Humio Server 1.6.5 Archive (2019-10-01)
Version? | Type? | Release Date? | Availability? | End of Support | Security Updates | Upgrades From? | Config. Changes? |
---|---|---|---|---|---|---|---|
1.6.5 | Archive | 2019-10-01 | Cloud | 2020-11-30 | No | 1.5.19 | No |
JAR Checksum | Value |
---|---|
MD5 | d5426b20769fcea54bdb6a296a1b853d |
SHA1 | 7881a0845b5548afd23938731abef349259a5002 |
SHA256 | d4d8abbdc1e4730141a36718101444377ed106c8a3db8557e572ca3eaec35bfa |
SHA512 | b71876f6d325b99cadbabb65acd304ca713ffb62e319a29941e16c5645540bf824a07c0fcc3d8820c735ae1e027bff1fad05ec150b36bf5bb973bc71d34ac274 |
Bug Fixes and Performance Improvements
No data migration required, but see version 1.6.3.
Fixed in this release
Summary
Redefined the
event-latency
metric to start measuring after parsing the events, just before inserting them into the ingest queue in Kafka. This metric is the basis of autosharding decisions and other scheduling priority choices internally and thus needs to reflect the time spent on the parts influenced by those decisions.Support reading events from the ingest queue in both the format written by 1.6.3 and older and 1.6.4.
The new metric
event-latency-repo/<repo>
includes time spent parsing too and is heavily influenced by the size of the bulks of events being posted to Humio.Apply the extra Kafka properties from config also on deleteRecordsBefore requests.
Improved performance of internal jobs calculating the data for the cluster management pages.
The new metric
ingest-queue-latency
measures the latency of events through the ingest queue in Kafka from the "send" to kafka and until it has been received by the Digest node.