From: | Guillaume Lelarge <guillaume(at)lelarge(dot)info> |
---|---|
To: | Heikki Linnakangas <hlinnakangas(at)vmware(dot)com> |
Cc: | PostgreSQL Developers <pgsql-hackers(at)postgresql(dot)org> |
Subject: | Re: Misunderstanding on the FSM README file |
Date: | 2014-12-09 07:28:41 |
Message-ID: | CAECtzeVdVL_h_Qr0jqtHQHk3OvqKCZapCXez6Uf5W5P7z0+oGg@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
2014-12-07 15:07 GMT+01:00 Heikki Linnakangas <hlinnakangas(at)vmware(dot)com>:
> On 12/07/2014 02:03 PM, Guillaume Lelarge wrote:
>
>> Hi,
>>
>> I've been reading the FSM README file lately
>> (src/backend/storage/freespace/README), and I'm puzzled by one of the
>> graph
>> (the binary tree structure of an FSM file). Here it is:
>>
>> 4
>> 4 2
>> 3 4 0 2 <- This level represents heap pages
>>
>> Shouldn't the last line be:
>> 4 3 2 0
>>
>> (ie, highest number of free space on the left node, lowest on the right
>> one)
>>
>> Probably just nitpicking, but still, I'm wondering if I missed something
>> out.
>>
>
> No, that's not how it works. Each number at the bottom level corresponds
> to a particular heap page. The first number would be heap page #0 (which
> has 3 units of free space), the second heap page #1 (with 4 units of free
> space) and so forth. Each node on the upper levels stores the maximum of
> its two children.
>
>
Oh OK. Thanks Heikki, that makes perfect sense.
--
Guillaume.
http://blog.guillaume.lelarge.info
http://www.dalibo.com
From | Date | Subject | |
---|---|---|---|
Next Message | Michael Paquier | 2014-12-09 07:44:30 | Re: Proposal : REINDEX SCHEMA |
Previous Message | Amit Langote | 2014-12-09 07:19:11 | Re: On partitioning |