[ https://issues.jboss.org/browse/TEIID-4251?page=com.atlassian.jira.plugin... ] Steven Hawkins commented on TEIID-4251: --------------------------------------- I'm trying to avoid using the allow-materialization-management construct in order to remove the need for an explicit status table or mvstatus function usage. To do that though requires some code modifications if any of the materialization scripts will still be run even if allow management is turned off. The other approach is to leave teiid out of it as much as possible. With the pg container include a create script: CREATE EXTENSION postgres_fdw; CREATE SERVER teiid_server ...; CREATE USER MAPPING FOR pg_user ...; -- for each schema with a materialized view CREATE SCHEMA remote_schema ...; IMPORT FOREIGN SCHEMA remote_schema (limit to all materializations in the schema) ...; -- for each materialized view in the schema CREATE MATERIALIZED VIEW xxx_mat AS SELECT * FROM remote_schema.xxx OPTION NO CACHE WITH NO DATA; create key / index information on xxx_mat ... By specifying option no cache we avoid having to worry about any load order and can concurrently load all. There would then need to be an initial load set of scripts, REFRESH MATERIALIZED VIEW xxx_mat; And a set of refresh scripts (with the concurrent option only against materialized views that have a unique key): REFRESH MATERIALIZED VIEW CONCURRENTLY xxx_mat; These could be run off of cron jobs, which could also be associated with the pg container image. -- This message was sent by Atlassian Jira (v7.12.1#712002)