We've run into multiple situations recently where clients were effectively unable to run CREATE INDEX or REINDEX
(CONCURRENTLYor otherwise) due to the load that they imposed on the system, mostly in I/O. I'd like to propose a
cost-baseddelay system for maintenance operations such as these, paralleling the system for (auto)VACUUM. Before I
pulltogether a code proposal, I wanted to preflight the idea to make sure it has not been considered and rejected
already,or find out if anyone is working on it already.
