RES: RES: Delivery by Demand

2005-11-04 Thread Fabricio Mota
OK, Shawn, nice tip.
I really didn't know if it was actually performed by server or by client.
But I'll study the MySQL client protocols.

But it is still strange and needs investigation, because as I remember, I've
submit a prove fire to the oracle server. The prove was:
I've sent a really really really heavy query, containing several tables
cartesian joins, resulting in too many data (billions of billions of
records), and we've imaginate it could spend at least many hours. For my
surprising, the server reponse was imediate, with a only first page.

Do you think that the server continues the actual processing until the end,
despite no more pages being requested anymore?

My Regards

FM
  -Mensagem original-
  De: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]
  Enviada em: sexta-feira, 4 de novembro de 2005 02:06
  Para: Fabricio Mota
  Cc: mysql@lists.mysql.com
  Assunto: Re: RES: Delivery by Demand



  Yes, it is a client-side behavior to the extent that the MySQL server does
not page through data. It gets the complete results unless you ask for a
LIMIT, then it stops building results after it meets the criteria of your
LIMIT.

  I really do not know much about Oracle administration and communication
protocols so I am just guessing.I believe that even your Oracle clients
had to ask for data in pages instead of the full set.  Are you sure your
Oracle server was really holding those results for you and only delivering
batches of 100 records?  That seems very much like a client-side behavior
that was just hard for you to notice. It could have been designed as part of
the client library As I said, I just don't know but I know others on the
list have had some extensive Oracle experience. Maybe one of them can weigh
in on this

  With MySQL, the behavior you want to emulate is definitely something you
control from the client-end either by using the LIMIT clause or by pulling
down single rows in batches. You have to remember, though, that while the
client is processing it's latest batch of rows that the server still holds
onto a complete result set and has to maintain an open connection to your
client. It's really in the best interest of performance for your client to
spend as little time getting the data from the server. That means that you
should only write queries that ask for the data you actually need and you
should get the data out of the server as soon as possible. That way the
server has more resources available to deal with other queries.

  Shawn Green
  Database Administrator
  Unimin Corporation - Spruce Pine



  Fabricio Mota [EMAIL PROTECTED] wrote on 11/03/2005 10:52:34 PM:

   Shawn,
  
   So are you telling me that it's a configuration in Client, but not in
   Server?
  
   FM
 -Mensagem original-
 De: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]
 Enviada em: quarta-feira, 2 de novembro de 2005 17:37
 Para: Fabricio Mota
 Cc: mysql@lists.mysql.com
 Assunto: Re: Delivery by Demand
  
  
  
  
 Fabricio Mota [EMAIL PROTECTED] wrote on 11/02/2005 10:23:46
AM:
  
  Hi all,
 
  In the past, I worked as a Oracle user. I've noted that in oracle
(or
   maybe
  in that configuration), when we request a great amount of data, such
   like:
 
  select * from million_records_table
 
  It does not delivers the entire table at the first moment. It
delivers a
  little amount of data - such like a single page containing about 100
  records - and awaits the cursor request the Record No 101 for fetch
the
   next
  set of data.
 
  In MySQL - at least, with the default configuration I have used - it
   does
  not happens. It delivers all the million of records existent in the
   table,
  inconditionally. I know that there is the clause LIMIT N, to limit
the
  first N records existing in the query, but is there a way to warrant
a
  delivery by demand, such Oracle does, without to have to alter the
  application's SQL code?
 
  Thank you
  
 I know that in the C-API (and others) there are two commands to
retrieve
   records from the server. One is mysql_store_result() which will bring
all of
   your results into your machine in a single pull.
  
 The second is mysql_use_result(). That command sets up a transfer
process
   of pulling the rows from the server one at a time. If you need 100 rows
of
   data, you issue 100 mysql_fetch_row() commands. You are responsible for
   storing the records for later re-use.
  
 Is that the flexibility you are looking for?
  
 http://dev.mysql.com/doc/refman/5.0/en/mysql-use-result.html
  
 Shawn Green
 Database Administrator
 Unimin Corporation - Spruce Pine


RES: Heikki: What will become of InnoDb once MySQL license runs out

2005-11-04 Thread Fabricio Mota
There are several opinions... I would like to forecast the real intentions
of Oracle...

FM

-Mensagem original-
De: Ezra Taylor [mailto:[EMAIL PROTECTED]
Enviada em: sexta-feira, 4 de novembro de 2005 23:46
Cc: mysql@lists.mysql.com
Assunto: Re: Heikki: What will become of InnoDb once MySQL license runs
out


To Mysql users:
 Just to remind you all, Oracle is a
business that expects to make money.  As you all know, Mysql is a
threat to the fat cats such as Oracle,DB2,MSSql and others.  If you
think Oracle is going to play fair with us then you will believe that
crack will one day be a multi vitamin.  For those you that don't know
what crack is, it's a drug that will fuck your life up.


Ezra Taylor

On 11/4/05, Heikki Tuuri [EMAIL PROTECTED] wrote:
 Mike,

 - Original Message -
 From: mos [EMAIL PROTECTED]
 Newsgroups: mailing.database.myodbc
 Sent: Saturday, November 05, 2005 12:11 AM
 Subject: Heikki: What will become of InnoDb once MySQL license runs out


  Heikki,
  I am about to start a large MySQL project that requires transactions and
I
  need to know if InnoDb will be around for MySQL after MySQL's license
for
  InnoDb runs out in 2006.

 the current GPL version of MySQL/InnoDB will of course be available then
by
 the very nature of the GPL license.

 The MySQL AB - Innobase Oy OEM agreement is about commercial non-GPL
 MySQL/InnoDB licenses. About that agreement I want to refer to the
official
 press release of Oracle Corporation:
 http://www.oracle.com/innodb/index.html

  If yes, will you still be supporting it or will
  it be up to MySQL AB?

 I want to refer to the official press release where Charles Rozwat,
Oracle's
 Executive Vice President in charge of Database and Middleware Technology
 says: Oracle intends to continue developing the InnoDB technology and
 expand our commitment to open source software.

  TIA
 
  Mike

 Regards,

 Heikki Tuuri
 Vice President, server technology
 Oracle Corporation


 --
 MySQL General Mailing List
 For list archives: http://lists.mysql.com/mysql
 To unsubscribe:
http://lists.mysql.com/[EMAIL PROTECTED]



--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:
http://lists.mysql.com/[EMAIL PROTECTED]


-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]



RES: Delivery by Demand

2005-11-03 Thread Fabricio Mota
Shawn,

So are you telling me that it's a configuration in Client, but not in
Server?

FM
  -Mensagem original-
  De: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]
  Enviada em: quarta-feira, 2 de novembro de 2005 17:37
  Para: Fabricio Mota
  Cc: mysql@lists.mysql.com
  Assunto: Re: Delivery by Demand




  Fabricio Mota [EMAIL PROTECTED] wrote on 11/02/2005 10:23:46 AM:

   Hi all,
  
   In the past, I worked as a Oracle user. I've noted that in oracle (or
maybe
   in that configuration), when we request a great amount of data, such
like:
  
   select * from million_records_table
  
   It does not delivers the entire table at the first moment. It delivers a
   little amount of data - such like a single page containing about 100
   records - and awaits the cursor request the Record No 101 for fetch the
next
   set of data.
  
   In MySQL - at least, with the default configuration I have used - it
does
   not happens. It delivers all the million of records existent in the
table,
   inconditionally. I know that there is the clause LIMIT N, to limit the
   first N records existing in the query, but is there a way to warrant a
   delivery by demand, such Oracle does, without to have to alter the
   application's SQL code?
  
   Thank you

  I know that in the C-API (and others) there are two commands to retrieve
records from the server. One is mysql_store_result() which will bring all of
your results into your machine in a single pull.

  The second is mysql_use_result(). That command sets up a transfer process
of pulling the rows from the server one at a time. If you need 100 rows of
data, you issue 100 mysql_fetch_row() commands. You are responsible for
storing the records for later re-use.

  Is that the flexibility you are looking for?

  http://dev.mysql.com/doc/refman/5.0/en/mysql-use-result.html

  Shawn Green
  Database Administrator
  Unimin Corporation - Spruce Pine


Delivery by Demand

2005-11-02 Thread Fabricio Mota
Hi all,

In the past, I worked as a Oracle user. I've noted that in oracle (or maybe
in that configuration), when we request a great amount of data, such like:

select * from million_records_table

It does not delivers the entire table at the first moment. It delivers a
little amount of data - such like a single page containing about 100
records - and awaits the cursor request the Record No 101 for fetch the next
set of data.

In MySQL - at least, with the default configuration I have used - it does
not happens. It delivers all the million of records existent in the table,
inconditionally. I know that there is the clause LIMIT N, to limit the
first N records existing in the query, but is there a way to warrant a
delivery by demand, such Oracle does, without to have to alter the
application's SQL code?

Thank you


RES: Hi all

2005-10-31 Thread Fabricio Mota
In true, I have a project to develop a new (revolutionary) engine for mySQL,
and so increase many times applications performance, and reduce enterprises
cost with hardware infraestruture. So I'm looking for peoples to help me to
develop this project - or appoval by MySQL AB.

My regards,

Fabricio Mota

-Mensagem original-
De: Stewart Smith [mailto:[EMAIL PROTECTED]
Enviada em: segunda-feira, 31 de outubro de 2005 05:08
Para: Fabricio Mota
Cc: mysql@lists.mysql.com; [EMAIL PROTECTED];
[EMAIL PROTECTED]
Assunto: Re: Hi all


On Sun, 2005-10-30 at 00:14 -0300, Fabricio Mota wrote:
 Anybody has worked as a developer to MySQL Database Server features?

If you mean people who work for MySQL hacking on Cluster, then yes, we
exist.
--
Stewart Smith, Software Engineer
MySQL AB, www.mysql.com
Office: +14082136540 Ext: 6616
VoIP: [EMAIL PROTECTED]
Mobile: +61 4 3 8844 332

Jumpstart your cluster:
http://www.mysql.com/consulting/packaged/cluster.html


-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]



RES: RES: Hi all

2005-10-31 Thread Fabricio Mota
Ok, could you suggest me where I may find it with more occurrences?

Another problem is: MySQL is a very big project, and I have no UML (or any
abstraction) that could help me to understand it!! Can you tell me where
could I find this kind of documentation?

My Regards

Fabricio Mota

-Mensagem original-
De: Stewart Smith [mailto:[EMAIL PROTECTED]
Enviada em: segunda-feira, 31 de outubro de 2005 23:14
Para: Fabricio Mota
Cc: mysql@lists.mysql.com; [EMAIL PROTECTED];
[EMAIL PROTECTED]
Assunto: Re: RES: Hi all


On Mon, 2005-10-31 at 21:16 -0300, Fabricio Mota wrote:
 In true, I have a project to develop a new (revolutionary) engine for
mySQL,
 and so increase many times applications performance, and reduce
enterprises
 cost with hardware infraestruture. So I'm looking for peoples to help me
to
 develop this project - or appoval by MySQL AB.

You certainly don't need anybody's permission. (distributing under a
non-GPL license would). Best bet is to start hacking and ask when you
get stuck.

There are a couple of articles around on writing a storage engine - as
well as the best doc ever - the code.

(btw, this probably should be on internals, not mysql, cluster or
plusplus list)
--
Stewart Smith, Software Engineer
MySQL AB, www.mysql.com
Office: +14082136540 Ext: 6616
VoIP: [EMAIL PROTECTED]
Mobile: +61 4 3 8844 332

Jumpstart your cluster:
http://www.mysql.com/consulting/packaged/cluster.html


-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]



RES: RES: Hi all

2005-10-31 Thread Fabricio Mota
Ok, David. Thank you, 

I'll take a good look on it!

My Regards,

Fabricio Mota

-Mensagem original-
De: Logan, David (SST - Adelaide) [mailto:[EMAIL PROTECTED]
Enviada em: segunda-feira, 31 de outubro de 2005 23:36
Para: Fabricio Mota; Stewart Smith
Cc: mysql@lists.mysql.com; [EMAIL PROTECTED];
[EMAIL PROTECTED]
Assunto: RE: RES: Hi all


Hi Fabricio,

You can find a lot of information in the manual, more specifically at
http://dev.mysql.com/doc/refman/5.0/en/example-storage-engine.html

This would be a very good place to start. Otherwise, as suggested, ask
questions on the internals list.

Regards

David Logan
Database Administrator
HP Managed Services
148 Frome Street,
Adelaide 5000
Australia

+61 8 8408 4273 - Work
+61 417 268 665 - Mobile
+61 8 8408 4259 - Fax


-Original Message-
From: Fabricio Mota [mailto:[EMAIL PROTECTED] 
Sent: Tuesday, 1 November 2005 12:54 PM
To: Stewart Smith
Cc: mysql@lists.mysql.com; [EMAIL PROTECTED];
[EMAIL PROTECTED]
Subject: RES: RES: Hi all

Ok, could you suggest me where I may find it with more occurrences?

Another problem is: MySQL is a very big project, and I have no UML (or
any
abstraction) that could help me to understand it!! Can you tell me where
could I find this kind of documentation?

My Regards

Fabricio Mota

-Mensagem original-
De: Stewart Smith [mailto:[EMAIL PROTECTED]
Enviada em: segunda-feira, 31 de outubro de 2005 23:14
Para: Fabricio Mota
Cc: mysql@lists.mysql.com; [EMAIL PROTECTED];
[EMAIL PROTECTED]
Assunto: Re: RES: Hi all


On Mon, 2005-10-31 at 21:16 -0300, Fabricio Mota wrote:
 In true, I have a project to develop a new (revolutionary) engine for
mySQL,
 and so increase many times applications performance, and reduce
enterprises
 cost with hardware infraestruture. So I'm looking for peoples to help
me
to
 develop this project - or appoval by MySQL AB.

You certainly don't need anybody's permission. (distributing under a
non-GPL license would). Best bet is to start hacking and ask when you
get stuck.

There are a couple of articles around on writing a storage engine - as
well as the best doc ever - the code.

(btw, this probably should be on internals, not mysql, cluster or
plusplus list)
--
Stewart Smith, Software Engineer
MySQL AB, www.mysql.com
Office: +14082136540 Ext: 6616
VoIP: [EMAIL PROTECTED]
Mobile: +61 4 3 8844 332

Jumpstart your cluster:
http://www.mysql.com/consulting/packaged/cluster.html


-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]


-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]



Hi all

2005-10-29 Thread Fabricio Mota
Hi all,

Anybody has worked as a developer to MySQL Database Server features?

Thank you

Fabricio Mota


--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]