On our quest to learn about Oracle's Data Pump utility it has often been compared to the old export and import (exp & imp) utilities that we have all grown to love (or hate). This article is where where Data Pump takes a detour from these old utilities and begins to shine. This article will explore some of the export modes available and give examples on how to export selected object types and dependencies those objects
Oracle recommends that RAC databases be managed with srvctl, an Oracle-supplied tool that was first introduced with 9i RAC. The 10g version of srvctl is slightly different from the 9i implementation. In this article, we will look at how -- and why -- to manage your 10g databases with srvctl.
Since we are all familiar with Oracle’s original export (exp) utility, and in my opinion Data Pump will be replacing exp soon, I thought it would be good to start off getting familiar with this utility by some relatively simple Data Pump exports (expdp) that are similar to the way we have used exp in the past. In particular the FULL export.
DBAs wanting to create a 10g Real Applications Cluster face many configuration decisions. One of the more potentially confusing decisions involves the choice of filesystems. Gone are the days when DBAs simply had to choose between "raw" and "cooked". DBAs setting up a 10g RAC can still choose raw devices, but they also have several filesystem options, and these options vary considerably from platform to platform. Further, some storage options cannot be used for all the files in the RAC setup. This article gives an overview of the RAC storage options available.
When Oracle released RMAN (Recovery Manager) in Oracle 8 they changed the way databases can be backed up and recovered in the event of disaster. Unfortunately, Oracle shops have been slow to embrace RMAN often times because the change required a leap of faith into the new backup / recovery process as well changes to scripts, procedures, etc. This article will introduce the reader to RMAN and explain why every DBA should use it.
The last couple articles I have written focused on meta-data or DDL extraction for Oracle. The search for a part III to those articles lead me to Oracle's Data Pump utility. Not necessarily for the data movement piece but because it has an API for meta-data. Well even though I have been using 10g for quite some time, I have yet to use Data Pump. I thought this would be a great way to introduce myself, and possibly you the reader, to this new utility. This article will serve as a basic introduction to Data Pump and then in subsequent articles we will walk through the new command line options for Data Pump's export and import (expdp & impdp), and look at the PL/SQL packages DBMS_DATAPUMP and DBMS_METADATA.
This article is the result of observations of the UNDO tablespace of Oracle 9i and Oracle 10g in various situations. We start with a simple query showing how to monitor the amount of undo generated in a session for a specific time. We investigate the creation, expansion, and resize of UNDO tablespace, and the issues that guide the reuse of UNDO segments. The impact of parameters like UNDO_RETENTION in Oracle 9i and UNDO_RETENTION and the GUARANTEE clause in CREATE UNDO statements is discussed using simple reproducible examples.
In this article James continues to explore the Oracle's Metadata API and provides a powerful function to compare objects and schemas and print the DDL required to bring them in sync.
This article shows how Oracle's Heterogeneous Services can be configured to allow a database to connect to a Microsoft Access database using standard databases links. The method described can be used to connect to MS-Access from about any platform - Unix/ Linux or Windows.
In this article James explores the Oracle's Metadata API (DBMS_METADATA) and shows how database users can extract object definitions (DDL statements) from an Oracle database without having to go through a stack of dictionary views.