Re: Tricky bugs in concurrent index build - Mailing list pgsql-hackers

From Zeugswetter Andreas DCP SD
Subject Re: Tricky bugs in concurrent index build
Date
Msg-id E1539E0ED7043848906A8FF995BDA579014DB7DB@m0143.s-mxs.net
Whole thread Raw
In response to Re: Tricky bugs in concurrent index build  (Tom Lane <tgl@sss.pgh.pa.us>)
Responses Re: Tricky bugs in concurrent index build
Re: Tricky bugs in concurrent index build
List pgsql-hackers
> > What bothers me about what we have now is that we have optional
> > keywords before and after INDEX, rather than only between
> CREATE and INDEX.
>
> Yeah, putting them both into that space seems consistent to
> me, and it will fix the problem of making an omitted index
> name look like a valid command.
>
> I'm not sure I should be opening this can of worms, but do we
> want to use a different keyword than CONCURRENTLY to make it
> read better there?

precedent syntax (Oracle, Informix) uses the keyword ONLINE at the end:CREATE INDEX blabla_x0 ON blabla (a,b) ONLINE;

I'd stick with that.

Andreas


pgsql-hackers by date:

Previous
From: "Bort, Paul"
Date:
Subject: Re: [Pgsqlrpms-hackers] Safer auto-initdb for RPM init
Next
From: Zoltan Boszormenyi
Date:
Subject: Re: Performance testing of COPY (SELECT) TO