2020 年 1 月January 2020

這些功能和 Azure Databricks 平臺改進已于2020年1月發行。These features and Azure Databricks platform improvements were released in January 2020.

注意

發行是暫存的。Releases are staged. 在初始發行日期之後,您的 Azure Databricks 帳戶可能不會更新到一周。Your Azure Databricks account may not be updated until up to a week after the initial release date.

本月已看到 Azure Databricks 平臺版本3.9 和3.11 的版本。This month saw the release of Azure Databricks platform versions 3.9 and 3.11. 沒有版本3.10 或3.8。There was no release of versions 3.10 or 3.8. 3.7 版是穩定性和 bug-僅限修正版本。Version 3.7 was a stability and bug-fix-only release.

即將推出:工作區、集區和叢集標記會傳播至 DBU 使用量詳細資料和 Azure VM,以提供更佳的成本管理報告Coming soon: workspace, pool, and cluster tags propagate to DBU usage details and Azure VMs for better cost management reporting

我們將于2月10日發行標記傳播,以 Azure Databricks 使用量詳細資料和 Azure Vm。On February 10th, we will release tag propagation to Azure Databricks usage details and Azure VMs. 新的標記傳播功能結合了 Azure Databricks 工作區標籤, (也就是資源群組標籤) 、集區標籤和叢集標籤,並將它們傳播至 Databricks DBU 使用量詳細資料和 Azure Vm 作為資源標記。The new tag propagation feature combines Azure Databricks workspace tags (that is, resource group tags), pool tags, and cluster tags and propagates them to the Databricks DBU usage details and Azure VMs as resource tags. 您將能夠在 Azure 成本管理入口網站和使用量詳細資料匯出中看到合併的標記資訊,讓您更清楚地瞭解 Azure Databricks 的使用量 (擁有權總成本) 以及商務單位和團隊的精確屬性。You will be able to see the combined tag information in the Azure Cost Management portal and in usage detail exports, giving you better visibility into Azure Databricks usage (total cost of ownership) and accurate attribution to business units and teams.

Azure Databricks 和 Azure Lighthouse 現在可以在同一個訂用帳戶中上線Azure Databricks and Azure Lighthouse can now live in the same subscription

2020年1月29日January 29, 2020

所有現有的 Azure Databricks 工作區都已從使用受控鎖定遷移到 拒絕指派All existing Azure Databricks workspaces have migrated from using Managed Locks to Deny Assignments. 所有建立的新工作區都會有拒絕指派。All new workspaces created will have Deny Assignments. 這不會變更任何現有的行為,而且安全性層級會維持不變。This does not change any existing behavior, and the level of security remains the same. 雖然您可以將使用 Azure Databricks 的訂用帳戶上線,但管理租使用者中的使用者目前無法在委派的訂用帳戶上啟動 Azure Databricks 工作區。While you can onboard subscriptions that use Azure Databricks, users in the managing tenant can’t launch Azure Databricks workspaces on a delegated subscription at this time.

適用於 Genomics GA 的 Databricks Runtime 6.3Databricks Runtime 6.3 for Genomics GA

2020 年 1 月 22 日January 22, 2020

Genomics 的 Databricks Runtime 6.3 是以 Databricks Runtime 6.3 為基礎。Databricks Runtime 6.3 for Genomics is built on top of Databricks Runtime 6.3. 它包含從 Databricks Runtime 6.2 Genomics 的許多增強功能和升級。It includes many improvements and upgrades from Databricks Runtime 6.2 for Genomics.

主要功能包括:The key features are:

  • 支援 Delta 資料表作為聯合基因型分型管線的輸入Support for Delta tables as input to the joint genotyping pipeline
  • 讀取 VCFs 時自動注釋剖析Automatic annotation parsing when reading VCFs
  • 改良的 multiallelic variant 分隔器Improved multiallelic variant splitter
  • 更快速的線性和羅吉斯回歸函數Faster linear and logistic regression functions

如需詳細資訊,請參閱 Genomics 的完整 Databricks Runtime 6.3 (不支援的) 版本資訊。For details, see the complete Databricks Runtime 6.3 for Genomics (Unsupported) release notes.

Databricks Runtime 6.3 ML GADatabricks Runtime 6.3 ML GA

2020 年 1 月 22 日January 22, 2020

Databricks Runtime 6.3 ML GA 引進了許多程式庫升級,包括:Databricks Runtime 6.3 ML GA brings many library upgrades, including:

  • PyTorch:1.3.0 至1.3。1PyTorch: 1.3.0 to 1.3.1
  • torchvision:0.4.1 至0.4。2torchvision: 0.4.1 to 0.4.2
  • MLflow:1.4.0 至1.5。0MLflow: 1.4.0 to 1.5.0
  • Hyperopt:0.2.1 至0.2。2Hyperopt: 0.2.1 to 0.2.2

如需詳細資訊,請參閱 Machine Learning (不支援的完整 Databricks Runtime 6.3) 版本資訊。For details, see the complete Databricks Runtime 6.3 for Machine Learning (Unsupported) release notes.

Databricks Runtime 6.3 GADatabricks Runtime 6.3 GA

2020 年 1 月 22 日January 22, 2020

Databricks Runtime 6.3 GA 引進了新功能、增強功能以及許多 bug 修正。Databricks Runtime 6.3 GA brings new features, improvements, and many bug fixes.

此版本引進改良的平行存取。This release introduces improved concurrency. 主要功能包括:The key features are:

  • 改善所有 Delta Lake 作業的並行Improved concurrency for all Delta Lake operations
  • 改進對壓縮檔案壓縮的支援Improved support for file compaction
  • 改善僅限插入合併的效能Improved performance for insert-only merge

如需詳細資訊,請參閱完整的 Databricks Runtime 6.3 (不支援的) 版本資訊。For details, see the complete Databricks Runtime 6.3 (Unsupported) release notes.

預設會啟用差異快取Delta cache enabled by default

2020年1月7-14 日:版本3。9January 7-14, 2020: Version 3.9

針對所有支援的 Databricks Runtime 版本,現在預設會在 Lsv2 系列實例上啟用差異快取。The Delta cache is now enabled by default on Lsv2 series instances for all supported Databricks Runtime releases. 請參閱 使用 Delta cachingSee Use Delta caching.

現在可設定叢集標準自動調整步驟Cluster standard autoscaling step is now configurable

2020年1月7-14 日:版本3。9January 7-14, 2020: Version 3.9

標準自動調整的第一個步驟預設會新增8個節點。By default the first step of standard autoscaling adds 8 nodes. 現在您可以在叢集 Spark 設定中設定步驟值。Now you can set the step value in the cluster Spark configuration. 請參閱 標準自動調整。See Standard autoscaling.

SCIM API 支援「取得使用者」和「取得群組」的分頁 (公開預覽)SCIM API supports pagination for Get Users and Get Groups (Public Preview)

2020年1月7-14 日:版本3。9January 7-14, 2020: Version 3.9

SCIM API 現在支援取得使用者和取得群組的分頁。The SCIM API now supports pagination for Get Users and Get Groups. 當您指定 startIndexcount 查詢參數時,SCIM 會傳回使用者/群組的子集。When you specify the startIndex and count query parameters, SCIM will return a subset of users/groups. startIndex參數是第一個結果的以1為基礎的索引。The startIndex parameter is the 1-based index of the first result. count參數是要傳回之使用者或群組的最大數目。The count parameter is the maximum number of users or groups to return. 這可確保 SCIM 用戶端的擴充性,並簡化 Azure Databricks 系統管理員的 SCIM 呼叫。This ensures scalability for the SCIM Client and simplifies SCIM calls for Azure Databricks admins. 請參閱 SCIM APISee SCIM API.

檔案瀏覽器區隔線寬度已增加至 240pxFile browser swimlane widths increased to 240px

2020年1月7-14 日:版本3。9January 7-14, 2020: Version 3.9

增加的寬度可減少滑鼠停留物件的需求,以查看完整的檔案名。The increased width reduces the need to mouse over objects to see the full filename.

Databricks Runtime 3.5 LTS 支援結束Databricks Runtime 3.5 LTS support ends

2020年1月2日January 2, 2020

支援 Databricks Runtime 3.5 LTS (長期支援) 于1月2日結束。Support for Databricks Runtime 3.5 LTS (Long Term Support) ended on January 2. 請參閱 Databricks 執行時間支援生命週期See Databricks runtime support lifecycle.