2018 年 5 月May 2018

發行是暫存的。Releases are staged. 您的 Azure Databricks 帳戶可能會在初始發行日期的一周後才更新。Your Azure Databricks account may not be updated until a week after the initial release date.

一般資料保護規定 (GDPR)General Data Protection Regulation (GDPR)

2018月24日:2.72 版May 24, 2018: Version 2.72

為了符合歐盟一般資料保護規定 (GDPR) 的需求(在2018年5月25日生效),我們對 Azure Databricks 平臺進行了一些修改,讓您能夠更充分掌控帳戶和使用者層級的資料保留。To meet the requirements of the European Union General Data Protection Regulation (GDPR), which goes into effect on May 25, 2018, we have made a number of modifications to the Azure Databricks platform to provide you with more control of data retention at both the account and user level. 更新包括:Updates include:

  • 叢集刪除:使用 UI 或叢集 API 永久刪除叢集設定。Cluster delete: permanently delete a cluster configuration using the UI or the Clusters API. 請參閱 刪除叢集。See Delete a cluster.
  • 版本2.71 中發行的工作區清除 () :永久刪除工作區物件,例如整個筆記本、個別筆記本資料格、個別筆記本批註,以及筆記本修訂歷程記錄。Workspace purge (released in version 2.71): permanently delete workspace objects, such as entire notebooks, individual notebook cells, individual notebook comments, and notebook revision history. 請參閱 管理工作區儲存體See Manage workspace storage.
  • 筆記本修訂歷程記錄清除:Notebook revision history purge:
    • 針對已定義的時間範圍,永久刪除工作區中所有筆記本的修訂歷程記錄。Permanently delete the revision history of all notebooks in a workspace for a defined time frame. 請參閱 管理工作區儲存體See Manage workspace storage.
    • 永久刪除單一筆記本修訂或筆記本的整個修訂歷程記錄。Permanently delete a single notebook revision or the entire revision history of a notebook. 請參閱 版本控制See Version control.

如需刪除 Azure Databricks 服務或取消 Azure 帳戶的相關資訊,請參閱 管理您的訂用帳戶。For information about deleting your Azure Databricks service or canceling your Azure account, see Manage your subscription.

Azure Databricks 使用者必須屬於 Azure AD 租用戶Azure Databricks users must belong to Azure AD tenant

2018月24日:2.72 版May 24, 2018: Version 2.72

使用者現在可以登入 Azure Databricks,只有當他們屬於) 工作區的 Azure Active Directory (Azure AD Azure Databricks 租 使用者時才會登入。Users can now sign in to Azure Databricks only if they belong to the Azure Active Directory (Azure AD) tenant of the Azure Databricks workspace. 如果您有不屬於 Azure AD 租使用者的使用者,您可以 將它們新增為標準或來賓使用者If you have users who do not belong to the Azure AD tenant, you can add them as standard or guest users.

HorovodEstimatorHorovodEstimator

2018月29日:2.72 版May 29, 2018: Version 2.72

已新增適用于 HorovodEstimator 的檔和筆記本,這是利用 Uber 的 Horovod 架構的 MLlib 樣式估算器 API。Added documentation and a notebook for HorovodEstimator, an MLlib-style estimator API that leverages Uber’s Horovod framework. HorovodEstimator 可促進 Spark 資料框架上深度類神經網路的分散式、多重 GPU 訓練,簡化了 Spark 中的 ETL 與 TensorFlow 中模型定型的整合。HorovodEstimator facilitates distributed, multi-GPU training of deep neural networks on Spark DataFrames, simplifying the integration of ETL in Spark with model training in TensorFlow. 請參閱 HorovodEstimator:使用 Horovod 的分散式深度學習和 Apache Spark MLlibSee HorovodEstimator: distributed deep learning with Horovod and Apache Spark MLlib.

MLeap ML 模型匯出MLeap ML Model Export

2018月24日:2.72 版May 24, 2018: Version 2.72

在 Azure Databricks 上使用 MLeap 新增了檔和筆記本。Added documentation and notebooks on using MLeap on Azure Databricks. MLeap 可讓您從 Apache Spark 和 scikit-learn 部署機器學習管線-學習到可移植的格式和執行引擎。MLeap allows you to deploy machine learning pipelines from Apache Spark and scikit-learn to a portable format and execution engine. 請參閱 MLEAP ML 模型匯出See MLeap ML Model Export.

更多的 GPU 叢集類型Even more GPU cluster types

2018月24日:2.72 版May 24, 2018: Version 2.72

除了我們在版本2.71 中新增的 Azure NC 實例類型 (NC12 和 NC24) ,我們現在支援 NC12s_v3 叢集上的 NCv3 實例類型系列 (NC6s_v3NC24s_v3) Azure Databricks。In addition to the Azure NC instance types (NC12 and NC24) that we added in Release 2.71, we now support the NCv3 instance type series (NC6s_v3, NC12s_v3, and NC24s_v3) on Azure Databricks clusters. NC 和 NCv3 實例提供 Gpu 來處理電源影像處理、文字分析,以及其他具有運算挑戰和需求優異效能的機器學習和深度學習工作。NC and NCv3 instances provide GPUs to power image processing, text analysis, and other machine learning and deep learning tasks that are computationally challenging and demand superior performance.

請參閱 已啟用 GPU 的叢集。See GPU-enabled clusters.

筆記本資料格:隱藏和顯示Notebook cells: hide and show

2018月24日:2.72 版May 24, 2018: Version 2.72

新的指標和訊息可讓您更輕鬆地在隱藏筆記本資料格內容時加以顯示。New indicators and messaging make it easier to show Notebook cell contents after they’ve been hidden. 請參閱 隱藏和顯示資料格內容See Hide and show cell content.

2018 年 5 月 22 日May 22, 2018

我們以更好的搜尋工具取代了我們的 doc 網站搜尋。We have replaced our doc site search with a better search tool. 您將在未來幾周內看到更多搜尋改進。You’ll see even more search improvements over the coming weeks.

注意

如果您在部署新搜尋之後馬上試用,搜尋可能會中斷。Search may look broken if you try it shortly after the new search is deployed. 您只需清除瀏覽器快取,就能查看新的搜尋體驗。Just clear your browser cache to see the new search experience.

適用於 Machine Learning 的 Databricks Runtime 4.1 ML (Beta)Databricks Runtime 4.1 ML for Machine Learning (Beta)

2018 5 月17日May 17, 2018

Databricks Runtime ML (Beta 版) 提供適用于機器學習和資料科學的現成環境。Databricks Runtime ML (Beta) provides a ready-to-go environment for machine learning and data science. 它包含多個熱門的程式庫,包括 TensorFlow、Keras 和 XGBoost。It contains multiple popular libraries, including TensorFlow, Keras, and XGBoost.

Databricks Runtime ML 可讓您使用分散式 TensorFlow 訓練所需的所有程式庫來啟動 Databricks 叢集。Databricks Runtime ML lets you start a Databricks cluster with all of the libraries required for distributed TensorFlow training. 它可確保在 TensorFlow 和 CUDA/cuDNN 之間,叢集 (上包含的程式庫相容性,例如) ,而且相較于使用 init 腳本,它會大幅減少叢集的啟動時間。It ensures the compatibility of the libraries included on the cluster (between TensorFlow and CUDA / cuDNN, for example) and substantially decreases the cluster start-up time compared to using init scripts.

注意

Databricks Runtime 4.1 ML 僅適用于 Premium SKU。Databricks Runtime 4.1 ML is available only in the Premium SKU.

請參閱 Databricks Runtime 4.1 ML (不支援) 的完整版本資訊。See the complete release notes for Databricks Runtime 4.1 ML (Unsupported).

Databricks DeltaDatabricks Delta

2018 5 月17日May 17, 2018

Databricks Delta 現已在私人預覽中提供給 Azure Databricks 使用者使用。Databricks Delta is now available in Private Preview to Azure Databricks users. 請洽詢您的客戶經理或註冊 https://databricks.com/product/databricks-deltaContact your account manager or sign up at https://databricks.com/product/databricks-delta. 此版本代表預期即將推出 GA 版本的候選版。This release represents a candidate release in anticipation of the upcoming GA release.

如需詳細資訊,請參閱 Databricks Runtime 4.1 (不支援的) Delta Lake 和差異引擎指南For more information, see Databricks Runtime 4.1 (Unsupported) and Delta Lake and Delta Engine guide.

影像資料類型的 Display() 支援Display() support for image data types

2018 5 月17日May 17, 2018

在 Databricks Runtime 4.1 中,現在會將 display() 包含影像資料類型的資料行轉譯為 RICH HTML。In Databricks Runtime 4.1, display() now renders columns containing image data types as rich HTML.

請參閱 影像See Images.

GPU 叢集類型GPU cluster types

2018月15日:2.71 版May 15, 2018: Version 2.71

我們很高興宣佈 Azure Databricks 叢集上 (NC12 和 NC24) 的 Azure NC 實例類型支援。We’re pleased to announce support for Azure NC instance types (NC12 and NC24) on Azure Databricks clusters. NC 實例提供 Gpu 來處理電源影像處理、文字分析,以及其他具有運算挑戰和需求優異效能的機器學習和深度學習工作。NC instances provide GPUs to power image processing, text analysis, and other machine learning and deep learning tasks that are computationally challenging and demand superior performance.

Azure Databricks 也會提供針對 Gpu 設定的預先安裝 NVIDIA 驅動程式和程式庫,以及開始使用數個熱門深度學習程式庫的材質。Azure Databricks also provides pre-installed NVIDIA drivers and libraries configured for GPUs, along with material for getting started with several popular deep learning libraries.

另請參閱:See also:

秘密管理正式推出Secret management GA

2018月15日:2.71 版May 15, 2018: Version 2.71

秘密管理現已處於私人預覽狀態,現已正式推出。Secret management, which had been in private preview, is now GA. 它提供強大的工具,可用來管理您向外部資料源進行驗證所需的認證。It provides powerful tools for managing the credentials you need for authenticating to external data sources. 您可以使用 Databricks 秘密管理來儲存及參考筆記本和作業中的認證,而不是直接在筆記本中輸入您的認證。Instead of typing your credentials directly into a notebook, use Databricks secret management to store and reference your credentials in notebooks and jobs. 若要管理秘密,您可以使用 秘密 CLI 來存取 秘密 APITo manage secrets, you can use the Secrets CLI to access the Secrets API.

注意

密碼管理需要 Databricks Runtime 4.0 或更新版本,以及 Databricks CLI 0.7.1 或更新版本。Secret management requires Databricks Runtime 4.0 or above and Databricks CLI 0.7.1 or above.

請參閱 秘密管理See Secret management.

秘密 API 端點和 CLI 命令變更Secrets API endpoint and CLI command changes

2018月15日:2.71 版May 15, 2018: Version 2.71

已對秘密 API 端點進行下列變更:The following changes were made to the Secrets API endpoints:

  • 針對所有端點,根路徑已從變更 /secret/secretsFor all endpoints, the root path was changed from /secret to /secrets.
  • 若為秘密端點,則 /secret/secrets 已折迭為 /secrets/For the secrets endpoint, the /secret/secrets was collapsed to /secrets/.
  • write方法已變更為 putThe write method was changed to put.

Databricks CLI 0.7.1 包含秘密命令的更新,以配合這些更新的 API 端點。Databricks CLI 0.7.1 includes updates to Secrets commands to align with these updated API endpoints.

請參閱 秘密 API密碼管理See Secrets API and Secret management.

叢集釘選Cluster pinning

2018月15日:2.71 版May 15, 2018: Version 2.71

您現在可以將叢集釘選到群集清單。You can now pin a cluster to the Clusters list. 這可讓您保留在過去30天內終止的叢集設定。This lets you retain the configuration of clusters terminated over 30 days old.

釘選叢集Pin cluster

此外,[叢集] 頁面現在會顯示在30天內終止的所有叢集 (從7天) 增加。In addition, the Clusters page now displays all clusters that were terminated within 30 days (increased from 7 days).

請參閱 選叢集。See Pin a cluster.

叢集自動啟動Cluster autostart

2018月15日:2.71 版May 15, 2018: Version 2.71

在此版本之前,排程在叢集上執行的作業 Terminated 失敗。Before this release, jobs scheduled to run on Terminated clusters failed. 針對在 Azure Databricks 2.71 版和更新版本中建立的叢集,從 JDBC/ODBC 介面或指派給現有終止叢集的作業執行的命令會自動重新開機該叢集。For clusters created in Azure Databricks version 2.71 and above, commands from a JDBC/ODBC interface or a job run assigned to an existing terminated cluster automatically restarts that cluster. 請參閱 JDBC connect建立作業See JDBC connect and Create a job.

自動啟動可讓您將叢集設定為自動終止,而不需要手動介入,即可重新開機排程工作的叢集。Autostart allows you to configure clusters to autoterminate, without requiring manual intervention to restart the clusters for scheduled jobs. 此外,您可以排程在指定時間重新開機已終止叢集的工作,以排程叢集初始化。Furthermore, you can schedule cluster initialization by scheduling a job that restarts terminated clusters at a specified time.

系統會強制執行叢集存取控制,並照常檢查作業擁有者許可權。Cluster access control is enforced and job owner permissions are checked as usual.

工作區清除Workspace purging

2018月15日:2.71 版May 15, 2018: Version 2.71

為了符合歐盟一般資料保護規定 (GDPR) 的持續努力,我們新增了清除工作區物件的功能,例如整個筆記本、個別筆記本資料格、個別筆記本批註,以及筆記本修訂歷程記錄。As part of our ongoing effort to comply with the European Union General Data Protection Regulation (GDPR), we have added the ability to purge workspace objects, such as entire notebooks, individual notebook cells, individual notebook comments, and notebook revision history. 我們將在未來幾周發行更多的功能和檔,以支援 GDPR 合規性。We will release more functionality and documentation to support GDPR compliance in the coming weeks.

請參閱 管理工作區儲存體See Manage workspace storage.

Databricks CLI 0.7.1Databricks CLI 0.7.1

2018月10日May 10, 2018

Databricks CLI 0.7.1 包含秘密命令的更新,以配合更新的 API 端點。Databricks CLI 0.7.1 includes updates to Secrets commands to align with updated API endpoints.

請參閱 DATABRICKS CLI密碼管理See Databricks CLI and Secret management.