As seen in
TEIID-2041 debug logs can grow much too large. The two main culprits are full pre-planning of all subqueries and the amount of information shown on each node for each rule. To address the former we should instead only plan project node subqueries after the initial run of RuleAssignOutputElements, since typically not all columns will be used. For the latter we should consider putting a modified marker on a plannode that will get cleared after a toString to truncate the values shown for a node (in some cases we make modifications to objects held on the node, so it's not quite that simple).