I went through the README on HOT. That was really a nice and cool feature. Hats off to the person who thought about it.
Ihave a couple of doubts about it.<br /><br />a) In the README, there is a statement like this. <br clear="all"
/><pre>Inprinciple we could continue a HOT chain across<br />pages, but this would destroy the desired property of
beingable to<br />reclaim space with just page-local manipulations. Anyway, we don't<br />want to have to chase
throughmultiple heap pages to get from an index
<br />entry to the desired tuple, so it seems better to create a new index<br />entry for the new tuple.<br /></pre><br
/>Iam especially interested in the case of continuing the HOT chain across pages. When we are actually reclaiming
space,we should check the snapshot and reclaim it. If it is HOT updated, we will leave the top most tuple and take the
rest.So then the top most tuple will remain always and any index scan now has to make read two heap pages to reach the
targetentry.Is this the only reason, it was left out?<br /><br /><br />-- <br />Thanks,<br />Gokul.<br />CertoSQL
Project,<br/>Allied Solution Group.<br />(<a href="http://www.alliedgroups.com">www.alliedgroups.com</a>)