ITAGAKI Takahiro <itagaki(dot)takahiro(at)oss(dot)ntt(dot)co(dot)jp> writes:
> BTW, why REINDEX requires access exclusive lock? Read-only queries
> are forbidden during the operation now, but I feel they are ok
> because REINDEX only reads existing tuples. Can we do REINDEX
> holding only shared lock on the index?
No. When you commit the reindex, the old copy of the index will
instantaneously disappear; it will not do for someone to be actively
scanning that copy.
regards, tom lane