Serge Rielau wrote:[color=blue]
> darko wrote:[color=green]
> > Serge Rielau je napisao:[color=darkred]
> >> Bob Jones wrote:
> >>>> We're talking migration here. We're talking that you can often accomplish
> >>>> the same thing with two different databases ... using similar features or
> >>>> substitute functionality that one vendor has chosen to implement somewhat
> >>>> differently.
> >>> Perhaps you can elaborate on how these features/functionalities are
> >>> implemented in DB2.
> >>>
> >>> 1. Dropping a table column.[/color]
> >
> > It is a shame that DB2 is not (yet) capable to do this. It was large
> > disappointment for us, coming from Informix world. We need it NOW
> > (being in the development phase, developers demand such actions on a
> > weekly basis). What took you so long to implement this?[/color]
> A wise man once said products at IBM don't get delivered, they escape.
> Schema evolution is one such case. July 28th is the date.
> Since you are in the development phase, have you looked at the ALTER
> TABLE in Control Center? It does DROP COLUMN today.
>[/color]
Correct me if I am wrong:
It does by a sequence of activities that include unloading data from
the table, probably saving definitions of constraints, views etc.
related to that table, dropping (or renaming) it, recreating without
the column that was supposed to be dropped, restoring constraints,
views etc, reloading the data into the table. It is much easier that
way (through CC) than it would be manually, but it is painful when the
table is populated (we do have some small part of functionality in
production :). And CC, among other things being a Java app, is not the
speed (or low-resource-consumption) champion, and so far it gave us a
share of troubles. Very often my colleagues lament for lack of a tool
like ServerStudio for Informix, which for a decent price provides a
lot. Can someone suggest us an analogous tool for UDB?
[color=blue][color=green][color=darkred]
> >>> 2. Creating a bitmap index.[/color]
> > I think we just might make use of this feature if we had it.[/color]
> Might?
>[color=green][color=darkred]
> >>> 3. Moving an index from one tablespace to another or having indexes for a
> >>> table placed in different tablespaces.
> >>> 4. Changing default tablespace for an index.[/color]
> > Definitely, I would like to see these features in UDB.[/color]
> ... because you are used to them in IDS?
> BTW. In DB2 9 you can create a ranged partitioned table with one open
> range and place your indexes as you wish.
>[/color]
It seems to me that that feature is coming from Informix. Welcome!
[color=blue][color=green]
> > IMHO, it seems that DB2 UDB development is targeted in such a way that
> > the product should be attractive to corporate bean counters rather than
> > to DBAs and developers. In long term, I believe that database product
> > should be marketed to IT people and tailored to their needs primarily.
> > Market trends will show which the right way is.
> >
> > The recent Gartner and IDC reports on global market share of RDBMS
> > shows that Oracle holds >40% of total market share, while IBM is
> > <25%. (I am interested to hear which part of that percentage belongs to
> > Informix). If I remember well, after Informix takeover in 2001, IBM
> > held larger share that Oracle (according to the info I got from our
> > Informix/IBM representatives).[/color]
> You remember the Gartner numbers, not IDCs. In Gartner's counting (new
> licence revenue - at least in the past) Oracle and IBM RDBMS were and
> still are neck to neck. At the time of purchase Informix boosted that
> number for IBM by some 3% only (Gartner actually split Informix out in
> the first year). I'm sure you can find the actual number somewhere.
> The IDC numbers always (at least since I monitor them) put Oracle higher
> than IBM RDBMS.
>[color=green]
> > Being long-term Informix and IBM
> > customer, I wish IBM to increase its market share, among other means,
> > by implementing features that IT people want and need. It is a good
> > thing to offer automation for administration, but we need means for
> > setting things up manually at occasions. Informix did offer such means,
> > and I suppose Oracle does also.[/color]
> One of the claims to fame for Informix is the number of instances of IDS
> a single DBA can administer. For an application vendor or customer this
> is a compelling purchase reason.
> I visit a fair number of customers and business partners (including some
> posting here) and that message is very clear.
> Each individual customer may have their own pet-peeve, feature request,
> but they all agree that the more the DBMS does itself the better it is.
>[color=green]
> > I speak for myself only, and not for my employer.[/color]
> Who makes the purchase decisions, you or your employer?
> If your employer can fire half of your staff (let's make that "redeploy
> into value generating tasks") because of his choice of DBMS he will.
> A DBA on the other hand has little interest of making him/herself
> obsolete, learning a new product or completely new skill.
>[/color]
In our particular case, purchase decision, when related to IT, is made
according to recommendation from the IT. In fact, I was involved in
making decision to pick DB2 UDB for the system under development (at
the moment when Informix's future did not seem very promising regarding
improvements; the support was never an issue). We were all very
optimistic regarding UDB's future, expecting it will unify all the best
from both DB2 and IDS. We are not pessimists now, but are still waiting
for some promises to be delivered (like onstat commands...).
In my country, largest companies are in the size range of western SMBs,
so it's quite possible that I don't have a clue regarding the way of
doing big bucks business. But I believe that we are not the only one
whose management listens to IT when making decisions that relate to IT.
With IDS, I was able alone to administer more than a dozen of servers
in a low-capacity lines WAN, using very modest hardware, not very
efficient apps and database schemas, ancient OS (SCO) and achieving
uptimes up to >1 year. We have yet to prove that we will lessen the
labor needed to maintain UDB (I hope we will).
I have no worry that our DBA and system programming group will be
reduced, quite the contrary. It seems that data volumes, as well as
types of information that business needs, are increasing faster than
hardware and (self-managed) software are following. It is good that
there is automation available. It's just that we could make our clients
(which are other departments, that are using our services and makes
revenue to the company) happier if we were able to tune some things
manually, instead of delivering slower response or asking for larger
hammer :) (more powerful server). In a way it relates to points 3 and 4
in a list (regarding indexes, tablespaces and two smoking barrels :)
It seems to me that we are now well off-topic the original theme of the
thread. I take my share of responsibility for that.
[color=blue][color=green]
> > English is not my
> > native language, accordingly excuse my language mistakes.[/color]
> Mine neither, we shall excuse each other ;-)
>
> Cheers
> Serge
>
> --
> Serge Rielau
> DB2 Solutions Development
> IBM Toronto Lab
>
> IOD Conference
>
http://www.ibm.com/software/data/ond...ness/conf2006/[/color]
BTW, I am grateful for many answers and advices that you have shared
with this newsgroup, Serge. Your posts are important help for our trip
to DB2 UDB land.
Darko Krstic
Chief DBA
Nova banka ad
B&H
Disclaimer:
I speak for myself only, and not for my employer.
English is not my native language, accordingly excuse my language
mistakes.
