Friday, July 31, 2009

What does it feel to be twenty?

One day is a very long time if you have to wait for something to happen. Twenty years are as fast as a wink of an eye if you keep on making things happen.

The 80's, the Glowing Decade

Back in the 80's talking about mobile phones was not popular at all. The cassio fx100 was my programmable scientific calculator used to solve integrals by means of numerical methods. In the music, Madonna, Billy Joel, Elton John, Cindy Lauper, Aerosmith, Brian Adams, Dire Straits, Duran Duran, Eurythmics, Peter Gabriel, Huey Lewis & the news, Robert Palmer, Bruce Springsteen, Survivor, Tears for Fears, Van Halen, ZZ Top, only to name a few, were the singers we used to hear with the walkMan (some sort of iPod of the 80's) wich was our preferred music gadget to play audio tapes.
At that time the fluorescent pink, yellow, green and blue were the fashion colors found in every guard robe. People dressed in those colors easily glowed in the night with the Discotheque lights. These noisy colors still survive today only as text markers.
At that time Java was an Island found in the Pacific Ocean and Cobol was still the programming language for the business applications. Meanwhile Pascal was an emerging structured programming language which mostly remained in the research labs. Most of the people related to informatics at the time felt in love with Pascal but ended married with C Language until the word Java changed its widely known meaning to the programming language for the internet.

20 Years of professional Career devoted to Oracle

Oracle has been evolving and it is as challenging, aggressive and innovative as the first time I met it. It was Oracle version 5.0 the first Oracle version I used to issue my first SQL statements. 16 Mb RAM, two hard disks of 512 Mb and 320Mb inside an HP9000 runnig HPUX 9.0 were the leading edge technology standards capable of carrying the data payload required to move the Metropolitan University data. Today this amount of computer power is hilarious, but if compared with the average PC of the time (640K RAM + Extended memory + 40Mb HD + 256 Colors monitor) this was a powerful Server.
My professional career has evolved side by side with Oracle. I have run at the same pace Oracle runs, stuck to it as its shadow. During this time I have seen a lot of people come and go to the Oracle community, and very few of them still join me today. Even though very few things have changed. The two main Kernel directives were not very different from today's kernel directives,
I. High Availability
II. High Performance
Oracle is conceived as a highly fault tolerate rdbms, whose memory and physical structures provide a good balance between recoverability and performance. I have met all the major and minor releases since then up to the latest available 11g release today and I can say that even though Oracle's main directives are the same, each day it becomes more robust and more complex.
It's been a long journey that has run at a very fast pace. I have met a lot of people, I have met a lot of places. And at the end here I am still working with Oracle. I enjoy this professional activity as much as the first time I met Oracle, it's been some sort of falling in love at the first sight that has ended in a marriage so fresh and new as the first time.
This is the train of a never ending trip where a lot of people climb up and others descend from it, but the most important part of it is the trip itself, not the destination.

Thursday, July 23, 2009

ORA-00600: [LibraryCacheNotEmptyOnClose]

Recently a database in 9iR2 (9.2.0.8.0) was migrated to 10gR2 (10.2.0.4.0), this database was already having a hard time with memory issues and frequent ORA-04030 errors. At the time the workaround for this issue was to increase the value of the PGA_AGGREGATE_TARGET instance parameter. Since 10g consumes more resources than 9i this error re-appeared and the local operators found easy to rebounce the oracle instance when this problem was more frequently shown. However, after the migration to 10gR2 this procedures throwed the above mentioned ORA-00600 error.

Googling around I found that this errors is due to a bug (Bug 4483084 OERI[LibraryCacheNotEmptyOnClose] on shutdown), and it is harmless, so this can be ignored. There are people who have never seen this error but they started facing it after migration to 64 bit platforms.

According to Shervin Sheidaei, he stated that purging the memory region prior to the shutdown process using a shutdown event trigger the problem can be worked around.

Code taken from his blog:
CREATE or replace TRIGGER flush_shared_pool
BEFORE SHUTDOWN ON DATABASE
BEGIN
execute immediate 'ALTER SYSTEM FLUSH SHARED_POOL';
EXCEPTION
WHEN OTHERS THEN
RAISE_APPLICATION_ERROR (num => -20000, msg => 'Error flushing pool');
END;

Packt Publishing Press Release

Oracle 10g/11g Data and Database Management Utilities is a new book from Packt that helps Oracle professionals improve performance and manageability using the advanced features of Oracle Utilities. Written by Hector R. Madrid, a certified DBA and Java curriculum instructor for the Oracle University, this book helps users to master the utilities to optimize the efficiency, management, and performance of their daily database tasks.

Database interaction is a part of the daily routine for all database professionals, and by using Oracle Utility tools such as Oracle Data Pump and Oracle SQL*Loader, users can benefit from improved maintenance windows related to data management tasks, optimized backups, faster data transfers, and more reliable security, thereby allowing them to do more with the same amount of time and resources.

Using the advanced features of direct export or import utilities, readers will learn to improve performance and manageability of different databases. Readers will make use of the Oracle Scheduler to specify maintenance windows, assign priorities, configure job classes and many more features. With the help of the Oracle Universal Installer tool, the installation tasks becomes more efficient, and allows users to execute effective default and comprehensive database creations.

Readers will also learn about the Oracle Wallet Manager and how it is used to increase the security in an Oracle environment, protect the backups, and manage its certificates. Using OPatch, users will be able to manage software updates related to Critical Patch Updates (CPU) or individual patches and perform installations in batch environments. They will also learn to use DBCA to create and configure a database and manage it in an ASM environment.

Oracle professionals interested in working with several powerful tools to help improve performance and manageability of utilities among different databases will find this book useful. This book is out now and is available from Packt. Fore more information, please visit http://www.packtpub.com/oracle-10g-11g-data-and-database-management-utilities/book

Does your database look complicated? Are you finding it difficult to interact with it? Database interaction is a part of the daily routine for all database professionals. Using Oracle Utilities the user can benefit from improved maintenance windows, optimized backups, faster data transfers, and more reliable security and in general can do more with the same time and resources.

"Oracle 10g/11g Data and Database Management Utilities" is written using a practical approach that guides you through different practical scenarios. It provides a brief introduction to the topics; this way you can quickly get to know the main features, start being productive with the tool, and grow with it at a fast pace. If you wish to read through the book extract, please feel free to click here: Book Extract

Wednesday, July 22, 2009

My recently published Book

"Oracle 10g/11g Data and Database Management Utilities" is the tittle of my recently released book that was published by Packt Publishing. This book represents for me my debut as an Oracle Author. This book was written after all the cases I have faced for already 20 years of professional carreer as an Oracle Professional.


There are times when a process doesn't seem to end, and when it is about to reach the final stages something happens and it breaks up. If we had a tool to monitor and at lest stop the process before this fatal ending it would be great. The tool actually exists, but the problem with several DBA's or DBO's is that they don't know this tool exist and even if some of them know this exists, very few of them know how to master it so they can use it in a practical environment.
This is not the only case I expose in my book, there are other scenarios that relate to tools such as DataPump. Today there are a lot, of people out there that still talk about the traditional exp/imp. Data Pump is a very old tool, considering this was first released in 2003 along with the 10g Rel. 1 RDBMS, and since the technology gets obsolete in five years (avg.) this can be considered that Data Pump has already had three evolution stages (10gR1, 10gR2, 11gR1 and very soon 11gR2). What about exp/imp, it is more than an obsoleted tool that has already given what it has to give and is maintained only for upward compatibility.

Another tool has to do with sql*loader, you may say it is not new and you are right, but how about using it to load CLOB, BLOB files? Can this be used to create External Tables and have access to log files so they can be analyzed at the database level by means of SQL commands? The answer is yes, and SQL*Loader can be used in a very creative way. You can get a network performance meter by means of a sqlloader exeternal table, you can have another view of the alert log and you can query it directly from the database.

There are several topics I address in the book, but overall I do a practical approach, I introduce practical examples that can help the DBA in the day by day tasks so the DBA can perform its dutys in a more efficient way.

Yet another CRS-0223: Resource 'ResourceName' has placement error

Recently I faced this error. The scenario took place after a full RAC reboot process. The difference with error I had already documented here was that it was useless to restart each RAC service. Either way it was sending the placement error for all RAC related services. In this case I took a look at the Metalink Note: 726925.1 "srvctl start instance fails with PRKP-1001; srvctl trace shows error connecting to CRSD".

First I checked the alert.log file related to the cluster operation. It reported nothing unusual, everything seemed to be normal.
Next I took a look at the output of the command:

crsctl check crs

This command reported everything was working just as normal. Definitely it had to do with the syncronization at startup time. It is pretty weired since a normal node reboot should not lead to such inconsistency. I must point out that the environment used was 10gR2 (10.2.0.1.0) on RHEL4 (Red Hat Enterprise Linux AS release 4 (Nahant Update 3) 2.6.16 xenU (32-bit)), this was a scenario faced while I was teaching the RAC 10g course for Oracle. Since this environment is not patched at the start of the course, I would not be surprised to find out that this is due to an already filed bug.

The procedure was to kill (as root) all crsd.bin process on all participating nodes, then a simple crs_stop -all / crs_start -all was just enough to put everything back to normal.