Creating indices concurrently (provided that index creation *completes*)
is generally a much safer way to build an index (since it allows reads
*and* writes to the table during index creation).
However, due to the inherent tradeoffs of concurrent index creation, the
operation will almost always take *longer* than a traditional (blocking)
index creation. It's actually much *more* important to lint against
concurrent index creation when an exclusive lock is open & held.
Test is from a comment by @David-Wobrock (I corrected syntax, though):
https://github.com/3YOURMIND/django-migration-linter/pull/251#discussion_r1249298197
While it's always a dangerous *pattern* to create an index while a
transaction is open & holds an exclusive lock, in practice it's not
always a big deal (for example, indexing a table with very few rows can
be fast). Accordingly, downgrade to a warning.
David Cain