sgl-project/sglang · warning · HTTPException

Generation is still in-progress

Error message

Generation is still in-progress

What it means

The mesh job exists but has no file_path yet (or the file was deleted from disk), so content download returns HTTP 404 'Generation is still in-progress'. The job record exists but the artifact file is not ready.

Source

Thrown at python/sglang/multimodal_gen/runtime/entrypoints/openai/mesh_api.py:270

@router.get("/{mesh_id}/content")
async def download_mesh_content(
    mesh_id: str = Path(...), variant: Optional[str] = Query(None)
):
    job = await MESH_STORE.get(mesh_id)
    if not job:
        raise HTTPException(status_code=404, detail="Mesh not found")

    if job.get("url"):
        raise HTTPException(
            status_code=400,
            detail=f"Mesh has been uploaded to cloud storage. Please use the cloud URL: {job.get('url')}",
        )

    file_path = job.get("file_path")
    if not file_path or not os.path.exists(file_path):
        raise HTTPException(status_code=404, detail="Generation is still in-progress")

    ext = os.path.splitext(file_path)[1].lower()
    media_type = {
        ".glb": "model/gltf-binary",
        ".obj": "text/plain",
    }.get(ext, "application/octet-stream")

    return FileResponse(
        path=file_path, media_type=media_type, filename=os.path.basename(file_path)
    )

View on GitHub (pinned to 0132848349)

Solutions

1. Poll GET /{mesh_id} status until completed/failed before requesting content
2. Add exponential backoff between content-download attempts
3. Inspect job status: if it shows failed, re-submit instead of retrying download

Example fix

# before
resp = await client.get(f"/v1/mesh/{id}/content")
# after
while (job := (await client.get(f"/v1/mesh/{id}")).json())["status"] not in ("completed","failed"):
    await asyncio.sleep(2)
resp = await client.get(f"/v1/mesh/{id}/content")

job = (await client.get(f'/v1/mesh/{id}')).json()
if job['status'] != 'completed': sleep and re-poll

while True:
    r = await client.get(f'/v1/mesh/{id}/content')
    if r.status_code != 404 or 'in-progress' not in r.text: break
    await asyncio.sleep(2)

When it happens

Trigger: Polling GET /{mesh_id}/content immediately after job creation, before generation finished and wrote the output file.

Common situations: Impatient polling without backoff; job failed before writing output; artifact cleaned from disk by a janitor process.

Related errors

• Failed to process image source: {str(e)}
• Invalid request body: {e}
• An input image is required for mesh generation
• Mesh not found
• Mesh has been uploaded to cloud storage. Please use the clou

AI-assisted analysis of sgl-project/sglang@0132848349 (2026-08-28). Data as JSON: /api/errors/2cfbe3bc758cd2c6. Report an issue: GitHub.