Humio Server 1.24.4 LTS (2021-05-31)
| Version? | Type? | Release Date? | Availability? | End of Support | Security Updates | Upgrades From? | Config. Changes? |
|---|---|---|---|---|---|---|---|
| 1.24.4 | LTS | 2021-05-31 | Cloud | 2022-04-30 | No | 1.16.0 | No |
Hide file hashes
| JAR Checksum | Value |
|---|---|
| MD5 | 83588b8570c97dc8a89b46524ea8d92a |
| SHA1 | 15958a097eaba9689a43c557e1d6d14e2d566d6d |
| SHA256 | cafca561035aec00f5971a15cf283b757794c8a2706829518a05ce089f92ea88 |
| SHA512 | 15352a96bb6b4ed27ef1c7b238cf97142c36cfefe17be25bf671a706bd6afcf0c8ad8951a325838e259eeb52e643f3950b8f47f3186ecea049f34fc81e066eed |
These notes include entries from the following previous releases: 1.24.0, 1.24.1, 1.24.2, 1.24.3
Minor bug fixes, as well as fix a stackoverflow bug in large clusters.
Fixed in this release
Summary
Minor bug fixes and improvements.
Minor bug fixes and improvements.
Fix a stackoverflow that could occur during startup in larger clusters.
Other
Removed the
QUERY_QUOTA_EXCEEDED_PENALTYconfig (introduced in 1.19.0).Fixed an issue on the search page that prevented the event list from scrolling correctly.
Fixed a bug where ingestOnly nodes could not start on a more recent version that the statefull nodes in the cluster
Fixed an issue which prevented Safari users from seeing alert actions
Fixed an issue where cost spent in a long-running query got accounted as spent "now" when the query ended in terms of Query Quota
Fixed an issue which caused problems with forward/backward compatibility of LanguageVersion in GraphQL
Major changes: 1.23.0 and 1.23.1.
Fixed an issue where a repository with very high number of datasources could trigger an error writing an oversized message to kafka from the ingest-reader-leader thread
The query scheduler now prioritizes new queries started by a user based on the cumulative cost of recent queries started by that user. Added new config
QUERY_SPENT_FACTORwith the default value 0.5, which defines the weight of recent query costs when scheduling. Higher values mean that users with high recent query costs will see their queries penalized harder in the scheduling.Ensure that if the Kafka leader loop thread dies, it kills the Humio process. In rare cases it was possible for this thread to die, leaving the node incapable of performing digest work
Fixes an issue where the user would get stuck in infinite loading after having been invited into an organization
Fixed a scrolling issues on the Kafka cluster admin page
Allow reverse proxies using 10s as timeout to work also for a query that takes longer than that to initialize
Reduced off-heap memory usage