โ™Š๏ธ GemiNews ๐Ÿ—ž๏ธ (dev)

Demo 1: Embeddings + Recommendation Demo 2: Bella RAGa Demo 3: NewRetriever Demo 4: Assistant function calling

๐Ÿ—ž๏ธMarch 13, 2024

๐Ÿ—ฟSemantically Similar Articles (by :title_embedding)

March 13, 2024

2024-03-13 - (from GCP latest releases)

AlloyDB for PostgreSQL Feature AlloyDB now supports continuous backup and recovery, and scheduled backups on secondary clusters. When you create a secondary cluster, any backup plans on the primary cluster are automatically copied to the new secondary cluster. For more information, see About cross-region replication. Anti Money Laundering AI Announcement Released a new v4 engine versions for the commercial line of business, with more reliable tuning performance, in particular for small datasets. Apigee X Announcement As of March 13, 2024, the conversion of Apigee API Management organizations with Pay-as-you-go pricing provisioned before October 1, 2023, to Pay-as-you-go organizations that use updated attributes for pricing is complete, with the exception of one organization that requires customer action. The Apigee API Analytics add-on is enabled in converted organizations.The Analytics add-on can be disabled if it is not required. In addition, you can update your Pay-as-you-go environment types using the API. For more information on the updated pricing and enhanced features now available for these organizations, see Pay-as-you-go (updated attributes) overview. Updated pricing attributes will be reflected in March invoices. For billing questions related to this change, contact Google Cloud Billing support. Chronicle SOAR Announcement Release 6.2.51 is currently in Preview. Feature Jobs Enhancement When updating an integration, the jobs will now be updated automatically. This does not apply to any legacy jobs that were created before October 2023. The Marketplace integration will clearly identify the legacy jobs that are affected and provide instructions on how to proceed. In addition, legacy jobs are now marked as such in the Jobs Scheduler page so that you can take action and resolve issues beforehand. Changed APIs now documented The following APIs are not new, but with this Release are now formally documented in Swagger: AddOrUpdateEnvironmentRecords RemoveEnvironmentRecords Fixed Searching for cases from the last week doesn't produce results (ID #00269819) Fixed Email HTML Templates > Show Email Template not rendering styles (ID #00249556) Fixed SDK call for create entity failure displays the wrong error message (ID #48950075) Cloud Billing Feature You can now view granular Bigtable usage in the Cloud Billing Detailed export to BigQuery You can now view granular Bigtable instance cost data in the Google Cloud Billing detailed export. Use the resource.global_name field in the export to view and filter your detailed Bigtable instance usage. Review the schema of the Detailed cost data export. Feature You can now view granular Memorystore for Redis usage in the Cloud Billing Detailed export to BigQuery You can now view granular Memorystore for Redis cost data in the Google Cloud Billing detailed export. Use the resource.global_name and resource.name fields in the export to view and filter your detailed Memorystore for Redis usage. Review the schema of the Detailed cost data export. Cloud Composer Deprecated All preview versions of Cloud Composer 2 are past their security notifications end date and are not supported. If your environment uses a preview version of Cloud Composer 2, then please upgrade this environment to a supported version or re-create it using the latest version of Cloud Composer 2. Issue If you see frequent scheduler restarts in your Airflow 2.6.3 environment and the [scheduler]job_heartbeat_sec Airflow configuration option is set to a non-default value, you can fix this issue either by upgrading to Cloud Composer version 2.6.4 or by removing this option's override, so that it uses the default value. Cloud SQL for SQL Server Feature Cloud SQL now supports SQL Server Reporting Services (SSRS) on your instances. For more information, see Use SSRS for creating reports. Compute Engine Feature Generally available: You can use SSH-in-browser to connect to TPU VMs. For more information, see Connecting to a Cloud TPU. Dataform Announcement The 3.0.0-beta.0 version of the open-source Dataform framework is available. This update introduces significant changes, including, but not limited to, the following: Deprecation of dataform.json in favor of workflow_settings.yaml Stateless package installation by @dataform/cli Warehouse-agnostic compilation output You don't need to take immediate action to update your Dataform code. For more information, see the 3.0.0-beta.0 release on GitHub. Looker Announcement Looker 24.4 includes the following changes, features, and fixes. Expected Looker (original) deployment start: Monday, March 18, 2024 Expected Looker (original) final deployment and download available: Thursday, March 28, 2024 Expected Looker (Google Cloud core) deployment start: Monday, March 18, 2024 Expected Looker (Google Cloud core) final deployment: Monday, April 1, 2024 Breaking Query IDs can no longer be used to fetch queries or create render tasks through the API. The Get All Running Queries API endpoint is now restricted to admins only. Query slugs that are generated by Looker will be 32 characters instead of 7. Breaking Chrome is starting to deprecate third-party cookies as of January 2024. Because of Looker's dependency on third-party cookies to establish embed user sessions, this may impact your embed use case. For more information, see the Chrome is deprecating third-party cookies notice. Breaking Previously, custom visualizations would not call the updateAsync function after the vis config is updated via the custom visualization API. Now, the function will be called. If a custom visualization is set up to update the vis config every time updateAsync is called, it could cause excessive refreshes. If your custom visualization is fails to load after this update, double check your custom visualization code for unnecessary vis config updates. If you have a Looker (original) instance, you can also enable the Custom Vis Reliable Render Labs feature which causes Looker to suppress excess refresh behavior in custom visualizations. Feature The Performant Field Picker feature is now generally available. Feature When an instance has no projects, Looker will more prominently prompt users to create a model. Feature In the Create a model wizard, your selections are now saved even if you close steps without having completed the model creation process. Feature Adding a query slug to source queries in the merge query API response GET merge_queries/<merge_query_id> returns the query slug in addition to the ID. Feature The save_content permission now has two child permissions, save_dashboards and save_looks. These permissions let Looker admins exert finer control over the kinds of content that users can save. Fixed Only users who have access to dashboard extensions will be shown the Add Extension tile. Fixed Subtotals have been fixed for queries with order_by_field references in query streaming pathways. This feature now performs as expected. Fixed An issue where embed secrets might have been visible to non-admin users has been fixed. This feature now performs as expected. Fixed Looker now ignores all blank filter strings, including IS NOT. Fixed An issue has been fixed that caused small decimals to be displayed in scientific notation even when formatting was disabled. This feature now performs as expected. Fixed An issue has been fixed where the PDT Context Override toggle was improperly reflecting the ON state when it had been cleared prior. This feature now performs as expected. Fixed Performant field picker sorting behavior has been fixed. This feature now performs as expected. Fixed Downloading results from SQL Runner now only downloads the file and does not open the file in a new browser tab. Fixed Filter expressions including user attributes and OR logical conditions were being incorrectly populated when generating SQL. This feature now performs as expected. Fixed A change in the Snowflake dialect was ported to Kotlin to maintain parity. Snowflake column names with mixed cases are now properly quoted. Fixed Filter suggestion requests have been reduced while the user is typing. Because normal typing will invoke fewer requests, the load on the server will be reduced. Fixed An issue that caused single value change indicators to not render in Safari when dashboards scrolled has been fixed. This feature now performs as expected. Fixed The LookML Validator no longer hangs on a connection that references a deleted or malformed user attribute. The Validator also surfaces a detailed error when the user tests the connection. Fixed An issue has been fixed where extension documents would appear when hiding Looker document links was disabled. This feature now performs as expected. Fixed Content Validator has added support for field replacement within custom measure filters (across Looks, dashboard elements, and merge queries). Fixed Queries with order_by_field references and subtotals should render correctly in downloads / run_query APIs. Fixed Looker should now correctly handle cases where the sorts query had an empty string or was entirely empty. Fixed Previously, the All Results option was unavailable for schedules on Looks. This feature now performs as expected. Fixed On the new Admin - Users page, Looker Support users were shown as having never logged in even for currently logged-in users. This issue has been fixed and this feature now performs as expected. Fixed LookML-defined fields that are used in field filters will not be rejected from a set when the field requiring them is rejected from that set. This feature now performs as expected. Fixed Previously, the Errors and Broken Content dashboard appeared twice in the admin panel. This feature now performs as expected. Fixed A data validation message is now returned for waterfall charts when there are multiple measures and a hidden dimension. Fixed Looker now shows a clearer warning message when a user attempts to download a query with dimension fill and All Results enabled. Fixed Looker no longer imposes the Explore row limit of 5,000 on queries that are run using the run inline query API endpoint. Fixed Previously, the lookml_model_explore API endpoint would return a 500 error in certain cases. This feature now performs as expected. Fixed Errors about UI downloads are now more descriptive, similar to descriptive API error messages. Fixed Internal encryption has been migrated from AES-128 to AES-GCM-256 encryption. Feature The Disallow Numeric Query IDs legacy feature has been added to let users opt in to or out of query API changes. Feature The Advanced Features for New Schedules Page Labs feature is now available. This lets you sort and filter the list of scheduled plans on the Admin - Schedules page. Fixed Previously, when a dashboard was scheduled using PNG format and one of the tiles contained an empty note, the schedule would fail. This feature now works as expected. Fixed The Export function has been re-enabled, which lets Looker admins export data from a Looker (original) instance for import into a Looker (Google Cloud core) instance. Fixed Incorrect quoting in Snowflake views has been fixed. Fixed IAM checks for ephemeral users were disabled as a result of rendering issues for users who were logged in with SAML in Looker (Google Cloud core). Resource Manager Feature You can add tags at the time of creating folders and projects. These tags can be added as key-value pairs. For more information, see Add tags during folder creation and Add tags during project creation. This feature is currently in preview. Storage Transfer Service Feature Support for transfers from cloud and on-premises Hadoop Distributed File System (HDFS) sources is now generally available (GA). HDFS support allows for use cases such as migrating from on-premises storage to Cloud Storage, archiving data to free up on-premises storage space, replicating data to Google Cloud for business continuity, or transferring data to Google Cloud for analysis and processing. See Transfer from HDFS to Cloud Storage for details.

[Technology] ๐ŸŒŽ https://cloud.google.com/release-notes#March_13_2024 [๐Ÿง ] [v2] article_embedding_description: {:llm_project_id=>"Unavailable", :llm_dimensions=>nil, :article_size=>17887, :llm_embeddings_model_name=>"textembedding-gecko"}
[๐Ÿง ] [v1/3] title_embedding_description: {:ricc_notes=>"[embed-v3] Fixed on 9oct24. Only seems incompatible at first glance with embed v1.", :llm_project_id=>"unavailable possibly not using Vertex", :llm_dimensions=>nil, :article_size=>17887, :poly_field=>"title", :llm_embeddings_model_name=>"textembedding-gecko"}
[๐Ÿง ] [v1/3] summary_embedding_description:
[๐Ÿง ] As per bug https://github.com/palladius/gemini-news-crawler/issues/4 we can state this article belongs to titile/summary version: v3 (very few articles updated on 9oct24)

๐Ÿ—ฟarticle.to_s 

------------------------------
Title: March 13, 2024
[content]
AlloyDB for PostgreSQL
Feature
AlloyDB now supports continuous backup and recovery, and scheduled backups on secondary clusters. When you create a secondary cluster, any backup plans on the primary cluster are automatically copied to the new secondary cluster. For more information, see About cross-region replication.
Anti Money Laundering AI
Announcement
Released a new v4 engine versions for the commercial line of business, with more reliable tuning performance, in particular for small datasets.
Apigee X
Announcement
As of March 13, 2024, the conversion of Apigee API Management organizations with Pay-as-you-go pricing provisioned before October 1, 2023, to Pay-as-you-go organizations that use updated attributes  for pricing is complete, with the exception of one organization that requires customer action.

The Apigee API Analytics add-on is enabled in converted organizations.The Analytics add-on can be disabled if it is not required. In addition, you can update your Pay-as-you-go environment types using the API.

For more information on the updated pricing and enhanced features now available for these organizations, see Pay-as-you-go (updated attributes) overview.

Updated pricing attributes will be reflected in March invoices. For billing questions related to this change, contact Google Cloud Billing support.
Chronicle SOAR
Announcement
Release 6.2.51 is currently in Preview. 
Feature
Jobs Enhancement

When updating an integration, the jobs will now be updated automatically.
This does not apply to any legacy  jobs that were created before October 2023.

The Marketplace integration will clearly identify the legacy jobs that are affected and provide instructions on how to proceed. 

In addition, legacy jobs are now marked as such in the Jobs Scheduler page so that you can take action and resolve issues beforehand. 
Changed
APIs now documented

The following APIs are not new, but with this Release are now formally documented in Swagger:

AddOrUpdateEnvironmentRecords

RemoveEnvironmentRecords
Fixed
Searching for cases from the last week doesn't produce results (ID #00269819)
Fixed
Email HTML Templates > Show Email Template not rendering styles (ID #00249556)
Fixed
SDK call for create entity failure displays the wrong error message (ID #48950075)
Cloud Billing
Feature
You can now view granular Bigtable usage in the Cloud Billing Detailed export to BigQuery

You can now view granular Bigtable instance cost data in the Google Cloud Billing detailed export. Use the resource.global_name field in the export to view and filter your detailed Bigtable instance usage.

Review the schema of the Detailed cost data export.
Feature
You can now view granular Memorystore for Redis usage in the Cloud Billing Detailed export to BigQuery

You can now view granular Memorystore for Redis cost data in the Google Cloud Billing detailed export. Use the resource.global_name and resource.name fields in the export to view and filter your detailed Memorystore for Redis usage.

Review the schema of the Detailed cost data export.
Cloud Composer
Deprecated
All preview versions of Cloud Composer 2 are past their security notifications end date and are not supported. If your environment uses a preview version of Cloud Composer 2, then please upgrade this environment to a supported version or re-create it using the latest version of Cloud Composer 2.
Issue
If you see frequent scheduler restarts in your Airflow 2.6.3 environment and the [scheduler]job_heartbeat_sec Airflow configuration option is set to a non-default value, you can fix this issue either by upgrading to Cloud Composer version 2.6.4 or by removing this option's override, so that it uses the default value.
Cloud SQL for SQL Server
Feature
Cloud SQL now supports SQL Server Reporting Services (SSRS) on your instances. For more information, see Use SSRS for creating reports.
Compute Engine
Feature
Generally available: You can use SSH-in-browser to connect to TPU VMs. For more information, see Connecting to a Cloud TPU.
Dataform
Announcement
The 3.0.0-beta.0 version of the open-source Dataform framework is available. This update introduces significant changes, including, but not limited to, the following:


Deprecation of dataform.json in favor of workflow_settings.yaml
Stateless package installation by @dataform/cli
Warehouse-agnostic compilation output


You don't need to take immediate action to update your Dataform code.

For more information, see the 3.0.0-beta.0 release on GitHub.
Looker
Announcement
Looker 24.4 includes the following changes, features, and fixes.

Expected Looker (original) deployment start: Monday, March 18, 2024

Expected Looker (original) final deployment and download available: Thursday, March 28, 2024

Expected Looker (Google Cloud core) deployment start: Monday, March 18, 2024

Expected Looker (Google Cloud core) final deployment: Monday, April 1, 2024
Breaking
Query IDs can no longer be used to fetch queries or create render tasks through the API. The Get All Running Queries API endpoint is now restricted to admins only. Query slugs that are generated by Looker will be 32 characters instead of 7.
Breaking
Chrome is starting to deprecate third-party cookies as of January 2024. Because of Looker's dependency on third-party cookies to establish embed user sessions, this may impact your embed use case. For more information, see the Chrome is deprecating third-party cookies notice.
Breaking
Previously, custom visualizations would not call the updateAsync function after the vis config is updated via the custom visualization API. Now, the function will be called. If a custom visualization is set up to update the vis config every time updateAsync is called, it could cause excessive refreshes.

If your custom visualization is fails to load after this update, double check your custom visualization code for unnecessary vis config updates. If you have a Looker (original) instance, you can also enable the Custom Vis Reliable Render Labs feature which causes Looker to suppress excess refresh behavior in custom visualizations.
Feature
The Performant Field Picker feature is now generally available.
Feature
When an instance has no projects, Looker will more prominently prompt users to create a model.
Feature
In the Create a model wizard, your selections are now saved even if you close steps without having completed the model creation process.
Feature
Adding a query slug to source queries in the merge query API response GET merge_queries/<merge_query_id> returns the query slug in addition to the ID.
Feature
The save_content permission now has two child permissions, save_dashboards and save_looks. These permissions let Looker admins exert finer control over the kinds of content that users can save.
Fixed
Only users who have access to dashboard extensions will be shown the Add Extension tile.
Fixed
Subtotals have been fixed for queries with order_by_field references in query streaming pathways. This feature now performs as expected.
Fixed
An issue where embed secrets might have been visible to non-admin users has been fixed. This feature now performs as expected.
Fixed
Looker now ignores all blank filter strings, including IS NOT.
Fixed
An issue has been fixed that caused small decimals to be displayed in scientific notation even when formatting was disabled. This feature now performs as expected.
Fixed
An issue has been fixed where the PDT Context Override toggle was improperly reflecting the ON state when it had been cleared prior. This feature now performs as expected.
Fixed
Performant field picker sorting behavior has been fixed. This feature now performs as expected.
Fixed
Downloading results from SQL Runner now only downloads the file and does not open the file in a new browser tab.
Fixed
Filter expressions including user attributes and OR logical conditions were being incorrectly populated when generating SQL. This feature now performs as expected.
Fixed
A change in the Snowflake dialect was ported to Kotlin to maintain parity. Snowflake column names with mixed cases are now properly quoted.
Fixed
Filter suggestion requests have been reduced while the user is typing. Because normal typing will invoke fewer requests, the load on the server will be reduced.
Fixed
An issue that caused single value change indicators to not render in Safari when dashboards scrolled has been fixed. This feature now performs as expected.
Fixed
The LookML Validator no longer hangs on a connection that references a deleted or malformed user attribute. The Validator also surfaces a detailed error when the user tests the connection.
Fixed
An issue has been fixed where extension documents would appear when hiding Looker document links was disabled. This feature now performs as expected.
Fixed
Content Validator has added support for field replacement within custom measure filters (across Looks, dashboard elements, and merge queries).
Fixed
Queries with order_by_field references and subtotals should render correctly in downloads / run_query APIs.
Fixed
Looker should now correctly handle cases where the sorts query had an empty string or was entirely empty.
Fixed
Previously, the All Results option was unavailable for schedules on Looks. This feature now performs as expected.
Fixed
On the new Admin - Users page, Looker Support users were shown as having never logged in even for currently logged-in users. This issue has been fixed and this feature now performs as expected.
Fixed
LookML-defined fields that are used in field filters will not be rejected from a set when the field requiring them is rejected from that set. This feature now performs as expected.
Fixed
Previously, the Errors and  Broken Content dashboard appeared twice in the admin panel. This feature now performs as expected.
Fixed
A data validation message is now returned for waterfall charts when there are multiple measures and a hidden dimension.
Fixed
Looker now shows a clearer warning message when a user attempts to download a query with dimension fill and All Results enabled.
Fixed
Looker no longer imposes the Explore row limit of 5,000 on queries that are run using the run inline query  API endpoint.
Fixed
Previously, the lookml_model_explore API endpoint would return a 500 error in certain cases. This feature now performs as expected.
Fixed
Errors about UI downloads are now more descriptive, similar to descriptive API error messages.
Fixed
Internal encryption has been migrated from  AES-128 to AES-GCM-256 encryption.
Feature
The Disallow Numeric Query IDs legacy feature has been added to let users opt in to or out of query API changes.
Feature
The Advanced Features for New Schedules Page Labs feature is now available. This lets you sort and filter the list of scheduled plans on the Admin - Schedules page.
Fixed
Previously, when a dashboard was scheduled using PNG format and one of the tiles contained an empty note, the schedule would fail. This feature now works as expected.
Fixed
The Export function has been re-enabled, which lets Looker admins export data from a Looker (original) instance for import into a Looker (Google Cloud core) instance.
Fixed
Incorrect quoting in Snowflake views has been fixed.
Fixed
IAM checks for ephemeral users were disabled as a result of rendering issues for users who were logged in with SAML in Looker (Google Cloud core).
Resource Manager
Feature
You can add tags at the time of creating folders and projects. These tags can be added as key-value pairs. For more information, see Add tags during folder creation and Add tags during project creation. This feature is currently in preview. 
Storage Transfer Service
Feature
Support for transfers from cloud and on-premises Hadoop Distributed File System (HDFS) sources is now generally available (GA). 

HDFS support allows for use cases such as migrating from on-premises storage to Cloud Storage, archiving data to free up on-premises storage space, replicating data to Google Cloud for business continuity, or transferring data to Google Cloud for analysis and processing.

See Transfer from HDFS to Cloud Storage for details.
[/content]

PublishedDate: 2024-03-13
Category: Technology
NewsPaper: GCP latest releases
{"id"=>1331,
"title"=>"March 13, 2024",
"summary"=>nil,
"content"=>"AlloyDB for PostgreSQL\n

Feature

\n

AlloyDB now supports continuous backup and recovery, and scheduled backups on secondary clusters. When you create a secondary cluster, any backup plans on the primary cluster are automatically copied to the new secondary cluster. For more information, see About cross-region replication.

\nAnti Money Laundering AI\n

Announcement

\n

Released a new v4 engine versions for the commercial line of business, with more reliable tuning performance, in particular for small datasets.

\nApigee X\n

Announcement

\n

As of March 13, 2024, the conversion of Apigee API Management organizations with Pay-as-you-go pricing provisioned before October 1, 2023, to Pay-as-you-go organizations that use updated attributes for pricing is complete, with the exception of one organization that requires customer action.

\n\n

The Apigee API Analytics add-on is enabled in converted organizations.The Analytics add-on can be disabled if it is not required. In addition, you can update your Pay-as-you-go environment types using the API.

\n\n

For more information on the updated pricing and enhanced features now available for these organizations, see Pay-as-you-go (updated attributes) overview.

\n\n

Updated pricing attributes will be reflected in March invoices. For billing questions related to this change, contact Google Cloud Billing support.

\nChronicle SOAR\n

Announcement

\n

Release 6.2.51 is currently in Preview.

\n

Feature

\n

Jobs Enhancement

\n\n

When updating an integration, the jobs will now be updated automatically.\nThis does not apply to any legacy jobs that were created before October 2023.

\n\n

The Marketplace integration will clearly identify the legacy jobs that are affected and provide instructions on how to proceed.

\n\n

In addition, legacy jobs are now marked as such in the Jobs Scheduler page so that you can take action and resolve issues beforehand.

\n

Changed

\n

APIs now documented

\n\n

The following APIs are not new, but with this Release are now formally documented in Swagger:

\n\n

AddOrUpdateEnvironmentRecords

\n\n

RemoveEnvironmentRecords

\n

Fixed

\n

Searching for cases from the last week doesn't produce results (ID #00269819)

\n

Fixed

\n

Email HTML Templates > Show Email Template not rendering styles (ID #00249556)

\n

Fixed

\n

SDK call for create entity failure displays the wrong error message (ID #48950075)

\nCloud Billing\n

Feature

\n

You can now view granular Bigtable usage in the Cloud Billing Detailed export to BigQuery

\n\n

You can now view granular Bigtable instance cost data in the Google Cloud Billing detailed export. Use the resource.global_name field in the export to view and filter your detailed Bigtable instance usage.

\n\n

Review the schema of the Detailed cost data export.

\n

Feature

\n

You can now view granular Memorystore for Redis usage in the Cloud Billing Detailed export to BigQuery

\n\n

You can now view granular Memorystore for Redis cost data in the Google Cloud Billing detailed export. Use the resource.global_name and resource.name fields in the export to view and filter your detailed Memorystore for Redis usage.

\n\n

Review the schema of the Detailed cost data export.

\nCloud Composer\n

Deprecated

\n

All preview versions of Cloud Composer 2 are past their security notifications end date and are not supported. If your environment uses a preview version of Cloud Composer 2, then please upgrade this environment to a supported version or re-create it using the latest version of Cloud Composer 2.

\n

Issue

\n

If you see frequent scheduler restarts in your Airflow 2.6.3 environment and the [scheduler]job_heartbeat_sec Airflow configuration option is set to a non-default value, you can fix this issue either by upgrading to Cloud Composer version 2.6.4 or by removing this option's override, so that it uses the default value.

\nCloud SQL for SQL Server\n

Feature

\n

Cloud SQL now supports SQL Server Reporting Services (SSRS) on your instances. For more information, see Use SSRS for creating reports.

\nCompute Engine\n

Feature

\n

Generally available: You can use SSH-in-browser to connect to TPU VMs. For more information, see Connecting to a Cloud TPU.

\nDataform\n

Announcement

\n

The 3.0.0-beta.0 version of the open-source Dataform framework is available. This update introduces significant changes, including, but not limited to, the following:

\n\n
    \n
  • Deprecation of dataform.json in favor of workflow_settings.yaml
    • \n
  • Stateless package installation by @dataform/cli
    • \n
  • Warehouse-agnostic compilation output
    • \n
\n\n

You don't need to take immediate action to update your Dataform code.

\n\n

For more information, see the 3.0.0-beta.0 release on GitHub.

\nLooker\n

Announcement

\n

Looker 24.4 includes the following changes, features, and fixes.

\n\n

Expected Looker (original) deployment start: Monday, March 18, 2024

\n\n

Expected Looker (original) final deployment and download available: Thursday, March 28, 2024

\n\n

Expected Looker (Google Cloud core) deployment start: Monday, March 18, 2024

\n\n

Expected Looker (Google Cloud core) final deployment: Monday, April 1, 2024

\n

Breaking

\n

Query IDs can no longer be used to fetch queries or create render tasks through the API. The Get All Running Queries API endpoint is now restricted to admins only. Query slugs that are generated by Looker will be 32 characters instead of 7.

\n

Breaking

\n

Chrome is starting to deprecate third-party cookies as of January 2024. Because of Looker's dependency on third-party cookies to establish embed user sessions, this may impact your embed use case. For more information, see the Chrome is deprecating third-party cookies notice.

\n

Breaking

\n

Previously, custom visualizations would not call the updateAsync function after the vis config is updated via the custom visualization API. Now, the function will be called. If a custom visualization is set up to update the vis config every time updateAsync is called, it could cause excessive refreshes.

\n\n

If your custom visualization is fails to load after this update, double check your custom visualization code for unnecessary vis config updates. If you have a Looker (original) instance, you can also enable the Custom Vis Reliable Render Labs feature which causes Looker to suppress excess refresh behavior in custom visualizations.

\n

Feature

\n

The Performant Field Picker feature is now generally available.

\n

Feature

\n

When an instance has no projects, Looker will more prominently prompt users to create a model.

\n

Feature

\n

In the Create a model wizard, your selections are now saved even if you close steps without having completed the model creation process.

\n

Feature

\n

Adding a query slug to source queries in the merge query API response GET merge_queries/<merge_query_id> returns the query slug in addition to the ID.

\n

Feature

\n

The save_content permission now has two child permissions, save_dashboards and save_looks. These permissions let Looker admins exert finer control over the kinds of content that users can save.

\n

Fixed

\n

Only users who have access to dashboard extensions will be shown the Add Extension tile.

\n

Fixed

\n

Subtotals have been fixed for queries with order_by_field references in query streaming pathways. This feature now performs as expected.

\n

Fixed

\n

An issue where embed secrets might have been visible to non-admin users has been fixed. This feature now performs as expected.

\n

Fixed

\n

Looker now ignores all blank filter strings, including IS NOT.

\n

Fixed

\n

An issue has been fixed that caused small decimals to be displayed in scientific notation even when formatting was disabled. This feature now performs as expected.

\n

Fixed

\n

An issue has been fixed where the PDT Context Override toggle was improperly reflecting the ON state when it had been cleared prior. This feature now performs as expected.

\n

Fixed

\n

Performant field picker sorting behavior has been fixed. This feature now performs as expected.

\n

Fixed

\n

Downloading results from SQL Runner now only downloads the file and does not open the file in a new browser tab.

\n

Fixed

\n

Filter expressions including user attributes and OR logical conditions were being incorrectly populated when generating SQL. This feature now performs as expected.

\n

Fixed

\n

A change in the Snowflake dialect was ported to Kotlin to maintain parity. Snowflake column names with mixed cases are now properly quoted.

\n

Fixed

\n

Filter suggestion requests have been reduced while the user is typing. Because normal typing will invoke fewer requests, the load on the server will be reduced.

\n

Fixed

\n

An issue that caused single value change indicators to not render in Safari when dashboards scrolled has been fixed. This feature now performs as expected.

\n

Fixed

\n

The LookML Validator no longer hangs on a connection that references a deleted or malformed user attribute. The Validator also surfaces a detailed error when the user tests the connection.

\n

Fixed

\n

An issue has been fixed where extension documents would appear when hiding Looker document links was disabled. This feature now performs as expected.

\n

Fixed

\n

Content Validator has added support for field replacement within custom measure filters (across Looks, dashboard elements, and merge queries).

\n

Fixed

\n

Queries with order_by_field references and subtotals should render correctly in downloads / run_query APIs.

\n

Fixed

\n

Looker should now correctly handle cases where the sorts query had an empty string or was entirely empty.

\n

Fixed

\n

Previously, the All Results option was unavailable for schedules on Looks. This feature now performs as expected.

\n

Fixed

\n

On the new Admin - Users page, Looker Support users were shown as having never logged in even for currently logged-in users. This issue has been fixed and this feature now performs as expected.

\n

Fixed

\n

LookML-defined fields that are used in field filters will not be rejected from a set when the field requiring them is rejected from that set. This feature now performs as expected.

\n

Fixed

\n

Previously, the Errors and Broken Content dashboard appeared twice in the admin panel. This feature now performs as expected.

\n

Fixed

\n

A data validation message is now returned for waterfall charts when there are multiple measures and a hidden dimension.

\n

Fixed

\n

Looker now shows a clearer warning message when a user attempts to download a query with dimension fill and All Results enabled.

\n

Fixed

\n

Looker no longer imposes the Explore row limit of 5,000 on queries that are run using the run inline query API endpoint.

\n

Fixed

\n

Previously, the lookml_model_explore API endpoint would return a 500 error in certain cases. This feature now performs as expected.

\n

Fixed

\n

Errors about UI downloads are now more descriptive, similar to descriptive API error messages.

\n

Fixed

\n

Internal encryption has been migrated from AES-128 to AES-GCM-256 encryption.

\n

Feature

\n

The Disallow Numeric Query IDs legacy feature has been added to let users opt in to or out of query API changes.

\n

Feature

\n

The Advanced Features for New Schedules Page Labs feature is now available. This lets you sort and filter the list of scheduled plans on the Admin - Schedules page.

\n

Fixed

\n

Previously, when a dashboard was scheduled using PNG format and one of the tiles contained an empty note, the schedule would fail. This feature now works as expected.

\n

Fixed

\n

The Export function has been re-enabled, which lets Looker admins export data from a Looker (original) instance for import into a Looker (Google Cloud core) instance.

\n

Fixed

\n

Incorrect quoting in Snowflake views has been fixed.

\n

Fixed

\n

IAM checks for ephemeral users were disabled as a result of rendering issues for users who were logged in with SAML in Looker (Google Cloud core).

\nResource Manager\n

Feature

\n

You can add tags at the time of creating folders and projects. These tags can be added as key-value pairs. For more information, see Add tags during folder creation and Add tags during project creation. This feature is currently in preview.

\nStorage Transfer Service\n

Feature

\n

Support for transfers from cloud and on-premises Hadoop Distributed File System (HDFS) sources is now generally available (GA).

\n\n

HDFS support allows for use cases such as migrating from on-premises storage to Cloud Storage, archiving data to free up on-premises storage space, replicating data to Google Cloud for business continuity, or transferring data to Google Cloud for analysis and processing.

\n\n

See Transfer from HDFS to Cloud Storage for details.

\n\n ",
"author"=>nil,
"link"=>"https://cloud.google.com/release-notes#March_13_2024",
"published_date"=>Wed, 13 Mar 2024 07:00:00.000000000 UTC +00:00,
"image_url"=>nil,
"feed_url"=>"https://cloud.google.com/release-notes#March_13_2024",
"language"=>nil,
"active"=>true,
"ricc_source"=>"feedjira::v1",
"created_at"=>Sun, 31 Mar 2024 21:23:49.015411000 UTC +00:00,
"updated_at"=>Mon, 21 Oct 2024 16:58:27.495202000 UTC +00:00,
"newspaper"=>"GCP latest releases",
"macro_region"=>"Technology"}
Edit this article
Back to articles