Re: Inserts and row chaining

From: ZEITGEIST <restaurant_at_the.end.of.the.universe>
Date: 1997/01/22
Message-ID: <5c4083$j5a@news1.ee.net>#1/1

In article <32e6d779.3583911_at_n5.gbso.net>, chuckh_at_dvol.com says...
>
>
>
>I think the question is something like this...
>
>If a block is already on the freelist (i.e. it fell below pctused but
>hasn't reached pctfree yet), and there's only 50 bytes remaining until
>it hits pctfree, and you have a 100 byte row, can it be inserted into
>that block?
>
>Please corrent me if I'm wrong but I think the answer is no. Only an
>update can cause a chaned row.
>--
>Chuck Hamilton
>chuckh_at_dvol.com
>

Thank you, Chuck. This was my question exactly. However, I think you can get them on an INSERT if you have a LONG or LONG RAW field. I always thought that, but the O'Reilly "Oracle Performance Tuning" book was a bit vague on stating that this would be the only case. That is what caused me to doubt my initial thoughts.

Thanks again.

Brett Cunningham
zgeist_at_ee.net Received on Wed Jan 22 1997 - 00:00:00 CST