Zum Inhalt springen

Databricks Release Notes

Einträge
1.189
Quellen
1
Zuletzt aktualisiert

Folge Databricks, um die Release Notes in deinen Feed zu holen.

Databricks

Genie: mehr Entity Matching und SQL-Editor-Korrektur

Genie unterstützt nun bis zu 60 Entity-Matching-Spalten, und der SQL-Editor im Chat verschwindet nicht mehr, wenn die generierte Abfrage gelöscht wird.

  • More entity matching in Genie: Genie can now support up to 60 entity matching (previously value dictionary) columns. See Tune Genie Agent quality.

  • Genie Agent (formerly Genie Space) SQL editor fix: Resolved an issued that caused the SQL editor in a chat window disappear if the user deleted the generated query. You can now use the editor to delete and rewrite queries.

Originalquelle(öffnet in neuem Tab)Problem melden

Databricks

Query-Insights: neue Events für Query History und Query Profile

Der Besuch der Seite Query History löst nun das Event listHistoryQueries aus, das Öffnen eines Query Profiles das Event getHistoryQuery.

  • Query insights improvements: Visiting the Query History page now emits the listHistoryQueries event. Opening a query profile now emits the getHistoryQuery event.

Originalquelle(öffnet in neuem Tab)Problem melden

Databricks

Salesforce Data Cloud File Sharing Connector (Public Preview)

Der neue Connector bietet eine Zero-Copy-Alternative zur Query Federation, bei der Databricks Daten über Salesforce-DaaS-APIs direkt aus dem Cloud-Objektspeicher liest, ohne JDBC und mit besserer Performance bei großen Datenmengen.

The Salesforce Data Cloud File Sharing connector offers an alternative zero-copy solution for querying Salesforce Data Cloud. When you use file federation instead of query federation, Databricks calls Salesforce Data-as-a-Service (DaaS) APIs to read data in the underlying cloud object storage location directly. Queries are run on Databricks compute without using JDBC.

Compared to query federation, file federation is ideal for federating a large amount of data. It offers improved performance for reading files from multiple data sources and better pushdown capabilities.

For more information, see Lakehouse Federation for Salesforce Data 360 File Sharing.

Originalquelle(öffnet in neuem Tab)Problem melden

Databricks

Korrekte job_name-Werte in system.billing.usage

Das Feld usage_metadata.job_name in system.billing.usage enthält nun die vom Nutzer vergebenen Jobnamen statt der Task Keys, außer bei einmaligen Job-Läufen, die weiterhin den Task Key tragen.

The usage_metadata.job_name value in the system.billing.usage table now correctly contains job names. Previously, this value was populated with task keys instead of the user-provided job names. This change does not apply to one-time job runs, which continue to be logged with the task key.

See Billable usage system table reference.

Originalquelle(öffnet in neuem Tab)Problem melden

Databricks

Mosaic AI Vector Search: storage-optimierte Endpunkte (Public Preview)

Mosaic AI Vector Search bietet storage-optimierte Endpunkte mit größerer Kapazität (über eine Milliarde Vektoren bei Dimension 768), 10–20x schnellerer Indexierung und auf viele Vektoren optimierter Preisgestaltung.

Mosaic AI Vector Search now offers the option of storage-optimized endpoints. Storage-optimized endpoints have a larger capacity (over one billion vectors at dimension 768) and provide 10-20x faster indexing. Pricing is optimized for the larger number of vectors. For details, see Vector Search endpoint options.

Originalquelle(öffnet in neuem Tab)Problem melden

Databricks

History Sharing standardmäßig aktiv für Delta Sharing

History Sharing ist für Databricks-to-Databricks Delta Sharing ab Databricks Runtime 16.2 standardmäßig aktiviert, um die Leseleistung von Tabellen zu verbessern.

History sharing is enabled by default (for Databricks Runtime 16.2 and above) to improve table read performance for Databricks-to-Databricks Delta Sharing. See Improve table read performance with history sharing.

Originalquelle(öffnet in neuem Tab)Problem melden

Databricks

Unity Catalog HTTP-Verbindungen mit OAuth User-to-Machine Per User (Preview)

Bei Unity Catalog HTTP-Verbindungen können sich einzelne Nutzer nun mit eigenen OAuth-Zugangsdaten bei externen Diensten anmelden, statt dass ein gemeinsamer OAuth-Credential genutzt wird.

Individual users can now use their own OAuth credentials to sign into external services when using a Unity Catalog HTTP connection. Previously, the only OAuth option was to share a single OAuth credential for the connection regardless of the user.

See Connect to external HTTP services.

Originalquelle(öffnet in neuem Tab)Problem melden

Databricks

sourceIpAddress in Audit Logs ohne Portnummer

Bei bestimmten Authentifizierungs- und Autorisierungs-Audit-Logs wird die fälschlich angehängte Portnummer (0) im Feld sourceIPAddress ab Anfang August 2024 schrittweise entfernt.

Due to a bug, certain authorization and authentication audit logs include a port number in addition to the IP in the sourceIPAddress field (for example, "sourceIPAddress":"10.2.91.100:0"). The port number, which is logged as 0, does not provide any real value and is inconsistent with the rest of the Databricks audit logs. To enhance the consistency of audit logs, Databricks plans to change the format of the IP address for these audit log events. This change will gradually roll out starting in early August 2024.

If the audit log contains a sourceIpAddress of 0.0.0.0, Databricks might stop logging it.

Originalquelle(öffnet in neuem Tab)Problem melden

Databricks

Metrik-Ansichten in Public Preview

Unity Catalog-Metrik-Ansichten sind jetzt in der Public Preview verfügbar und bieten eine zentrale Definition wiederverwendbarer und verwalteter Geschäftskennzahlen für Berichte und Warnungen.

  • Unity Catalog metric views are now in Public Preview: Metric views provide a centralized way to define and manage consistent, reusable, and governed core business metrics. They abstract complex business logic into a centralized definition, enabling organizations to define key performance indicators once and use them consistently across reporting tools like dashboards, Genie Agents (formerly Genie Spaces), and alerts. Use a SQL warehouse running on the Preview channel (2025.16) to work with metric views. See Unity Catalog metric views.

Originalquelle(öffnet in neuem Tab)Problem melden

Databricks

Dashboards: Catalog-/Schema-Auswahl im Dataset-Editor und Filterfix

Im Data-Tab eines Dashboard-Entwurfs lassen sich Catalog und Schema auswählen, sodass Abfragen nur den Tabellennamen verwenden können, und Vollbild-Widgets zeigen mit globalen Filtern wieder Daten an.

  • Catalog and schema picker in the dataset editor: You can now select a catalog and schema from the SQL editor in the Data tab of a draft dashboard. With the catalog and schema selected, you can write queries that reference a table name only, instead of a fully-qualified table name. You can still use a table's three-level namespace to reference it in a query.

Catalog and schema picker in the AI/BI dashboard dataset editor.

  • Fix for global filters on full-screen widgets: An issue that prevented full-screen widgets from displaying data when used with global filters has been resolved.

Originalquelle(öffnet in neuem Tab)Problem melden

Databricks

Genie Knowledge Store, Query-Vorschläge und Fehlerbehebungen

Genie erhält einen Knowledge Store mit Format Assistance und Bearbeitung von Spaltenbeschreibungen auf Space-Ebene, schlägt populäre Abfragen als Beispiel-SQL vor, vereinfacht die Bereitstellung von Tabellenmetadaten und behebt einen Fehler bei der Benchmark-Ausführung.

  • Benchmark execution bug fix in Genie: A bug where benchmarks would not execute immediately after creating a benchmark run has been fixed.

  • Popular query suggestions in Genie: Genie will now suggest popular queries as example SQL for newly added tables to the space. Only popular queries that the author has access to are suggested. See Create a Genie Agent.

  • Table metadata improvement in Genie: The process to provide table metadata to Genie has been changed to simplify data transformation for Genie workloads.

  • Genie knowledge store: The knowledge store allows authors to build more reliable, accurate Genie Agents (formerly Genie Spaces). Authors can now annotate, edit, and manage Genie datasets at the space level using the following features:

    • Increased accuracy with format assistance: Genie now provides format assistance (previously example values) when you add tables to a space, helping it better understand the data in your tables and generate more accurate responses. See Edit column metadata.
    • Edit column descriptions in Genie: You can now edit table and column descriptions at the Genie-space level. This allows you to fine-tune metadata descriptions within a specific the Genie Agent (formerly Genie Space). See Edit column metadata. …

Originalquelle(öffnet in neuem Tab)Problem melden

Databricks

Dashboards: Filter-Erweiterungen und Visualisierungsverbesserungen

Dashboards erhalten unter anderem das Ausschließen ausgewählter Werte in Multi-Select-Filtern, ein globales Filter-Panel, kompaktere PDF-Snapshots ohne Leerraum und konstante Referenzlinien in kartesischen Diagrammen.

  • Exclude selected values in filter widgets: Dashboard viewers can now include or exclude the values they select from multi-select filter widgets.

A point map visulization showing average number of bedrooms by price. A filter with multiple values selected excludes selected values from the chart.

  • Drop-down selection fix: Fixed an issue where selection options in a drop-down list appeared out of order.

  • Global filters panel: Dashboard authors can now add filters to a global filters panel, which applies across all dashboard pages.

  • Relative date and time preview improvement: Previews for relative date and time controls are now hidden to avoid confusion caused by time zone differences.

  • Dashboard PDF snapshots improvement: Dashboard PDF snapshots now wrap the content of the dashboard and no longer include white space.

  • Constant reference line annotations: Horizontal and vertical constants reference line annotations for quantitative axes in cartesian chart types, such as area, bar, line, scatter, heatmap, and histogram charts.

Originalquelle(öffnet in neuem Tab)Problem melden

Databricks

Genie: Usage Guidance und schnellere Ergebnisanzeige

In Genie lässt sich Usage Guidance für SQL-Beispiele ergänzen, Abfrageergebnisse erscheinen schneller mit Statusanzeigen wie Running query und Still thinking, und leere Visualisierungen werden nicht mehr gelegentlich angezeigt.

  • Usage guidance addition in Genie: You can now add Usage Guidance to clarify how and when SQL examples should be used.

Usage guidance text entry.

  • Improved query result display in Genie: Query results now show more quickly. It now displays Running query until the query result is ready and Still thinking if additional work is required.

  • UI polling update in Genie: The UI now starts polling for result data sooner rather than waiting until after the visualization and sample questions are generated.

  • SQL function de-emphasis in Genie: The option to add SQL functions has been de-emphasized, making adding example SQL the default action.

  • Improved result delivery timing in Genie: Descriptions are now generated after the SQL is run, thereby reducing the time it takes to get results.

  • Fixed empty visualizations issue in Genie: Fixed an issue where empty visualizations were occasionally rendering in Genie.

Originalquelle(öffnet in neuem Tab)Problem melden

Databricks

Neue Alert-Version in Beta

Eine neue Alert-Version in Beta bündelt Abfrage, Bedingungen, Zeitpläne und Benachrichtigungsziele in einer Oberfläche, wobei Legacy-Alerts weiterhin parallel nutzbar sind.

A new version of alerts is now in Beta. This version simplifies creating and managing alerts by consolidating query setup, conditions, schedules, and notification destinations into a single interface. You can still use legacy alerts alongside the new version. See Databricks SQL alerts.

Originalquelle(öffnet in neuem Tab)Problem melden

Databricks

Dashboards: Genie-Abfragen, Berechnungen und Zeitplan-Unterstützung

Genie in Dashboards fragt nur noch Dashboard-Datasets ab, benutzerdefinierte Berechnungen unterstützen isnull, isnotnull und den null-Operator, Zeitpläne funktionieren auch ohne eingebettete Credentials, und Tooltips sowie Zahlenformatierung wurden verbessert.

  • Genie in dashboards query update: Genie in dashboards will now only query from dashboard datasets, instead of the underlying tables within dashboard datasets.

  • Custom calculation support in dashboards: Added support for custom calculations using isnull, isnotnull functions, and the null operator.

  • AI/BI dashboard schedule support: AI/BI dashboard schedules now support dashboards without embedded credentials.

  • Dashboard tooltip and number formatting fixes:

    • Tooltip fix: Fixed tooltips to display an empty value when the field value is an empty string ("").
    • Number formatting support: Number formatting in table and pivot table cells from notebooks, SQL editor, and legacy dashboards can now be carried over to AI/BI dashboards.

Originalquelle(öffnet in neuem Tab)Problem melden

Databricks

Databricks SQL Version 2025.16

Databricks SQL Version 2025.16 ist jetzt im Preview-Channel verfügbar und bringt neue Funktionen, Verhaltensänderungen und Fehlerbehebungen.

Databricks SQL version 2025.16 is now available in the Preview channel. Review the following section to learn about new features, behavioral changes, and bug fixes.

Originalquelle(öffnet in neuem Tab)Problem melden

Databricks

Dashboards: Mehrfachauswahl von Daten, Point Maps per KI, Bundles

Im Dialog „Add data“ lassen sich mehrere Tabellen und Views gleichzeitig hinzufügen, Point Maps können per Prompt generiert werden, Dashboard-Tasks sind in Declarative Automation Bundles nutzbar, und Entwurfs-Dashboards lassen sich nicht mehr per API planen.

  • Select and add multiple tables and views: Users can now select and add multiple tables and views at once from the Add data dialog.

  • Draft dashboard scheduling bug fix: Fixed a bug that allowed draft dashboards to be scheduled using the API. Now, only published dashboards can be scheduled.

  • AI-assisted point maps in dashboards: You can now use natural-language prompts to generate point map charts in dashboards.

  • Deploy dashboard tasks using Declarative Automation Bundles: You can now use dashboard tasks within a job resource when deploying bundles. To learn more about Declarative Automation Bundles, see What are Declarative Automation Bundles?. To learn more about programmatically configuring a dashboard task in a job, see the Jobs API.

Originalquelle(öffnet in neuem Tab)Problem melden

Databricks

Genie: Mehr Werte beim Entity Matching und Point Maps

Entity-Matching-Spalten in Genie unterstützen nun 1024 statt 255 Werte pro Spalte, und Point Maps werden in Genie unterstützt.

  • Increased entity matching support in Genie: Each entity matching (previously value dictionary) column in Genie now supports 1024 values per column, up from 255. Click Refresh on your existing entity matching columns to increase cardinality support. See Tune Genie Agent quality.

  • Point maps support in Genie: Point maps are now supported in Genie.

Originalquelle(öffnet in neuem Tab)Problem melden

Databricks

Dashboards: Mehr Daten bei abgeschnittenen Diagrammen anzeigen

Bei Diagrammen mit abgeschnittenen Ergebnissen lassen sich über „Render more data“ mehr Daten anzeigen.

  • View truncated data: You can now view more data in charts with truncated results by clicking Render more data.

Originalquelle(öffnet in neuem Tab)Problem melden

Databricks

Genie: Folgefragen bei Textantworten und bessere Anweisungsverarbeitung

Genie schlägt jetzt auch bei Textantworten ohne SQL oder Ergebnis Folgefragen vor und berücksichtigt relevante Anweisungen anhand des Prompts besser.

  • Follow-up questions for text responses in Genie: Genie now suggests follow-up questions for text responses that include no generated SQL or results. Previously, follow-up questions appeared only after responses with a result set. Now, prompts such as Explain the data set also return suggested follow-up questions to help Genie Agent (formerly Genie Space) users explore the data.

Follow-up questions are suggested for a text-based response

  • Improved instruction processing in Genie: Genie's chain-of-thought reasoning has been refined to prioritize and adhere to the most relevant instructions based on the user prompt.

Originalquelle(öffnet in neuem Tab)Problem melden