reorganizing indexes

[bdho...@acm.org]

How does one determine if an index is in need of reorganizing? How
does one reorganize an index?

The information provided by rmu/analyze/indexes does not seem to
include the amount of fragmentation. Am I overlooking something?

Thank you

Rdb 7.0-61
VMS 7.1
emulated VAX 4000-105a
It would be great to upgrade, but those that hold the purse strings
don't want to upgrade. sigh

[Italiano]

In the rare occasion that I have rebuild index was because the min/max
path to reach node's leaf or data row was very different.
Check documentation on RMU/ANALYZE/PLACEMENT, from there you can see
the I/O needed to reach node's leaf and data row.
Wich problem are you facing? Rdb do not have needs to do "RE-ORG" or
"RE-DO" commands.

>... amount of fragmentation...
Infact Rdb do not have this problem unless you are speaking about
"record fragmentation" ("record chain"? using Oracle vocabulary)

ciao
GIo

[bdho...@acm.org]

On Apr 25, 3:31 am, Italiano <gio...@gmail.com> wrote:
> In the rare occasion that I have rebuild index was because the min/max
> path to reach node's leaf or data row was very different.
> Check documentation on RMU/ANALYZE/PLACEMENT, from there you can see
> the I/O needed to reach node's leaf and data row.
> Wich problem are you facing? Rdb do not have needs to do "RE-ORG" or
> "RE-DO" commands.
>
> >... amount of fragmentation...
>
> Infact Rdb do not have this problem unless you are speaking about
> "record fragmentation" ("record chain"? using Oracle vocabulary)

I don't have a problem with indexes or Rdb. I wasn't sure if I needed
to reorganize the indexes myself or if Rdb takes care of it. If I
understand your reply, Rdb reorganizes the indexes as needed.

Many many winters ago, I had to occasionally reorganize the indexes of
some other database system to keep response times in the acceptable
range. I wasn't sure if databases had gotten smart enough to take
care of that task themselves.

Thank you, Giovanni, for information