!/bin/bash
#
# Container Doctor — diagnose and fix common container health issues
#
# Usage:
#   sudo ./scripts/container-doctor.sh              # Run locally on node
#   ./scripts/container-doctor.sh user@host          # Run remotely via SSH
#
# Fixes:
#   1. Stale podman ps/stats processes (>10 = pileup)
#   2. Orphaned conmon/crun processes holding ports
#   3. System tor conflicting with container tor
#   4. Tor hidden service directory permissions (must be 700)
#   5. SearXNG read-only root / cap-drop ALL
#   6. Bitcoin Knots prune+txindex conflict
#   7. Containers stuck with exit code 127 (binary not found)
#   8. Stopped core containers (rootless restart policy workaround)
#
# Safe to run multiple times (idempotent). Never blocks deploy (exit 0 always).
#

set -o pipefail

# Source pinned image versions (single source of truth)
SCRIPT_DIR="$(cd "$(dirname "$0")" && pwd)"
[ -f "$SCRIPT_DIR/image-versions.sh" ] && . "$SCRIPT_DIR/image-versions.sh"

FIXES_APPLIED=0
CHECKS_PASSED=0
FIX_NAMES=()

log() { echo "[$(date +%H:%M:%S)] DOCTOR: $*"; }

run_fix() {
    local name="$1"
    shift
    if "$@"; then
        FIXES_APPLIED=$((FIXES_APPLIED + 1))
        FIX_NAMES+=("$name")
    else
        CHECKS_PASSED=$((CHECKS_PASSED + 1))
    fi
}

# ── Fix 1: Stale podman processes ────────────────────────────
fix_stale_podman() {
    local count
    count=$(pgrep -f "podman (ps|stats)" 2>/dev/null | wc -l)
    count=${count:-0}
    if [ "$count" -gt 10 ]; then
        log "Killing $count stale podman ps/stats processes"
        pkill -f "podman (ps|stats)" 2>/dev/null || true
        sleep 2
        local after
        after=$(pgrep -f "podman (ps|stats)" 2>/dev/null | wc -l)
        after=${after:-0}
        log "Reduced from $count to $after"
        return 0
    fi
    return 1
}

# ── Fix 2: Orphaned conmon holding ports ─────────────────────
fix_orphaned_conmon() {
    local fixed=false
    # Find conmon processes whose containers no longer exist
    local pids
    pids=$(pgrep -f "conmon.*--exit-command" 2>/dev/null || true)
    if [ -z "$pids" ]; then
        return 1
    fi

    # Doctor runs as root but containers are rootless under archipelago user.
    # Must check container existence using the rootless podman database.
    local PODMANCMD="sudo -u archipelago XDG_RUNTIME_DIR=/run/user/1000 podman"

    for pid in $pids; do
        # Extract container ID from conmon args
        local cid
        cid=$(tr '\0' ' ' < /proc/"$pid"/cmdline 2>/dev/null | grep -oP '(?<=-c )[a-f0-9]{64}' || true)
        if [ -z "$cid" ]; then
            continue
        fi
        # Check if container still exists in rootless podman
        if ! $PODMANCMD inspect "$cid" &>/dev/null; then
            local port_info
            port_info=$(ss -tlnp 2>/dev/null | grep "pid=$pid" | grep -oP ':\K\d+' | head -3 | tr '\n' ',' | sed 's/,$//')
            log "Killing orphaned conmon pid=$pid (ports: ${port_info:-none})"
            kill "$pid" 2>/dev/null || kill -9 "$pid" 2>/dev/null || true
            fixed=true
        fi
    done

    $fixed && return 0 || return 1
}

# ── Fix 3: Ensure system Tor is running (preferred over container) ──
fix_system_tor_conflict() {
    # System Tor is preferred over container Tor.
    # If archy-tor container exists, remove it and use system Tor instead.
    if podman ps -a --format '{{.Names}}' 2>/dev/null | grep -qE '^archy-tor$'; then
        podman stop archy-tor 2>/dev/null || true
        podman rm -f archy-tor 2>/dev/null || true
        log "Removed archy-tor container (system Tor is preferred)"
    fi

    # Ensure system Tor is enabled and running
    if command -v tor >/dev/null 2>&1; then
        if ! systemctl is-active tor@default >/dev/null 2>&1; then
            systemctl enable tor tor@default 2>/dev/null || true
            systemctl start tor tor@default 2>/dev/null || true
            log "Started system Tor"
            return 0
        fi
    fi

    return 1
}

# ── Fix 4: Tor hidden service permissions ────────────────────
fix_tor_permissions() {
    local fixed=false
    local tor_dirs=("/var/lib/archipelago/tor" "/var/lib/tor")

    for base in "${tor_dirs[@]}"; do
        if [ ! -d "$base" ]; then
            continue
        fi
        while IFS= read -r dir; do
            local perms
            perms=$(stat -c '%a' "$dir" 2>/dev/null)
            if [ "$perms" != "700" ]; then
                chmod 700 "$dir"
                log "Fixed permissions on $dir ($perms -> 700)"
                fixed=true
            fi
        done < <(find "$base" -maxdepth 1 -name "hidden_service_*" -type d 2>/dev/null)
    done

    # If we fixed permissions, restart system Tor to pick up the changes
    if $fixed; then
        systemctl restart tor@default 2>/dev/null || true
        return 0
    fi
    return 1
}

# ── Fix 5: SearXNG read-only / cap-drop ─────────────────────
fix_searxng() {
    if ! podman ps -a --format '{{.Names}}' 2>/dev/null | grep -q '^searxng$'; then
        return 1
    fi

    local state
    state=$(podman inspect searxng --format '{{.State.Status}}' 2>/dev/null || true)
    local readonly_root
    readonly_root=$(podman inspect searxng --format '{{.HostConfig.ReadonlyRootfs}}' 2>/dev/null || true)
    local cap_drop
    cap_drop=$(podman inspect searxng --format '{{.HostConfig.CapDrop}}' 2>/dev/null || true)

    # Fix if: exited, or has read-only root, or has cap-drop ALL
    local needs_fix=false
    if [ "$state" = "exited" ]; then
        needs_fix=true
    fi
    if [ "$readonly_root" = "true" ]; then
        needs_fix=true
    fi
    if [[ "$cap_drop" == *"ALL"* ]] || [[ "$cap_drop" == *"all"* ]]; then
        needs_fix=true
    fi

    if ! $needs_fix; then
        return 1
    fi

    log "Recreating SearXNG (readonly=$readonly_root, cap_drop=$cap_drop, state=$state)"

    # Get current port mapping
    local port
    port=$(podman inspect searxng --format '{{range $k,$v := .HostConfig.PortBindings}}{{$k}}={{range $v}}{{.HostPort}}{{end}}{{println}}{{end}}' 2>/dev/null | head -1)
    local host_port="${port##*=}"
    host_port="${host_port:-8888}"

    # Kill any stale conmon holding the port
    local conmon_pid
    conmon_pid=$(ss -tlnp 2>/dev/null | grep ":${host_port} " | grep -oP 'pid=\K\d+' | head -1)

    podman stop searxng 2>/dev/null || true
    podman rm -f searxng 2>/dev/null || true

    if [ -n "$conmon_pid" ]; then
        kill -9 "$conmon_pid" 2>/dev/null || true
        sleep 2
    fi

    podman run -d \
        --name searxng \
        --restart=unless-stopped \
        --security-opt=no-new-privileges:true \
        --tmpfs /tmp:rw,noexec,nosuid,size=256m \
        -v searxng-config:/etc/searxng:rw \
        -v searxng-cache:/var/cache/searxng:rw \
        -p "${host_port}:8080" \
        --memory=512m \
        "${SEARXNG_IMAGE}" 2>&1 || true

    log "SearXNG recreated (no readonly, no cap-drop ALL)"
    return 0
}

# ── Fix 6: Bitcoin Knots prune+txindex conflict ──────────────
fix_bitcoin_txindex() {
    if ! podman ps -a --format '{{.Names}}' 2>/dev/null | grep -q '^bitcoin-knots$'; then
        return 1
    fi

    # Check if bitcoin.conf has prune enabled
    local conf="/var/lib/archipelago/bitcoin/bitcoin.conf"
    if [ ! -f "$conf" ] || ! grep -q '^prune=' "$conf"; then
        return 1
    fi

    # Check if container args include txindex
    local cmd
    cmd=$(podman inspect bitcoin-knots --format '{{json .Config.Cmd}}' 2>/dev/null || true)
    if ! echo "$cmd" | grep -q "txindex"; then
        return 1
    fi

    log "Bitcoin Knots: prune+txindex conflict detected"

    # Get current config
    local image
    image=$(podman inspect bitcoin-knots --format '{{.ImageName}}' 2>/dev/null)
    local network
    network=$(podman inspect bitcoin-knots --format '{{.HostConfig.NetworkMode}}' 2>/dev/null)

    # Read per-installation RPC password
    local SECRETS_DIR="/var/lib/archipelago/secrets"
    local BTC_RPC_PASS="archipelago"
    if [ -f "$SECRETS_DIR/bitcoin-rpc-password" ]; then
        BTC_RPC_PASS=$(cat "$SECRETS_DIR/bitcoin-rpc-password")
    fi

    # Ensure bitcoin.conf has all RPC settings
    if ! grep -q 'rpcuser=' "$conf"; then
        cat > "$conf" <<BCONF
server=1
prune=550
rpcuser=archipelago
rpcpassword=$BTC_RPC_PASS
rpcbind=0.0.0.0
rpcallowip=127.0.0.1/32
rpcallowip=10.88.0.0/16
rpcport=8332
listen=1
printtoconsole=1
BCONF
        log "Updated bitcoin.conf with full RPC settings"
    fi

    # Remove stale txindex if present
    if [ -d "/var/lib/archipelago/bitcoin/indexes/txindex" ]; then
        find /var/lib/archipelago/bitcoin/indexes/txindex -type f -delete 2>/dev/null
        rmdir /var/lib/archipelago/bitcoin/indexes/txindex 2>/dev/null || true
        log "Removed stale txindex directory"
    fi

    # Recreate without txindex
    podman stop bitcoin-knots 2>/dev/null || true
    podman rm -f bitcoin-knots 2>/dev/null || true
    sleep 2

    # Kill stale conmon on port 8332/8333
    for p in 8332 8333; do
        local cpid
        cpid=$(ss -tlnp 2>/dev/null | grep ":${p} " | grep -oP 'pid=\K\d+' | head -1)
        if [ -n "$cpid" ]; then
            kill -9 "$cpid" 2>/dev/null || true
        fi
    done
    sleep 1

    local net_arg=""
    if [ -n "$network" ] && [ "$network" != "bridge" ] && [ "$network" != "host" ]; then
        net_arg="--network=$network"
    elif [ "$network" = "host" ]; then
        net_arg="--network=host"
    else
        net_arg="--network=archy-net"
    fi

    podman run -d \
        --name bitcoin-knots \
        --restart=always \
        $net_arg \
        -p 8332:8332 \
        -p 8333:8333 \
        -v /var/lib/archipelago/bitcoin:/home/bitcoin/.bitcoin \
        --memory=2g \
        --cap-drop=ALL \
        --cap-add=CHOWN \
        --cap-add=FOWNER \
        --cap-add=SETUID \
        --cap-add=SETGID \
        --cap-add=DAC_OVERRIDE \
        --security-opt=no-new-privileges:true \
        --health-cmd="bitcoin-cli -rpcuser=archipelago -rpcpassword=$BTC_RPC_PASS getblockchaininfo || exit 1" \
        --health-interval=30s \
        --health-retries=3 \
        "$image" 2>&1 || true

    log "Bitcoin Knots recreated without txindex (prune mode)"
    return 0
}

# ── Fix 7: Exit code 127 containers ─────────────────────────
fix_exit_127() {
    local containers
    containers=$(podman ps -a --format '{{.Names}} {{.Status}}' 2>/dev/null | grep 'Exited (127)' | awk '{print $1}' || true)

    if [ -z "$containers" ]; then
        return 1
    fi

    local fixed_names=()
    for name in $containers; do
        # Skip containers handled by other fixes
        if [ "$name" = "searxng" ]; then
            continue
        fi

        log "Container $name has exit code 127 — recreating"
        # Get image and create command for recreation
        local image
        image=$(podman inspect "$name" --format '{{.ImageName}}' 2>/dev/null || true)
        local create_cmd
        create_cmd=$(podman inspect "$name" --format '{{json .Config.CreateCommand}}' 2>/dev/null || true)

        podman rm -f "$name" 2>/dev/null || true

        if [ -n "$create_cmd" ] && [ "$create_cmd" != "null" ]; then
            # Re-run the original create command (strip the leading "podman" and "run")
            local recreate_args
            recreate_args=$(echo "$create_cmd" | python3 -c "
import json, sys
args = json.load(sys.stdin)
# Skip 'podman' and 'run', output the rest
print(' '.join(['\"' + a + '\"' if ' ' in a else a for a in args[2:]]))
" 2>/dev/null || true)

            if [ -n "$recreate_args" ]; then
                eval "podman run $recreate_args" 2>&1 || true
                fixed_names+=("$name")
                log "Recreated $name from original args"
            else
                fixed_names+=("$name(removed)")
                log "Removed $name — will be recreated on next deploy"
            fi
        else
            fixed_names+=("$name(removed)")
            log "Removed $name — will be recreated on next deploy"
        fi
    done

    [ ${#fixed_names[@]} -gt 0 ] && return 0 || return 1
}

# ── Fix 8: Rootless netns egress lost ────────────────────────
# Rootless podman uses pasta to give containers internet egress. If pasta's
# tap vanishes (host link flap, mount churn), the rootless-netns keeps inter-
# container traffic working but silently loses outbound. Bitcoin IBD stalls
# at 0 peers; package pulls fail. The only reliable repair is a stop-all/
# start-all cycle so pasta + aardvark-dns rebuild the netns from scratch.
fix_rootless_netns_egress() {
    local archi_uid
    archi_uid=$(id -u archipelago 2>/dev/null) || return 1

    # Locate the rootless-netns via aardvark-dns (it lives inside it).
    local aardvark_pid
    aardvark_pid=$(pgrep -U "$archi_uid" -f '^/usr/lib/podman/aardvark-dns' 2>/dev/null | head -1)
    [ -z "$aardvark_pid" ] && return 1   # no rootless network active

    # Host precheck: if the host itself can't reach the internet, no point
    # cycling containers — this is an upstream problem.
    if ! timeout 3 bash -c '</dev/tcp/1.1.1.1/443' 2>/dev/null; then
        return 1
    fi

    # Probe egress from inside the rootless-netns. One probe is noisy;
    # require two consecutive failures 10s apart to rule out transients.
    if timeout 3 nsenter -t "$aardvark_pid" -n bash -c '</dev/tcp/1.1.1.1/443' 2>/dev/null; then
        return 1   # first probe succeeded
    fi
    sleep 10
    aardvark_pid=$(pgrep -U "$archi_uid" -f '^/usr/lib/podman/aardvark-dns' 2>/dev/null | head -1)
    [ -z "$aardvark_pid" ] && return 1
    if timeout 3 nsenter -t "$aardvark_pid" -n bash -c '</dev/tcp/1.1.1.1/443' 2>/dev/null; then
        return 1   # recovered on its own
    fi

    log "Rootless-netns egress is broken (host online, container netns unreachable) — cycling"

    local PODMANCMD="sudo -u archipelago XDG_RUNTIME_DIR=/run/user/$archi_uid podman"
    local running
    running=$($PODMANCMD ps --format '{{.Names}}' 2>/dev/null)
    if [ -z "$running" ]; then
        log "  No running containers to cycle — skipping"
        return 1
    fi

    local count
    count=$(echo "$running" | wc -l)
    log "  Stopping $count running containers (graceful, 30s)..."
    $PODMANCMD stop --all --time 30 >/dev/null 2>&1
    sleep 5

    log "  Starting containers back up..."
    for c in $running; do
        $PODMANCMD start "$c" >/dev/null 2>&1 &
    done
    wait
    sleep 5

    aardvark_pid=$(pgrep -U "$archi_uid" -f '^/usr/lib/podman/aardvark-dns' 2>/dev/null | head -1)
    if [ -n "$aardvark_pid" ] && timeout 3 nsenter -t "$aardvark_pid" -n bash -c '</dev/tcp/1.1.1.1/443' 2>/dev/null; then
        log "  Rootless-netns egress restored ($count containers cycled)"
    else
        log "  WARN: egress still broken after cycle — may need manual intervention"
    fi
    return 0
}

# ── Fix 9: Restart stopped core containers ──────────────────
# Rootless Podman 4.x restart policies don't auto-restart on crash.
# This check restarts any exited core containers (tiers 0-2).
fix_stopped_core_containers() {
    local core_containers="bitcoin-knots lnd electrumx mempool-api archy-mempool-web archy-mempool-db archy-btcpay-db archy-nbxplorer btcpay-server"
    local restarted=()
    # Doctor runs as root but containers are rootless under archipelago user
    local PODMANCMD="sudo -u archipelago XDG_RUNTIME_DIR=/run/user/1000 podman"

    for name in $core_containers; do
        local state
        state=$($PODMANCMD inspect "$name" --format '{{.State.Status}}' 2>/dev/null || echo "missing")
        if [ "$state" = "exited" ] || [ "$state" = "stopped" ]; then
            log "Restarting stopped container: $name"
            $PODMANCMD start "$name" 2>/dev/null && restarted+=("$name") || true
        fi
    done

    [ ${#restarted[@]} -gt 0 ] && return 0 || return 1
}

# ── Main ─────────────────────────────────────────────────────

# If remote host provided, run via SSH
if [ -n "$1" ] && [ "$1" != "--local" ]; then
    REMOTE_HOST="$1"
    SSH_KEY="${ARCHIPELAGO_SSH_KEY:-$HOME/.ssh/archipelago-deploy}"
    SSH_OPTS="-o StrictHostKeyChecking=no -o ServerAliveInterval=15 -o ServerAliveCountMax=4 -i $SSH_KEY"

    log "Running container doctor on $REMOTE_HOST"

    # Copy script to remote and execute
    scp $SSH_OPTS "$0" "$REMOTE_HOST:/tmp/container-doctor.sh" 2>/dev/null
    ssh $SSH_OPTS "$REMOTE_HOST" "sudo bash /tmp/container-doctor.sh --local" 2>&1
    exit 0
fi

# Running locally (on the node itself)
log "Starting container health check"

run_fix "stale-podman"    fix_stale_podman
run_fix "orphaned-conmon" fix_orphaned_conmon
run_fix "system-tor"      fix_system_tor_conflict
run_fix "tor-permissions"  fix_tor_permissions
run_fix "searxng"         fix_searxng
run_fix "bitcoin-txindex" fix_bitcoin_txindex
run_fix "exit-127"        fix_exit_127
run_fix "netns-egress"    fix_rootless_netns_egress
run_fix "stopped-core"    fix_stopped_core_containers

echo ""
if [ $FIXES_APPLIED -gt 0 ]; then
    log "Done: $FIXES_APPLIED fixes applied (${FIX_NAMES[*]}), $CHECKS_PASSED checks passed"
else
    log "Done: all $CHECKS_PASSED checks passed — no fixes needed"
fi

exit 0