Hi, today my production druid cluster alerted me that a disk on a historical was full. Almost 500 Gigs of shards (smoosh files).. in the java temp space:
java.io.tmpdir=/mnt/tmp
Has anyone encountered this and utilized a viable way to remove them, or combat the issue? Any help is much appreciated!
Regards,
John