HP3000-L Archives

September 1995, Week 1

HP3000-L@RAVEN.UTC.EDU

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Leslie-Anne Bain <[log in to unmask]>
Reply To:
Leslie-Anne Bain <[log in to unmask]>
Date:
Wed, 6 Sep 1995 23:17:57 GMT
Content-Type:
text/plain
Parts/Attachments:
text/plain (31 lines)
Chris Breemer ([log in to unmask]) wrote:
 
> Does anybody know if an Allbase TID can change as the result of an UPDATE ?
> This seems to be the case with Oracle rowid's, but in the Allbase manuals
> I cannot find anything about it.
 
The TID of a row never changes between the time the row is inserted until it
is deleted.
 
If the length of the row changes so that it is not possible to physically
store the row on the original page, an "indirect" row is created.  The
original row is moved to another page that has space.  A pointer to the new
location (the pointer is itself a TID) is stored at the original location.
At most two I/Os are required to fetch an indirect row (if the length of the
row changes again and there is not space on the "new" page, the row will move
again.  The pointer on the page of the original TID is updated to point to the
final location).
 
By the way, you can use the STATIC INDIRECT screen in SQLMON to help determine
whether you should unload and reload a table.  It shows the percentage of
indirect rows in each table.
 
Leslie-Anne
 
______________________________________________________________________________
Leslie-Anne Bain
Hewlett-Packard Company                          Voice:  408/447-0666
19447 Pruneridge Ave, MS 47UX                    FAX:    408/447-7902
Cupertino, CA  95014-9913                        E-Mail: [log in to unmask]
______________________________________________________________________________

ATOM RSS1 RSS2