Enhance build-cloudinit-image.sh with improved error handling for decompression: add checks for xz installation, validate downloaded file type, and capture detailed error messages during decompression failures. Update deploy-to-proxmox.sh to restart cm4-dashboard service after enabling it, ensuring new code is loaded.

This commit is contained in:
nearxos
2026-02-19 15:38:33 +02:00
parent 39aa042dc9
commit ad00491487
3 changed files with 68 additions and 2 deletions

View File

@@ -66,8 +66,27 @@ if ! curl "${CURL_OPTS[@]}" "$URL" 2>"$CURL_ERR"; then
fi
write_status "decompressing" "Decompressing image…" "" ""
# Check we have a real xz file (not HTML error page)
if ! command -v xz >/dev/null 2>&1; then
write_status "error" "" "" "Decompress failed: xz not installed. Install xz-utils (apt install xz-utils)"
exit 1
fi
if [[ ! -s "$XZ_FILE" ]]; then
write_status "error" "" "" "Decompress failed: downloaded file is empty"
exit 1
fi
FILE_TYPE=$(file -b "$XZ_FILE" 2>/dev/null || true)
if [[ "$FILE_TYPE" == *"HTML"* ]] || [[ "$FILE_TYPE" == *"text"* ]] && [[ "$FILE_TYPE" != *"XZ"* ]]; then
write_status "error" "" "" "Decompress failed: download is not an image (got: ${FILE_TYPE:0:80})"
exit 1
fi
IMG_FILE="$TEMP_DIR/image.img"
xz -d -k -f "$XZ_FILE" || { write_status "error" "" "" "Decompress failed"; exit 1; }
XZ_ERR="$TEMP_DIR/xz_err.txt"
if ! xz -d -k -f "$XZ_FILE" 2>"$XZ_ERR"; then
err_detail=$(head -c 300 "$XZ_ERR" 2>/dev/null | tr '\n' ' ' | sed 's/"/\\"/g')
write_status "error" "" "" "Decompress failed: ${err_detail:-xz exited with error}"
exit 1
fi
[[ -f "$IMG_FILE" ]] || { write_status "error" "" "" "image.img not found after decompress"; exit 1; }
write_status "injecting" "Mounting boot partition and injecting cloud-init…" "" ""

View File

@@ -317,7 +317,8 @@ fi
pct exec "$CTID" -- cp /opt/cm4-provisioning/dashboard/cm4-dashboard.service /etc/systemd/system/
pct exec "$CTID" -- systemctl daemon-reload
pct exec "$CTID" -- systemctl enable --now cm4-dashboard
log "LXC: cm4-dashboard enabled and started."
pct exec "$CTID" -- systemctl restart cm4-dashboard
log "LXC: cm4-dashboard enabled and restarted (new code loaded)."
# --- LXC: optional SSH (root password + SSH key from deploy env) ---
if [[ -n "${DEPLOY_SSH_KEY_B64:-}" ]] || [[ -n "${DEPLOY_LXC_PWD_B64:-}" ]]; then

View File

@@ -0,0 +1,46 @@
#!/usr/bin/env bash
# Run on the Proxmox host (root@10.130.60.224) to diagnose cloud-init build failures.
# Usage: scp this script to the host, then: bash troubleshoot-cloudinit-build.sh
set -e
echo "=== Cloud-init build troubleshoot ==="
echo ""
echo "1. xz available?"
if command -v xz >/dev/null 2>&1; then
xz --version | head -1
else
echo " NOT FOUND. Install: apt install -y xz-utils"
fi
echo ""
echo "2. /tmp space (need several GB for decompress)?"
df -h /tmp
echo ""
echo "3. Provisioning dir and last request?"
PROV_DIR="${CM4_PROVISIONING_DIR:-/var/lib/cm4-provisioning}"
[[ -f /opt/cm4-provisioning/env ]] && source /opt/cm4-provisioning/env
echo " PROV_DIR=$PROV_DIR"
if [[ -f "$PROV_DIR/build_cloudinit_request.json" ]]; then
echo " Last request URL: $(python3 -c "import json; print(json.load(open('$PROV_DIR/build_cloudinit_request.json')).get('url','?'))" 2>/dev/null || echo '?')"
else
echo " No request file (build may have already run)."
fi
echo ""
echo "4. Last build status?"
if [[ -f "$PROV_DIR/build_cloudinit_status.json" ]]; then
cat "$PROV_DIR/build_cloudinit_status.json" | python3 -m json.tool 2>/dev/null || cat "$PROV_DIR/build_cloudinit_status.json"
else
echo " No status file."
fi
echo ""
TEST_DIR=$(mktemp -d)
trap "rm -rf $TEST_DIR" EXIT
echo "5. Quick xz decompress test?"
echo "test content" > "$TEST_DIR/f.txt"
xz -k "$TEST_DIR/f.txt" 2>/dev/null && xz -d -k -f "$TEST_DIR/f.txt.xz" 2>/dev/null && echo " OK" || echo " FAILED"
echo ""
echo "Done. If xz is missing, run on host: apt update && apt install -y xz-utils"