vm-bench/backend/app.py

268 lines
9.8 KiB
Python

"""
Backend API — Proxmox Image Conversion Engine
Runs on the Proxmox host (srv2) at http://10.2.0.2:9000
All API routes live under /api/v1/.
Matches open-api.yaml v1.1.0 spec exactly.
"""
from __future__ import annotations
import logging
from logging.handlers import RotatingFileHandler
from pathlib import Path
from fastapi import FastAPI, HTTPException, Request
from fastapi.middleware.cors import CORSMiddleware
from models import (
JobSubmissionRequest,
JobStatusResponse,
CleanupRequest,
CleanupResponse,
AnalyzeRequest,
AnalyzeResponse,
AnalyzeStatusResponse,
CloneRequest,
HealthResponse,
ErrorResponse,
)
from converter import extract_if_needed, discover_disk, detect_os, detect_efi, start_analysis, get_analysis_status
from provisioner import submit_job, get_job_status, cleanup_staging, clone_vm
# ---------------------------------------------------------------------------
# Logging
# ---------------------------------------------------------------------------
logger = logging.getLogger("backend")
logger.setLevel(logging.INFO)
# Console → systemd journal
_ch = logging.StreamHandler()
_ch.setFormatter(logging.Formatter("%(asctime)s %(levelname)s %(message)s"))
logger.addHandler(_ch)
# File handler → shared log directory
LOG_DIR = Path(__file__).resolve().parent.parent / "logs"
LOG_DIR.mkdir(parents=True, exist_ok=True)
_fh = RotatingFileHandler(
LOG_DIR / "vm-bench-backend.log", maxBytes=10 * 1024 * 1024, backupCount=5,
)
_fh.setFormatter(logging.Formatter("%(asctime)s %(levelname)s %(name)s %(message)s"))
logger.addHandler(_fh)
logger.info("Backend starting — log file: %s", LOG_DIR / "vm-bench-backend.log")
# ---------------------------------------------------------------------------
# Constants
# ---------------------------------------------------------------------------
API_PREFIX = "/api/v1"
# ---------------------------------------------------------------------------
# App
# ---------------------------------------------------------------------------
app = FastAPI(
title="Proxmox Image Conversion Engine",
version="1.1.0",
docs_url="/docs",
)
app.add_middleware(
CORSMiddleware,
allow_origins=["*"],
allow_methods=["*"],
allow_headers=["*"],
)
# ---------------------------------------------------------------------------
# Helpers
# ---------------------------------------------------------------------------
def _validate_filename(filename: str) -> None:
"""Reject filenames that attempt path traversal."""
if ".." in filename or filename.startswith("/"):
raise HTTPException(status_code=400, detail="Invalid filename: path traversal not allowed")
# ---------------------------------------------------------------------------
# Routes — all under /api/v1/
# ---------------------------------------------------------------------------
@app.get(f"{API_PREFIX}/health", response_model=HealthResponse)
def health() -> HealthResponse:
"""Health check."""
return HealthResponse(status="ok")
@app.get(f"{API_PREFIX}/storage/pools")
def storage_pools():
"""Return available Proxmox storage pools that support VM disk images."""
pools = []
try:
cfg = Path("/etc/pve/storage.cfg")
if cfg.exists():
lines = cfg.read_text()
current = {}
for line in lines.split("\n"):
stripped = line.strip()
if not stripped or stripped.startswith("#"):
continue
if ":" in stripped and not line.startswith(" "):
if current and current.get("name"):
pools.append(current)
parts = stripped.split(":", 1)
stype = parts[0].strip()
sname = parts[1].strip() if len(parts) > 1 else ""
current = {"name": sname, "type": stype}
elif line.startswith(" ") and current:
if " " in stripped:
key, val = stripped.split(" ", 1)
current[key] = val
if current and current.get("name"):
pools.append(current)
except Exception:
pass
result = []
for p in pools:
if p.get("type") in ("lvmthin", "zfspool", "rbd", "dir"):
content = p.get("content", "")
if "images" in content:
result.append({
"name": p["name"],
"type": p["type"],
})
if not result:
result.append({"name": "local-lvm", "type": "lvmthin"})
return {"pools": result}
@app.get(f"{API_PREFIX}/sessions")
def list_sessions():
"""Return existing session directories with staged files that can be resumed."""
tmp = Path("/mnt/converter/tmp") if Path("/mnt/converter/tmp").exists() else Path(__file__).resolve().parent.parent / "tmp"
sessions = []
try:
for entry in sorted(tmp.iterdir(), reverse=True):
if not entry.is_dir():
continue
in_dir = entry / "in"
if not in_dir.is_dir():
continue
session_files = []
total_bytes = 0
has_disks = False
all_files = sorted([f for f in in_dir.rglob("*") if f.is_file()], key=lambda f: f.stat().st_size, reverse=True)
for f in all_files:
sz = f.stat().st_size
total_bytes += sz
rel = str(f.relative_to(in_dir))
session_files.append({"name": rel, "size_bytes": sz, "size_gb": round(sz / (1024**3), 1)})
if f.suffix.lower() in {".vmdk", ".qcow2", ".img", ".raw", ".vhd", ".vhdx"}:
has_disks = True
if not session_files:
continue
archives = {".7z", ".zip", ".rar", ".tar.gz", ".tgz", ".tar", ".gz", ".bz2", ".xz", ".ova"}
has_source = has_disks or any(f["name"].lower().endswith(tuple(archives)) for f in session_files)
sessions.append({
"session_id": entry.name,
"files": session_files[:10],
"total_size_gb": round(total_bytes / (1024**3), 1),
"has_disks": has_disks,
"resumable": has_source,
})
except Exception:
pass
return {"sessions": sessions}
@app.post(f"{API_PREFIX}/analyze", response_model=AnalyzeStatusResponse, status_code=202)
def analyze(req: AnalyzeRequest) -> AnalyzeStatusResponse:
"""Analyze a source file asynchronously: extraction + disk probing + OS detection."""
_validate_filename(req.source_filename)
logger.info("Analyze request: vmid=%d file=%s", req.vmid, req.source_filename)
try:
result = start_analysis(req.vmid, req.source_filename)
return AnalyzeStatusResponse(
analysis_id=result["analysis_id"],
vmid=req.vmid,
status=result["status"],
message=result["message"],
)
except Exception as exc:
logger.error("Analysis start failed: %s", exc)
raise HTTPException(status_code=400, detail=str(exc))
@app.get(f"{API_PREFIX}/analyze/{{analysis_id}}", response_model=AnalyzeStatusResponse)
def analyze_status(analysis_id: str) -> AnalyzeStatusResponse:
"""Poll async analysis status."""
try:
result = get_analysis_status(analysis_id)
return AnalyzeStatusResponse(**result)
except KeyError:
logger.warning("Analysis not found: %s", analysis_id)
raise HTTPException(status_code=404, detail=f"Analysis not found: {analysis_id}")
@app.post(f"{API_PREFIX}/jobs", response_model=JobStatusResponse, status_code=202)
def create_job(req: JobSubmissionRequest) -> JobStatusResponse:
"""Submit a conversion + provisioning job."""
logger.info("Job submit: vmid=%d name=%s disk=%s storage=%s",
req.vmid, req.vm_name,
req.boot_disk.source_filename, req.target_storage)
try:
result = submit_job(req)
logger.info("Job queued: %s", result.job_id)
return result
except Exception as exc:
logger.error("Job submission failed: %s", exc)
raise HTTPException(status_code=400, detail=str(exc))
@app.get(f"{API_PREFIX}/jobs/{{job_id}}", response_model=JobStatusResponse)
def job_status(job_id: str) -> JobStatusResponse:
"""Get current status of a conversion job."""
try:
result = get_job_status(job_id)
return result
except Exception:
logger.warning("Job not found: %s", job_id)
raise HTTPException(status_code=404, detail=f"Job not found: {job_id}")
@app.post(f"{API_PREFIX}/jobs/{{job_id}}/cleanup", response_model=CleanupResponse)
def cleanup_job(job_id: str, req: CleanupRequest) -> CleanupResponse:
"""Delete or preserve staging files for a job."""
try:
vmid = int(job_id.split("_")[1])
except (IndexError, ValueError):
vmid = 0
logger.info("Cleanup: job=%s vmid=%d delete=%s", job_id, vmid, req.delete_staging_files)
result = cleanup_staging(vmid, req.delete_staging_files, req.session_id)
return CleanupResponse(
job_id=result["job_id"],
action_taken=result["action_taken"],
message=result["message"],
)
@app.post(f"{API_PREFIX}/clone", response_model=JobStatusResponse, status_code=202)
def clone_vm_endpoint(req: CloneRequest) -> JobStatusResponse:
"""Clone an existing VM with a new ID and name."""
logger.info("Clone request: %d%d (%s)", req.source_vmid, req.target_vmid, req.target_name)
try:
return clone_vm(req)
except Exception as exc:
logger.error("Clone failed: %s", exc)
raise HTTPException(status_code=400, detail=str(exc))
# ---------------------------------------------------------------------------
# Run
# ---------------------------------------------------------------------------
if __name__ == "__main__":
import uvicorn
uvicorn.run(app, host="0.0.0.0", port=9000)