Falcon LogScale 1.69.0 GA (2022-12-13)

Version?Type?Release Date?Availability?End of SupportSecurity UpdatesUpgrades From?Downgrades To?Config. Changes?
1.69.0GA2022-12-13

Cloud

2024-01-31No1.44.01.26.0No

Available for download two days after release.

Bug fixes and updates.

Deprecation

Items that have been deprecated and may be removed in a future release.

  • The query function holtwinters() is now deprecated and will be removed along with the release of future version 1.73; therefore, its usage in alerts is not recommended.

  • The holtwinters() query function will be deprecated with the release of future version 1.68. From then, it cannot be expected to work in alerts, and it will be removed entirely with the release of version 1.72.

New features and improvements

  • Storage

    • Reduced CPU usage of background tasks for the case of high partition count and high datasource count.

  • Queries

    • Add support for GET+DELETE requests for queries by external Query ID without including the repository name in the URL. The new URL is /api/v1/queryjobs/QUERYID. Note that shared dashboard token authentication is not supported on this API. (The existing API on /api/v1/repositories/REPONAME/queryjobs/QUERYID remains unmodified and support POST requests for submit of queries.)

  • Other

    • Add bounds to maximum number of active notifications per user.

    • Added option to filter by group and role permission types in groupsPage and rolesPage queries.

    • Throttle publish to global-events topic internally based on time spent in recent transactions of the same type (digest-related writes are not throttled). See details at new configuration variable GLOBAL_THROTTLE_PERCENTAGE page. Also see the metric global-operation-time for measurements of the time spent.

Fixed in this release

  • Other

    • Allow creating Kafka topics even if a broker is down.

    • LogScale no longer considers every host to be alive for a period after rebooting. Only hosts marked as running in global will be considered alive. This fixes an issue where a query coordinator might pointlessly direct queries to dead nodes because the coordinator had recently booted.