FAQ（常见问题解答）¶
Is there a central place to view all user uploaded images within Foundry?¶
Yes; Control Panel provides Foundry administrators the ability to view all images pushed into the platform, as well as view and recall any vulnerabilities present in those images.
Are container-backed models compatible with live deployments?¶
Yes, container models can be used within live and batch deployments.
Is there a particular base image that should be used when creating images to push to Foundry?¶
No, there is no standard base image provided nor required. However, all images pushed must adhere to the image requirements.
Is there a standard API that the image must expose?¶
No; how Foundry interacts with the image will be defined by the model adapter implementation. A common pattern reflected in this example custom adapter is to construct the image to listen on a specific port for input, then have the model adapter send post requests.
What is the maximum image size for containerized models?¶
Typically, images larger than 22 GB will time out during the Docker push step. If your use case requires a larger image, contact your Palantir representative.
Does Foundry support the execution of Windows-based containers?¶
No; all images pushed into the platform must be built for the Linux platform as the entities in the Foundry Kubernetes cluster are Linux machines.
Can I use multiple containers within my model?¶
Yes; multiple images can be configured to back a model version, but there is no orchestration support. All containers will be launched simultaneously at execution time and it will not be possible to guarantee container start time ordering.
Are container-backed models supported on all Foundry instances?¶
No. All user-provided container workflows require the Rubix ↗ engine as the backing infrastructure. Also, you need to enable container workflows in Control Panel.
How do I enable telemetry on my model?¶
To enable telemetry on your model, create a new model version and toggle on Enable telemetry in the third step of model version creation. The image must have a shell executable in /bin/sh, and the image must support the shell commands set and tee.
Will telemetry work for all types of models?¶
No, telemetry for containerized models only works in Python transforms and live deployments, but will not emit container logs for batch deployments.
How can I test if my container has the necessary commands and settings for telemetry?¶
You can test this by running docker run --entrypoint /bin/sh <EXAMPLE_IMAGE>:<IMAGE_TAG_OR_DIGEST> -c 'set -a && tee -a' && echo "Telemetry compatible". An output of Telemetry compatible indicates telemetry can be enabled for this container.
中文翻译¶
常见问题解答¶
Foundry 中是否有集中查看所有用户上传镜像（image）的位置？¶
是的；控制面板（Control Panel）为 Foundry 管理员提供了查看所有推送到平台上的镜像的功能，同时还可以查看并召回这些镜像中存在的任何漏洞。
基于容器的模型（container-backed model）是否兼容实时部署（live deployment）？¶
是的，容器模型（container model）可用于实时和批量部署。
创建推送到 Foundry 的镜像时，是否需要使用特定的基础镜像（base image）？¶
不需要，平台既不提供也不要求使用标准基础镜像。不过，所有推送的镜像必须符合镜像要求。
镜像是否需要暴露标准 API？¶
不需要；Foundry 与镜像的交互方式将由模型适配器实现（model adapter implementation）定义。此示例自定义适配器中反映的常见模式是：构建镜像使其在特定端口监听输入，然后由模型适配器（model adapter）发送 POST 请求。
容器化模型（containerized model）的最大镜像大小是多少？¶
通常情况下，超过 22 GB 的镜像会在 Docker 推送步骤中超时。如果您的用例需要更大的镜像，请联系您的 Palantir 代表。
Foundry 是否支持运行基于 Windows 的容器？¶
不支持；所有推送到平台上的镜像必须为 Linux 平台构建，因为 Foundry Kubernetes 集群中的实体均为 Linux 机器。
我可以在模型中使用多个容器吗？¶
可以；可以配置多个镜像来支持一个模型版本，但不提供编排（orchestration）支持。所有容器将在执行时同时启动，无法保证容器的启动顺序。
所有 Foundry 实例是否都支持基于容器的模型？¶
不是。所有用户提供的容器工作流都需要 Rubix ↗ 引擎作为底层基础设施。此外，您需要在控制面板中启用容器工作流。
如何为我的模型启用遥测（telemetry）？¶
要为模型启用遥测，请创建一个新的模型版本，并在模型版本创建的第三步中开启启用遥测选项。镜像必须包含 /bin/sh 中的 shell 可执行文件，并且必须支持 set 和 tee 这两个 shell 命令。
遥测是否适用于所有类型的模型？¶
不适用，容器化模型的遥测仅适用于 Python 转换（Python transforms）和实时部署，但不会为批量部署（batch deployment）输出容器日志。
如何测试我的容器是否具备遥测所需的命令和设置？¶
您可以通过运行以下命令进行测试：docker run --entrypoint /bin/sh <EXAMPLE_IMAGE>:<IMAGE_TAG_OR_DIGEST> -c 'set -a && tee -a' && echo "Telemetry compatible"。如果输出结果为 Telemetry compatible，则表示该容器可以启用遥测。