订阅

UiPath Automation Suite

The UiPath Automation Suite Guide

2021.10

Context


With an ever-growing fleet of products in our arsenal, we at UiPath are continuously exploring new ways to pave a smooth path towards automation.

We are preoccupied with your struggles and want to diminish them. We know that installing, setting up, and maintaining a large assortment of products could sometimes feel laborious, so we put some thought into it and came up with an initiative meant to reduce the complexity of these tasks.

Curious how this idea materialized? Meet Automation Suite, a framework designed to offer a consistent, painless, and elegant experience when installing, upgrading, and managing your UiPath products.

 

Automation Suite in the spotlight


The philosophy behind Automation Suite is to provide an all-encompassing solution that would minimize the effort you invest in bothersome tasks so that you can instead focus on making the most out of our tools. To put it briefly, Automation Suite bets big on a consistent and predictable experience across all UiPath products.

但是,您可能会问:“Automation Suite 到底是什么”。它是一个基于 Linux 的解决方案,它将我们的大多数服务器产品、软件依赖项、常用平台组件和基础架构捆绑到单个安装程序中。简而言之,它可以确保您轻松获取整个 UiPath 套件。

 

A sneak peek at Automation Suite


Let’s dive into some of the goodies Automation Suite brings about.

Hyperautomation

Automation Suite is designed to encapsulate our entire line of server products. The only exceptions are Process Mining and Data Service, which are not shipped with Automation Suite 2021.10 but are planned for a future release.

It’s created to implement our end-to-end hyperautomation platform strategy, which should guarantee an experience similar to the one in the cloud. Just think of the benefits of Automation Cloud made available to on-premises users.

Automation Suite provides the same set of capabilities as in the cloud for user management/directory API, as well as for tenant and license management.

Easy installation, upgrade, and maintenance

Automation Suite establishes one single delivery and maintenance mechanism for multiple UiPath products. This approach translates into simplified installation, upgrade, and maintenance endeavors. Troubleshooting should be equally undemanding.

It puts at your disposal the required dependencies and ensures they are installed and running. Depending on the products your business requires, each service you opt for needs to bring their manifests, configurations, and dependencies in a specific format.

Automation Suite installers cover both online and offline (air-gapped) environments.

Support for the installer, entire product stack and included infrastructure, and Kubernetes containerization are provided by UiPath.

If you are wondering how the Automation Suite initiative will impact our .MSI installers, you can rest assured that we will continue to deliver them. However, they will continue to be standalone installers per product.

The .MSI installers will also play a role in the upgrade to Automation Suite. That is because, if you are an existing customer, you first need to upgrade to the latest .MSI version, and only then move to Automation Suite.

Scalability

Automation Suite is a scalable solution that allows you to start small and then expand according to your business needs. Needless to say, it is available in both a single-node and multi-node mode, and you can start with the respective setups based on your needs. We purposefully try to be mindful of your long-term strategies, regardless of their direction and magnitude, so no worries on this front.

 

Known issues and limitations


安装

  • 在安装时需要根访问权限才能设置文件夹和安装权限。

  • ODBC 连接不支持包含特殊字符的用户名。对于 AI Center 和 Document Understanding 的数据库用户名,请仅使用大写和小写字母。

  • Using the default value (1450) when configuring the MTU on Kubernetes Pod level could result in various issues for Azure Virtual Network. To prevent these problems, we recommend setting the Pod MTU to 1350. For more details, see Troubleshooting.

Automation Suite 门户

组织管理员必须具有关联的电子邮件地址,否则在尝试配置单个产品时会发生 400 错误。

Orchestrator

  • 带有警示的定期电子邮件不会发送给用户。为确保 Google 帐户的 SMTP 服务器正常工作,您需要在网络安全组上为配置的 SMTP 端口设置新的出站规则。在 Automation Suite 中为 SMTP 配置的设置应该会在 60 分钟内传播到 Orchestrator。

AI Center

  • On rare occasions, if you restart the machine two times consecutively, service deployment can get stuck because of the DATABASECHANGELOGLOCK lock not being released by one service. In this case you will see AI Center pods restarting continuously. If this happens, see the Troubleshooting section.

  • 上传活动时无法使用数据集的公共端点。数据集的公共端点仍可直接与 HTTP 调用一起使用。

  • 在 Automation Suite 离线环境中使用 AI Center 时,有些组件未正确加载。在无法访问互联网的计算机上打开“AI 应用程序”页面时,会发生这种情况。

  • In a rare scenario, due to cluster state, asynchronous operations like Skill Deployment or Package upload could be stuck for a long time. If DU Skill deployment is taking more than 2-3 hours, try deploying a simpler model (e.g, TemplateModel). If the model also takes more than an hour, then the mitigation is to restart AI Center services with the following commands:

    • kubectl -n uipath rollout restart deployment ai-deployer-deployment
    • kubectl -n uipath rollout restart deployment ai-trainer-deployment
    • kubectl -n uipath rollout restart deployment ai-pkgmanager-deployment
    • kubectl -n uipath rollout restart deployment ai-helper-deployment
    • kubectl -n uipath rollout restart deployment ai-appmanager-deployment

使用以下命令进行验证,等待 AI Center Pod 重新启动:

* `kubectl -n uipath get pods | grep ai-*`

以上所有 Pod 都应处于“正在运行”状态,并且容器状态应显示为 2/2。

Document Understanding

  • 在 AI Center 中创建 Document Understanding OCR ML 包时,其名称不能为 ocrOCR。请务必选择其他名称。

  • 有时,DefaultTenant 未启用 Document Understanding。这表明其自身无法创建数据标签会话。要解决此问题,请在该租户上禁用 AI Center,然后重新启用。请注意,您可能需要等待几分钟才能重新启用 AI Center。

  • Document Understanding 要求在 SQL Server 上启用“全文搜索”功能。否则,安装将失败,并且在这方面没有明确的错误消息。

  • 如果您已从最早的内部版本升级到最新版本,则 Document Understanding ML 技能预测可能无法正常工作,并显示prediction failed错误消息。建议的解决方法是从本地存储中删除以 auth_token_aicenter 开头的所有键,然后重新登录 AI Center。

  • Auto-retrain pods may remain in NotReady state after completion. Suggested workaround is to delete the NotReady pods following the command below:
    kubectl get pods -n uipath | grep autoretrain | grep "NotReady" | awk '{print $1}' | xargs kubectl delete pod

  • After backup and restore on baremetal setup, the Training Pipeline and the Data Labeling session may get stuck on AI Center.
    Suggested workaround for the Training Pipeline is to manually delete the ai-trainer-deployment pod in ALM (available at https://alm.hostname) and restart.
    Suggested workaround for the Data Labeling session is to manually delete the ai-app-manager pod in ALM (available at https://alm.hostname) and restart.

Insights

  • Insights 不支持高可用性。
  • 还原后,Insights Looker Pod 无法启动。您可以通过在 ArgoCD 用户界面中从 Insights 应用程序中删除 Looker Pod 来解决此问题。部署将创建一个新的 Pod,它应该能够成功启动。

Test Manager

  • 如果您的许可证是在登录时分配的,则打开 Test Manager 时可能无法检测到您的许可证分配。

如果发生这种情况,请执行以下步骤:

  1. 导航到 Test Manager。
  2. 从门户中注销。
  3. 重新登录。

Automation Hub

  • 无法将未在 Automation Suite 或 Active Directory 中定义电子邮件地址的用户添加到 Automation Hub。如果发生这种情况,请确保用户属于 Active Directory 组或 Automation Suite 组,并将该组添加到 Automation Hub。

2 个月前更新


2021.10


建议的编辑仅限用于 API 参考页面

您只能建议对 Markdown 正文内容进行编辑,而不能建议对 API 规范进行编辑。