v.4.14.0 Release

PRODUCTDATERELEASEADDITIONAL RESOURCES
Search2025-09-17FeatureKnown Issues, Cribl Lake Release Notes

Cribl Search 4.14.0 includes significant performance improvements, new capabilities, and important bug fixes.

Important Changes

These changes to Cribl Search might require you to check or modify existing queries, or other configuration, especially on saved searches.

Lakehouse Datasets Default to Lakehouse Query Execution

When you use Cribl Search to query a Cribl Lake Dataset that’s assigned to a Lakehouse, queries now default to Lakehouse execution. This returns results faster and more efficiently, but limits the results to the past 30 days, even if the underlying Lake Dataset has a longer retention period. To override the new default and search back beyond 30 days, precede your query with set lakehouse="off"; (see the docs for details).

Stricter Data Sending Permissions

Search Editors can no longer send search results to custom URLs. However, they can still send results to custom-named Stream Worker Groups. For details, see the Permissions section of the send operator doc (available starting Sep 17).

New Features

This release adds the following new capabilities.

Notebooks (Preview) for Data Investigation Storytelling

Notebooks enable data analysts to collaborate on complex, multi-stage investigations right within Cribl Search. In 14.4.0, we’re releasing Notebooks as a Preview feature. Here’s what you can do with it:

  • Iterate: Run multiple searches next to one another, for faster, more efficient investigations.
  • Annotate: Add complex notes using Markdown.
  • Share: Grant editing or read-only access to your Notebook.

See the Notebooks documentation for details.

Dashboard Interactions Support More Tokens

When you add interactions to Dashboard visualizations, you can now use new token types to help users drill down into underlying values:

  • $environment.baseUrl$ (static)
  • $environment.type$ (static)
  • $environment.workspace$ (static)
  • $field.name$ (dynamic, the name of the field clicked)
  • $field.value$ (dynamic, the value of the field clicked)
  • $rowData.<fieldname>$ (dynamic, extracts individual fields from returned events)

For details, see Add Interactions to Your Cribl Search Dashboard.

Cribl Search User API Credentials

Organization Owners and Organization Admins can now create API Credentials that assign Cribl Search User Permissions for individual Workspaces.

Cribl Terraform Provider (Preview)

The Cribl Terraform provider is now available in Preview from the Terraform Registry. Use the criblio provider to manage Cribl as part of your infrastructure-as-code workflows and make large deployments repeatable and reliable.

New Cribl.Cloud Regions

We are excited to now support creating Cribl.Cloud Organizations in Switzerland and Singapore AWS Regions (eu-central-2, ap-southeast-1).

Corrections

This release includes numerous fixes to various areas of Cribl Search, most notably:

IssueDescription
CRIBL-30786
percentile, median, and medianif functions now return consistent results, whether you’re searching a Lakehouse or not.
SEARCH-3389Searches executed via the Cribl Search API now use the same default time range values as UI searches. earliest defaults to -1h, and latest to now.
SEARCH-5899When exporting results to Cribl Lake (export to lake), the search logs now correctly show the number of bytes written. Before, the bytesOut and totalBytesOut fields would incorrectly show 0 in some scenarios.
SEARCH-9084Exporting results to Cribl Lake now correctly returns final search logs upon completion.
SEARCH-9844Searching large (>64 MB) Splunk journal files is now more reliable and scalable.
SEARCH-10396Searches that execute immediately will no longer flicker as Queued in the UI.
SEARCH-10461The number of executors no longer exceeds the max_executors value.
SEARCH-10553On the Cribl Search landing page, the Available Datasets and History sections now display a clearer wording for the last run time (XX minutes ago rather than in XX minutes).
SEARCH-10569We fixed an issue where having a large number of Datasets made the query box work slower.