Re: Informix beats Oracle
DA Morgan wrote:
> Not sure why you listed #2 unless I misunderstand your use of
> "lightweight."
I meant it's lighter than Oracle. Uses less hardware.
> Not sure why you listed #3 either. Given the current price of disk the
> Oracle footprint works out to about $11.25 at retail. Not exactly the
> kind of thing anyone is going to even notice with an enterprise
> database.
Glad to see that you use such a price for your "enterprise" storage systems.
Has you perfectly know, the price you pay per MB in enterprise systems has
nothing to do with this. Besides, the problem is not necessarily the price you
pay for the storage it needs. It's all that is associated with it. Let me
remind you that one of the features of IDS 11 is the configurable footprint.
If you bother to check the values you can save you will certainly put a smile
on your face... I did. But do you want to know why IBM did it? Customers
demand... Sometimes they want the database installation *really* small... try
to explain them that their request doesn't make sense...
> #6 makes sense but given Oracle has had locally managed tablespaces
> since 8i (1999) its a bit of old news.
You certainly know that things take time on customers side... Nevertheless we
could call it an IDS "innovation", since it used it ages before Oracle
implemented it...
> #7 fascinates me but I've yet to see anyone detail what specifically
> qualifies as "simplicity." What is it that is simple in product I but
> not simple in product O?
That's why you would need to actually know the product... I could give you
details, but they'd be my feelings. Things like onstat, ontape, onmode etc.
> Can't disagree with #8 but then I expect IBM would raise the price to
> what the market would bear if it would bear higher.
This is called evolution. Brought to you by knowledge... A few posts ago you
were trying to prove the opposite. Glad I could help ;)
> No #9.
Secrete feature... Definitively a killer!!!! :)
> #10 is dead with 10g as Rule Based is history. But third-party apps
> not being well written and not being kept up-to-date is hardly the
> fault of the underlying database product. Put those apps on SQL Server
> or DB2 and they are equally rotten.
They were bad written because someone told them to use that RBO instead of CBO.
We could consider this a nice feature of Informix since it always used CBO.
It's mature and works very well. By the way, the same applications were running
on DB2 CBO of course... Siebel is an example. Don't bother to look for, they
already moved to CBO.
> Ontape (#11) is good and has traditionally been the best but the
> capabilities added to RMAN in 10g (single block recovery of an online
> object) has it trumped.
>
> Truth is few people I know back up to tape anymore anyway. We all back
> up to cabinets full of SATA drives and ship the drives off-site.
I've done "block recovery" with ontape... Just a matter of putting dd to work
along it :). And when you need block recovery one of this happened:
- You made a mistake
- You hardware sucks
- Your database has bugs
You'd be amazed by the number of people that use ontape... One simple use for
it, is backip up to disk and allow the fs backup to send it to the nice
cabinets... this can impact your backup costs as you won't need the database
agent license. Just an example.
> I'd be interested in comparing limitations (#13). What limitations are
> you referring to? With regard to indexes ... all indexes in Oracle can
Add a column with a default value. Not exactly sure about 10g, but it usually
translate into a single transaction. Heavy on big tables. Informix makes this
an "inplace alter". Only the table definition is changed and this is immediately.
Btree scanner works behind the scenes to clean up your indexes. It's there for
ages, but has recently received lots of improvements.
> I'll discount 14 and 15 on your list as those are not "features" in the
> sense that the term is normally used but I will look into #1 as it is
> the only one I am not familiar with.
>
> Thanks for helping Ian out though they are not the 20 HA features
> he was pontificating about.. <g>
You keep writing without reading. I explicitly wrote that most of these were
not features, but reasons why we like to work with Informix. I also wrote that
this post was not intended to speak on behalf of others. Just my thoughts.
If you're not familiar with DSA (Dynamic Scalable Architecture) you really
should, or at least stop trying to give the idea you know Informix. It's the
IDS internal basics. It's another lovable non-feature. You'll probably find old
stuff about it. At a certain point in time Informix/IBM decided that this kind
of information was to technical to appeal to customers. Back into the nineties
I'd bet any PPT about Informix started with a slide or two about DSA.
Regards.
--
Fernando Nunes
Portugal
http://informix-technology.blogspot.com
My email works... but I don't check it frequently...
Fernando Nunes wrote:
> DA Morgan wrote:
>
>> Not sure why you listed #2 unless I misunderstand your use of
>> "lightweight."
>
> I meant it's lighter than Oracle. Uses less hardware.
I didn't disagree. But my point was that the difference works out to
about $11.25 at retail ... in other words the difference given today's
hardware prices is inconsequential.
>> Not sure why you listed #3 either. Given the current price of disk the
>> Oracle footprint works out to about $11.25 at retail. Not exactly the
>> kind of thing anyone is going to even notice with an enterprise
>> database.
>
> Glad to see that you use such a price for your "enterprise" storage
> systems.
> Has you perfectly know, the price you pay per MB in enterprise systems
> has nothing to do with this. Besides, the problem is not necessarily the
> price you pay for the storage it needs. It's all that is associated with
> it. Let me remind you that one of the features of IDS 11 is the
> configurable footprint.
> If you bother to check the values you can save you will certainly put a
> smile on your face... I did. But do you want to know why IBM did it?
> Customers demand... Sometimes they want the database installation
> *really* small... try to explain them that their request doesn't make
> sense...
I'm trying to follow you here but I'm a bit lost.
I can run essentially any RDBMS in 256MB of RAM on any Pentium. If I
don't need much more than that I go with MySQL or Oracle's free
Application Express. If I need more than that then I'm running SAP or
Siebel or PeopleSoft, etc. so the horsepower required is, again,
financially inconsequential. I can't imagine made the change they did
for the reason you list. Seems more likely it was due to their desire
to keep it from competing with DB2 and sell as an embedded database:
What CA did with Ingres before they through in the towel and gave it away.
>> #6 makes sense but given Oracle has had locally managed tablespaces
>> since 8i (1999) its a bit of old news.
>
> You certainly know that things take time on customers side...
> Nevertheless we could call it an IDS "innovation", since it used it ages
> before Oracle implemented it...
In that sense you are correct. I tend to think of innovation as being
something done during the current millenium. <g>
>> #7 fascinates me but I've yet to see anyone detail what specifically
>> qualifies as "simplicity." What is it that is simple in product I but
>> not simple in product O?
>
> That's why you would need to actually know the product... I could give
> you details, but they'd be my feelings. Things like onstat, ontape,
> onmode etc.
Lets take ontape for point of discussion. Here's what it takes to backup
and Oracle database:
$ rman target /RMAN> backup database;
can you make it simpler than that?
If you did would the benefit be palpable?
>> No #9.
>
> Secrete feature... Definitively a killer!!!! :)
Well then SQL Server has us all beaten. They've got more features that
don't exist than Informix and Oracle combined.
>> #10 is dead with 10g as Rule Based is history. But third-party apps
>> not being well written and not being kept up-to-date is hardly the
>> fault of the underlying database product. Put those apps on SQL Server
>> or DB2 and they are equally rotten.
>
> They were bad written because someone told them to use that RBO instead
> of CBO.
In some cases you are correct. But back in the 1980s that was all there
was: CBO showed up in 7.3 and it was, quite frankly, awful. Given the
providence of many of these products ... mainframe ported to RDBMS with
zero application of synapses this is not surprising.
> We could consider this a nice feature of Informix since it always used
> CBO. It's mature and works very well. By the way, the same applications
> were running on DB2 CBO of course... Siebel is an example. Don't bother
> to look for, they already moved to CBO.
Sippl and King formed Informix in 1980 several years after Larry and
company had a released product. No question in those days Informix was
at the cutting edge.
But in the field of software and technology where sea-change change 3-5
years ... that really is ancient history.
The question I would put to you with respect to innovation would be to
compare the features of Cheetah, just released, with Oracle 11g, soon to
be released. Is the innovation there or is it playing catch-up with SQL
Server?
>> Ontape (#11) is good and has traditionally been the best but the
>> capabilities added to RMAN in 10g (single block recovery of an online
>> object) has it trumped.
>>
>> Truth is few people I know back up to tape anymore anyway. We all back
>> up to cabinets full of SATA drives and ship the drives off-site.
>
> I've done "block recovery" with ontape... Just a matter of putting dd to
> work along it :). And when you need block recovery one of this happened:
> - You made a mistake
Never the cause.
> - You hardware sucks
Hard disks are hard disks. It happens.
> - Your database has bugs
Never the cause.
> You'd be amazed by the number of people that use ontape... One simple
> use for it, is backip up to disk and allow the fs backup to send it to
> the nice cabinets... this can impact your backup costs as you won't need
> the database agent license. Just an example.
Actually I wouldn't be. I expect 95+% of Informix shops use it. On the
Oracle side RMAN is, and has always been, free. There is no such entity
as separate licensing.
>> I'd be interested in comparing limitations (#13). What limitations are
>> you referring to? With regard to indexes ... all indexes in Oracle can
>
> Add a column with a default value. Not exactly sure about 10g, but it
> usually translate into a single transaction. Heavy on big tables.
> Informix makes this an "inplace alter". Only the table definition is
> changed and this is immediately.
In Oracle:
ALTER TABLE <table_name> MODIFY (<column_name> DEFAULT <default_value>);
Standard functionality in all products TTBOMK.
> Btree scanner works behind the scenes to clean up your indexes. It's
> there for ages, but has recently received lots of improvements.
As does the corresponding background process in Oracle: No difference.
Anyone that claims otherwise is smoking dope.
> You keep writing without reading. I explicitly wrote that most of these
> were not features, but reasons why we like to work with Informix.
And I wasn't disagreeing. I was just looking back to my original
challenge. I understood your point.
--
Daniel A. Morgan
University of Washington
damorgan@x.washington.edu (replace x with u to respond)
Puget Sound Oracle Users Group
www.psoug.org