From: | Robert Haas <robertmhaas(at)gmail(dot)com> |
---|---|
To: | Alvaro Herrera <alvherre(at)alvh(dot)no-ip(dot)org> |
Cc: | Masahiko Sawada <sawada(dot)mshk(at)gmail(dot)com>, Peter Geoghegan <pg(at)bowt(dot)ie>, PostgreSQL Hackers <pgsql-hackers(at)lists(dot)postgresql(dot)org> |
Subject: | Re: Rename dead_tuples to dead_items in vacuumlazy.c |
Date: | 2021-11-24 15:50:13 |
Message-ID: | CA+TgmoaEKSR40F7ZVxCfjpgyueKHN_ELKjQBV3kAatj2iuD==g@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
On Wed, Nov 24, 2021 at 9:51 AM Alvaro Herrera <alvherre(at)alvh(dot)no-ip(dot)org> wrote:
> On 2021-Nov-24, Robert Haas wrote:
> > Hmm. I think in my model an item and an item pointer and a line
> > pointer are all the same thing, but a TID is different. When I talk
> > about a TID, I mean the location of an item pointer, not its contents.
> > So a TID is what tells me that I want block 5 and the 4th slot in the
> > item pointer array. The item pointer tells me that the associate tuple
> > is at a certain position in the page and has a certain length.
>
> OK, but you can have item pointers that don't have any item.
> LP_REDIRECT, LP_DEAD, LP_UNUSED item pointers don't have items.
I guess so. I said before that I thought an item and an item pointer
were the same, but on reflection, that doesn't entirely make sense.
But I don't know that I like making item and tuple synonymous either.
I think perhaps the term "item" by itself is not very clear.
--
Robert Haas
EDB: http://www.enterprisedb.com
From | Date | Subject | |
---|---|---|---|
Next Message | Bharath Rupireddy | 2021-11-24 15:59:51 | Re: pg_replslotdata - a tool for displaying replication slot information |
Previous Message | Bharath Rupireddy | 2021-11-24 15:39:48 | Re: pg_replslotdata - a tool for displaying replication slot information |