and two additionnal idea:- check logs to find the initial SQL statement(s) that failed,- restart the cluster to free all that space
RegardsPAscal
--Sent from: https://www.postgresql-archive.org/PostgreSQL-novice-f2132464.html