Re: CLUSTER and clustered indices - Mailing list pgsql-hackers

From Tom Lane
Subject Re: CLUSTER and clustered indices
Date
Msg-id 26552.1132243110@sss.pgh.pa.us
Whole thread Raw
In response to CLUSTER and clustered indices  (Simon Riggs <simon@2ndquadrant.com>)
Responses Re: CLUSTER and clustered indices
List pgsql-hackers
Simon Riggs <simon@2ndquadrant.com> writes:
> When a table has been CLUSTERed on a particular index AND that index
> values is monotonically increasing, then it would be a bad move to use
> blocks from the FSM since this would tend to destroy the natural
> clustering sequence.

By the time there are any blocks in FSM to take, the original clean
index page sequence is doubtless history.  The pure-increasing-key
scenario you are thinking of never will have any FSM entries, so it's
moot.
        regards, tom lane


pgsql-hackers by date:

Previous
From: "Michael Paesold"
Date:
Subject: Re: Optional postgres database not so optional in 8.1
Next
From: Alvaro Herrera
Date:
Subject: Re: CLUSTER and clustered indices