This message was deleted.
# generals
Slackbot
10/06/2023, 8:46 PMThis message was deleted.
b
Ben Krug
10/06/2023, 10:53 PMYou should be able to - https://druid.apache.org/docs/latest/data-management/compaction/#segment-granularity-handling
a
Ankit
10/08/2023, 3:14 PMThanks @Ben Krug
j
Julien Calfayan
10/25/2023, 2:53 PMIf you have any other segment granularities in your cluster that are not Week, then it’s highly recommended to not use Week granularity.
We ran into that problem at Atlassian where we had datasources that had segments of Month for data older than 1 year, younger than a year was Week, and anything within the last week was Day.
The issue is that Week and Month segments don’t work well together and will cause overlapping segments due to the fact that Week segments use isoWeek intervals that don’t play well with Month intervals which go from first day of the month to first day of the month
b
Ben Krug
10/25/2023, 3:42 PMvery good point