Removed input in Elastic Agent Fleet System integration
With the release of Elastic Agent Fleet version 2.0.0, the “System” integration has been updated and now removes support for a deprecated log collection input method that relied on httpjson. This input is no longer available in the latest version of the integration.
To address this change, we’ve updated the NetEye Fleet policy definition by removing the httpjson input reference from the System integration template. This fix ensures continued compatibility with the new integration behavior.
However, please be aware that if you’re running a NetEye version 4.40 or previous, the default System integration definition must be updated manually. Failing to do so before an upgrade, if the System integration has been updated to version 2.0, can break the NetEye update or upgrade process.
To prevent this issue, NetEye users with versions prior to 4.41 must manually edit the following configuration files on every NetEye node:
Fix Kibana sysconfig migration We resolved a bug that occurred during the migration of Kibana to a multi-instance setup. The issue prevented the proper copying of custom configurations from /neteye/shared/kibana/conf/sysconfig/kibana-user-customization. As a result, these customizations were missing from the Kibana instances Read More
Fix monitoring object creation during deployment We fixed a bug that caused a monitoring object to be lost if it was created while a Diretor deployment was in progress. List of updated packages To solve the issues mentioned above, the Read More
Fix for Tornado Webhook collector We have fixed an issue where large incoming requests to the Webhook Collector could cause the system to become unresponsive over time. List of updated packages To solve the aforementioned issues, the following packages have Read More
Fixes for Elastic Stack Improved wait for cluster status during updates / upgrades When restarting the Elasticsearch cluster during the NetEye update / upgrade procedure, is it possible let the restart procedure going even if the cluster has a "yellow" Read More
Fix for Previously, database updates would fail if the GLPI DRBD resource was active on a node other than the primary, due to an inability to locate the necessary database configuration within the DRBD-managed volume. This fix ensures that GLPI Read More