8.13.0 release notes
edit8.13.0 release notes
editUpgrading to Enterprise Search 8.13.0? See Upgrading and migrating.
New features
edit-
We’ve added the following connectors to our catalog:
- Notion (beta)
- Jira Data Center (technical preview)
- Confluence Data Center (technical preview)
- Redis (technical preview)
-
The following connectors are now natively available in Elastic Cloud:
- Gmail
- Outlook
-
The following connectors are now generally available:
- Salesforce
- Google Drive
- Onedrive
-
The following connectors now support document level security (DLS):
- Servicenow
- Salesforce
- Jira Server & Data Center
-
We’ve added the ability to use API keys for native connectors running in Elastic Cloud, behind a feature flag. This means native connectors now require API keys.
- Native connectors created before 8.13.0 do not automatically have API keys upon upgrading. The attached indices of these connectors cannot be changed until the connector has been converted to use an API key. Refer to Native Connectors Enabling Api Keys For Upgraded Connectors for more instructions.
-
We have enabled streamlined incremental sync jobs for the bulk of our non-database connectors following connectors, with fine-grained, minute-level scheduling.
-
Refer to the documentation for the full list of connectors that support incremental syncs.
To enable incremental syncs, you’ll need to run the following API call in Kibana Dev Tools (or cURL):
POST .elastic-connectors/_update/<connector_id> { "doc": { "features": { "incremental_sync": { "enabled": true } } } }
-
-
All connectors can now write to any Elasticsearch index, without needing to
search-
prefix the index. This also makes it easy to attach connectors to existing indices.-
One exception is that we still require
search-
as a prefix for indices created by the Elastic crawler for now. -
Following this change, we replaced the "Only show
search-
optimized indices" toggle in the Indices tab with a "Only show crawler indices".
-
One exception is that we still require
- The connector service tries to apply settings and mappings if the connector is created programmatically using the Elasticsearch Connector APIs. When working with an existing index, previously only mappings were applied, but we’ve updated to apply settings as well.
-
The Connectors CLI can now authorize with Elasticsearch API keys by running
./bin/connectors login --method apikey.
-
The CLI command
./bin/connectors index list
now lists all open indices, and not just those with thesearch-*
prefix. -
We added the
name
argument to the Connectors CLI to allow you to specify a name for the connector when creating it. -
service.max_file_download_size
can now be set in your connectorsconfig.yml
file to limit the size of files the connectors will attempt to download. This config is only used for connectors that don’t use the self-hosted Data Extraction Service.
Bug fixes
edit- We have improved the output structure of documents created by access control syncs for document level security. These docs are now appropriately structured to create role descriptors. The structure for accessing identity and permissions has not changed. For details, refer to the this pull request. This can help reduce storage requirements by up to 50%.
-
Fixed an issue where removing trained models could result in warning messages like the following:
#! The default [remove_binary] value of 'false' is deprecated and will be set to 'true' in a future release. Set [remove_binary] explicitly to 'true' or 'false'...
Known issues
edit-
Enterprise Search: Enterprise Search nodes in Elastic Cloud may be erroneously removed from deployments. See connectors known issues.
-
Fixed in
8.14.3
-
Fixed in