Re: bitmap AM design - Mailing list pgsql-hackers

From Tom Lane
Subject Re: bitmap AM design
Date
Msg-id 19530.1109703941@sss.pgh.pa.us
Whole thread Raw
In response to Re: bitmap AM design  ("Victor Y. Yegorov" <viy@mits.lv>)
Responses Re: bitmap AM design
List pgsql-hackers
"Victor Y. Yegorov" <viy@mits.lv> writes:
> All lists (list of ctids, bitmaps) will only grow, no data will be
> deleted, as deletes will require relocation and possibly exclusive
> lock on the index.
> Extending lists will need only a short-term exclusive locks on the pages in
> the tails of each list.

Hmm, you seem to be envisioning that these are actually lists, not
arrays --- that is, to find the N'th page in a list requires traversing
list links starting at the first page.  That doesn't sound workable.
If you can't access all the entries in roughly constant time then the
thing is going to have problems with big tables.

> I have counter question for you: you've mentioned, that you want bitmaps in
> the 8.1. What kind of bitmaps you were speaking about?

In-memory is what I intend to work on.
        regards, tom lane


pgsql-hackers by date:

Previous
From: Tom Lane
Date:
Subject: Re: bitmap AM design
Next
From: pgsql@mohawksoft.com
Date:
Subject: Re: bitmap AM design