Re: Different flavors of relational database

Discussion created by ca.portal.admin on May 2, 2006
Onni, what do you do when the 'explain' shows a path that's not optimal?
Rework the SQL statement? Does IDMS/SQL have 'hints' like Oracle?
Besides, the file structures are different on midranges and pc's the
mainframes, and so is the underlying software technology I doubt that
any mainframe database can retrieve data as rapidly as a google search
of anything ALutz Petzold
This e-mail may contain confidential or privileged information If you
think you have received this e-mail in error, please advise the sender
by reply e-mail and then delete this e-mail immediately Thank you Aetna

IDMS Public Discussion Forum


FW: Different flavors of relational database
"Richard has raised some interesting points here, which of course has
generated more questions on my part:

1) I think the default index in IDMS/SQL serves the purpose below, but
we always drop ours as soon as an alternate key (CALC or index) is
defined. I am now questioning the wisdom of that. Perhaps the default
index should be kept in some cases. Would a CALC chain serve the same

2) The table in question was stored CALC. The area was only about 20%
full for the test, so we would have been doing a lot more I/O than with
a smaller area. Do you think this extra I/O could account for the large
difference in response time?

3) I don't think either SQL Server nor DB2 has a hash storage method.
Does anyone know if their storage methods leave free space embedded
within the data? I am trying to determine if their database would have
been physically smaller, thus requiring less I/O.

4) This table was originally stored via an index. We experienced
terrible page contention at INSERT time, and switched to CALC to
alleviate that, which worked beautifully. I have to wonder if SQL
Server or DB2 could handle the online production INSERT volume.