Re: Concurrently option for reindexdb

On August 25, 2014 10:35:20 PM CEST, Alvaro Herrera <alvherre(at)2ndquadrant(dot)com> wrote:
>Michael Paquier wrote:
>> On Tue, Aug 26, 2014 at 3:48 AM, Fujii Masao <masao(dot)fujii(at)gmail(dot)com>
>wrote:
>> > On Mon, Aug 25, 2014 at 4:33 PM, Sawada Masahiko
><sawada(dot)mshk(at)gmail(dot)com> wrote:
>> >> this might be difficult to call this as --concurrently.
>> >> It might need to be change the name.
>> >
>> > I'm OK to say that as --concurrently if the document clearly
>> > explains that restriction. Or --almost-concurrently? ;P
>> By reading that I am thinking as well about a wording with "lock",
>> like --minimum-locks.
>
>Why not just finish up the REINDEX CONCURRENTLY patch.

+many. Although I'm not sure if we managed to find a safe relation swap.

If not: How about adding ALTER INDEX ... SWAP which requires an exclusive lock but is fast and O(1)? Than all indexes can be created concurrently, swapped in a very short xact, and then dropped concurrently? 95% of all users would be happy with that and the remaining 5 would still be in a better position than today where the catalog needs to be hacked for that (fkeys, pkeys et al).

---
Please excuse brevity and formatting - I am writing this on my mobile phone.