When a table has too many partitions, do_tm_task('moveout') and mergeout cannot complete efficiently. This often occurs when tables are partitioned by day over multiple years. Reducing partition count by merging, archiving, or dropping old partitions allows Tuple Mover tasks to run successfully.
Environment
Product: Vertica Analytics Platform
Feature: Partitioned tables and Tuple Mover
Situation
Tuple Mover tasks (
mergeout, moveout) fail or stall due to excessive partition count.Table is likely partitioned by day, creating thousands of partitions over time.
Cause
Too many partitions increase ROS container fragmentation and prevent Tuple Mover from consolidating data efficiently.
