renderg-sdk 0.1.31

www.renderg.com

RenderG SDK

RenderG SDK 是面向 RenderG 云渲染平台 的 Python 客户端开发工具包， 帮助开发者将 DCC 软件的渲染工作流快速接入 RenderG 渲染农场。

功能特性

• 多 DCC 支持：内置 Maya、Houdini、Clarisse 场景文件分析模块
• 自动资产收集：分析场景依赖，生成资产清单（info.cfg），确保所有贴图、缓存、代理文件完整上传
• 高速文件传输：基于 Aspera 协议的高速上传/下载，支持联通、电信多线路选择
• 完整作业管理：创建、提交、暂停、恢复、加速、删除渲染作业及帧任务
• 实时状态监听：通过 MQTT 订阅作业完成通知，渲染结束后自动下载结果

环境要求

• Python 2 / Python 3
• 依赖包：requests、tenacity、paho-mqtt、typing_extensions

安装

pip install renderg-sdk

或从源码安装：

git clone https://github.com/renderg-repo/renderg-sdk.git
cd renderg-sdk
pip install -e .

快速开始

1. 创建配置文件

在项目目录下新建 config.json：

{
  "AUTH_KEY": "your_auth_key_here",
  "CLUSTER_ID": 27,
  "PROJECT_ID": 21479,
  "ENV_ID": 7715,
  "ZONE_ID": 1009,
  "RAM_LIMIT": "64G"
}

字段	说明	获取方式
AUTH_KEY	用户身份认证密钥	联系 RenderG 技术支持 获取
CLUSTER_ID	集群 ID	api.config.get_cluster_list()
PROJECT_ID	项目 ID	api.project.get_project_list()
ENV_ID	渲染环境 ID	api.env.get_env_list()
ZONE_ID	硬件区域 ID	api.config.get_zone_list()
RAM_LIMIT	内存限制	64G / 128G / 256G

2. 初始化 API

from renderg_api import RenderGAPI
from renderg_utils import utils

config = utils.read_json("./config.json")
api = RenderGAPI(auth_key=config["AUTH_KEY"], cluster_id=config["CLUSTER_ID"])

也可通过环境变量传入密钥，无需在代码中硬编码：

set RENDERG_AUTH_KEY=your_auth_key_here

api = RenderGAPI(cluster_id=27)  # 自动读取环境变量 RENDERG_AUTH_KEY

---

提交示例

Maya

import os
from renderg_utils import utils, log
from analyze_maya import AnalyzeMaya, ParamChecker
from renderg_api import RenderGAPI
from renderg_api.constants import TransferLines
from renderg_transfer.RGUpload import RenderGUpload
from renderg_transfer.RGDownload import RenderGDownload

config = utils.read_json("./config.json")
workspace = config.get("WORKSPACE", os.path.expandvars("%userprofile%/RenderG_WorkSpace"))

log.init_logging(log_dir=utils.get_workspace(workspace), console=True)
logger = log.get_logger()

api = RenderGAPI(auth_key=config["AUTH_KEY"], cluster_id=config["CLUSTER_ID"])

# 1. 分析场景，收集依赖资产
env_info = api.env.get_env_info_by_id(config["ENV_ID"])
analyze_obj = AnalyzeMaya(
    dcc_file=r"E:\Work\Scene\Maya\scene.ma",
    dcc_version="2025",
    project_path=env_info.get("project_path") or r"E:\Work\Scene\Maya",
    api=api,
    project_id=config["PROJECT_ID"],
    env_id=config["ENV_ID"],
    workspace=workspace,
    logger=logger,
)
analyze_obj.analyze()

# 2. 设置渲染参数
param_check_obj = ParamChecker(api, analyze_obj)
param_check_obj.set_render_layers({
    "defaultRenderLayer": {
        "ForceRenderFrames": "1-10",
        "RenderCameras": "",        # 留空使用场景默认相机
        "Renderable": True,
        "RenderWidth": "1920",
        "RenderHeight": "1080",
    }
})
param_check_obj.execute(
    ChunkSize=1,
    zone_id=config["ZONE_ID"],
    ram_limit=config["RAM_LIMIT"],
)

# 3. 上传资产
RenderGUpload(
    api=api,
    job_id=analyze_obj.job_id,
    info_path=analyze_obj.info_path,
    line=TransferLines.LINE_UNICOM,
    speed=200,
).upload()

# 4. 提交作业，开始渲染
api.job.submit_job(analyze_obj.job_id)

# 5. 等待完成后自动下载
RenderGDownload(
    api=api,
    job_id=analyze_obj.job_id,
    download_path="D:/renders/output",
    line=TransferLines.LINE_UNICOM,
    cluster_id=config["CLUSTER_ID"],
    speed=500,
).auto_download_after_job_completed()

Houdini

import os
from renderg_utils import utils, log
from analyze_houdini import AnalyzeHoudini, ParamChecker
from renderg_api import RenderGAPI
from renderg_api.constants import TransferLines
from renderg_transfer.RGUpload import RenderGUpload
from renderg_transfer.RGDownload import RenderGDownload

config = utils.read_json("./config.json")
workspace = config.get("WORKSPACE", os.path.expandvars("%userprofile%/RenderG_WorkSpace"))

log.init_logging(log_dir=utils.get_workspace(workspace), console=True)
logger = log.get_logger()

api = RenderGAPI(auth_key=config["AUTH_KEY"], cluster_id=config["CLUSTER_ID"])

# 1. 分析场景，收集依赖资产
analyze_obj = AnalyzeHoudini(
    dcc_file=r"E:\Work\Scene\Houdini\scene.hip",
    dcc_version="20.5.584",
    api=api,
    project_id=config["PROJECT_ID"],
    env_id=config["ENV_ID"],
    workspace=workspace,
    logger=logger,
)
analyze_obj.analyze()

# 2. 设置渲染参数（指定 ROP 节点和帧范围）
param_check_obj = ParamChecker(api, analyze_obj)
param_check_obj.set_render_nodes({"/out/mantra1": "1-10"})
param_check_obj.execute(
    ChunkSize=1,
    zone_id=config["ZONE_ID"],
    ram_limit=config["RAM_LIMIT"],
)

# 3. 上传资产
RenderGUpload(
    api=api,
    job_id=analyze_obj.job_id,
    info_path=analyze_obj.info_path,
    line=TransferLines.LINE_UNICOM,
    speed=200,
).upload()

# 4. 提交作业，开始渲染
api.job.submit_job(analyze_obj.job_id)

# 5. 等待完成后自动下载
RenderGDownload(
    api=api,
    job_id=analyze_obj.job_id,
    download_path="D:/renders/output",
    line=TransferLines.LINE_UNICOM,
    cluster_id=config["CLUSTER_ID"],
    speed=500,
).auto_download_after_job_completed()

Clarisse

import os
from renderg_utils import utils, log
from analyze_clarisse import AnalyzeClarisse, ParamChecker
from renderg_api import RenderGAPI
from renderg_api.constants import TransferLines
from renderg_transfer.RGUpload import RenderGUpload
from renderg_transfer.RGDownload import RenderGDownload

config = utils.read_json("./config.json")
workspace = config.get("WORKSPACE", os.path.expandvars("%userprofile%/RenderG_WorkSpace"))

log.init_logging(log_dir=utils.get_workspace(workspace), console=True)
logger = log.get_logger()

api = RenderGAPI(auth_key=config["AUTH_KEY"], cluster_id=config["CLUSTER_ID"])

# 1. 分析场景，收集依赖资产
analyze_obj = AnalyzeClarisse(
    dcc_file=r"E:\Work\Scene\Clarisse\scene.project",
    dcc_version="5.0 SP11",
    api=api,
    project_id=config["PROJECT_ID"],
    env_id=config["ENV_ID"],
    workspace=workspace,
    logger=logger,
)
analyze_obj.analyze()

# 2. 设置渲染参数（指定 Image 节点路径和帧范围）
param_check_obj = ParamChecker(api, analyze_obj)
param_check_obj.set_render_layers({
    "project://render/cam2.color_sun_chr": "1-100",
    "project://render/cam2.ID_chr":        "1-100",
})
param_check_obj.execute(
    ChunkSize=1,
    zone_id=config["ZONE_ID"],
    ram_limit=config["RAM_LIMIT"],
)

# 3. 上传资产
RenderGUpload(
    api=api,
    job_id=analyze_obj.job_id,
    info_path=analyze_obj.info_path,
    line=TransferLines.LINE_UNICOM,
    speed=200,
).upload()

# 4. 提交作业，开始渲染
api.job.submit_job(analyze_obj.job_id)

# 5. 等待完成后自动下载
RenderGDownload(
    api=api,
    job_id=analyze_obj.job_id,
    download_path="D:/renders/output",
    line=TransferLines.LINE_UNICOM,
    cluster_id=config["CLUSTER_ID"],
    speed=500,
).auto_download_after_job_completed()

---

作业管理

# 查询作业列表
jobs = api.job.get_job_list(page=1, count=20)

# 查询单个作业详情
info = api.job.get_jobs_info(job_id)

# 控制作业（支持列表批量操作）
api.job.stop_job([job_id])       # 暂停
api.job.start_job([job_id])      # 恢复
api.job.requeue_job([job_id])    # 重新排队
api.job.speedup_job([job_id])    # 加速
api.job.delete_job([job_id])     # 删除

# 查询并控制帧任务
tasks = api.task.get_task_list_by_job_id(job_id)
api.task.requeue_task(job_id, task_id_list=[234617690, 234617691])  # 重新渲染指定帧

错误处理

from renderg_api.exception import RenderGAPIError

try:
    api.job.submit_job(job_id)
except RenderGAPIError as e:
    print("错误码:", e.err_code)
    print("错误信息:", e.err_msg)
    print("请求地址:", e.request_url)

完整示例

更多详细示例请参考 example/ 目录：

• example/submit_maya.py — Maya 完整提交示例
• example/submit_houdini.py — Houdini 完整提交示例
• example/submit_clarisse.py — Clarisse 完整提交示例