Subscribe:

Monday, 1 November 2010

Something I learned about Oracle Database 11g RMAN restore


Last weekend (it was saturday night), I needed to restore a Development database from a old backup. I never did a RMAN restore before until last saturday. As the saying goes, “necessity is the mother of invention”. Though it is not really an invention (rman is there for a long time), for me, I learned to know about RMAN restore last week.
Our DBA was not available on Saturday. I needed to test few things on the Development and for that I need to restore a backup that was taken earlier couple of months ago. So I did the follow the procedures to restore the database using RMAN.
This could be a basic thing all the DBAs know about. But for me this is not something I do everyday. So this was new to me.
First, I ran the “shutdown immediate” command to shutdown my development database. Then I followed these steps to restore the database from a older backup taken by RMAN. Database was running on the Redhat Enterprise Linux Machine and the database version was 11.1.1.6.0.
$ rman

RMAN> list backup;
List of Backup Sets
===================

…….
I got the TAG details from here.
……
RMAN>  restore datafile ‘/u02/oradata/OIMTST/system01.dbf’ from tag = ‘BEFORERECON’;

RMAN>  restore datafile ‘/u02/oradata/OIMTST/sysaux01.dbf’ from tag = ‘BEFORERECON’;
….
RMAN> restore datafile ‘/u02/oradata/OIMTST/undotbs01.dbf’ from tag = ‘BEFORERECON’;

RMAN> restore datafile ‘/u02/oradata/OIMTST/users01.dbf’ from tag = ‘BEFORERECON’;

RMAN> restore datafile ‘/u02/oradata/OIMTST/oimtst4_tspace_01.dbf’ from tag = ‘BEFORERECON’;

RMAN> list backup of controlfile;

RMAN> restore controlfile to ‘/u02/oradata/OIMTST/control01a.ctl’ from tag = ‘TAG20100820T112653′

RMAN> quit
Recovery Manager complete.

$
Copying the Control Files:
============================
cd /u02/oradata/OIMTST  # The conrol files are located here.
cp control01a.ctl control01.ctl
cp control01a.ctl control02.ctl
cp control01a.ctl control03.ctl

$ sqlplus / as sysdba….
SQL> startup
ORACLE instance started.
Total System Global Area 1073131520 bytes
Fixed Size                  2151248 bytes
Variable Size             264244400 bytes
Database Buffers          801112064 bytes
Redo Buffers                5623808 bytes
Database mounted.
ORA-01589: must use RESETLOGS or NORESETLOGS option for database open

SQL>
SQL> alter database open resetlogs;

Database altered.
SQL>
Hurray!!!! It is success!!!
This was my first restore using RMAN. I knew the concepts earlier, but I didn’t really restore a database like this before. I thought of sharing this knowledge.
We will meet in another post. Until then

Wednesday, 29 September 2010

Checking on Oracle Fusion Applications


Sun Blogger Vijay Tatkar wrote in his blog about the eight technology innovations mentioned by Larry Ellision during his Oracle Open World Keynote speech during last week. Nearly half of them were Sun Hardware related (such as Exadata, ExaLogicSun ultraSPARC t3 etc). Here is the list:

  1. Fusion Apps
  2. Unbreakable Linux Kernel
  3. Solaris Express 11
  4. unltraSPARC t3 chip
  5. mySQL 5.5
  6. exadata
  7. exalogic
  8. Java 7 and 8
Since the beginning, I am always interested to know more about Fusion Apps, mainly out of curiosity. Oracle Fusion Applications were formally introduced during the Oracle Open World last week (during Open World 2010). As per Oracle Release, this was one of the major innovation or next big thing for Oracle. As you are aware, Fusion Middleware Products were released already. Now, it is time to talk about the Fusion Applications.
You may be already aware; I started my IT career as a Web Developer in a small web hosting company. I used to write perl CGI code and hosting them on Linux Servers running with Apache Web Server and mySQL Database. I got bored (or I wanted a change, I am not sure!) with that job and then moved into the Unix System Administration. I worked as a Sun Solaris Admin and HP Unix Admin for some time. Then I worked in both Peoplesoft System Administration for nearly 7 years and currently working in an Oracle Identity Management (which is part of Oracle Fusion Middleware products) project for nearly past one year.
So, the question is “now what?” And how can we develop the knowledge for Fusion Apps Administration.
I am not sure when Fusion Apps will be deployed full-fledged instead of the other ERP Applications like Peoplesoft. I don’t think it will be near soon, but may be after few years, Oracle may buy in customers who are going to do a new implementation of some ERP Applications.
You know what, Fusion middleware for Fusion Apps is like PeopleTools for Peoplesoft Applications. PeopleTools Technology is the abstract layer on top of which all the Peoplesoft Applications run on. PeopleTools was originally built on C and C++ and finally evolved into a Java Technology. However I still feel some of them are C++ code. While Fusion Middleware is more Java and J2EE apps, I believe Fusion Apps will be more J2EE apps than Peoplesoft. I need to spend little bit more time on implementing a Fusion Middleware and an application. As of now, I only worked on Identity Management Product Sure and little but of Oracle Portal Technologies.
For an IT Infrastructure Administrator like me (who mainly works on Oracle Server Technologies), I think understanding the Fusion Middleware Stack will be important.
Talk to you later. Until then

Thursday, 12 August 2010

Peoplesoft Connectors for Oracle Identity Manager – Part I


Introduction

Couple of weeks ago, I attended an Oracle Webcast titled “Introducing Oracle Identity Management 11g”. That webcast was about introducing the remaining components of Oracle Identity Management Product Suite which is part of the Oracle Fusion Middleware 11g (we can call it as a second set of product release!).

During the first phase release of Oracle Fusion Middleware Components, Oracle released the few components such as Oracle Internet Directory (OID), Oracle Virtual Directory (OVD) etc. Along with couple of other components, following are the major software releases (as part of second release) of the new Oracle Identity Management 11g Product Suite:
  • Oracle Identity Manager
  • Oracle Access Manager
  • Oracle Identity Analytics
  • … and few others …
In Identity Management, Oracle Identity Management 11g product suite provides Identity and Access Management (IAM) functions along with compliance/security related solutions. In Oracle Identity Management 11g, as usual, more features are added such as security development platform, integration with Fusion Middleware.
In this blog series, I am going to talk more about the Oracle Identity Manager (OIM) Product. Let us first understand about the Oracle Identity Manager Product and its features, and then we will talk more about various options available for integrating Peoplesoft Systems with Oracle Identity Manager Product. I used my personal experience with the product and referred the Oracle Identity Manager 11g Release 1 documentation for these. These are various guides available as part of Oracle Fusion Middleware Documentation. If you need in-depth knowledge about this product, you need to refer these manuals. Let’s understand OIM product first.

About Oracle Identity Manager

One of the Major and important Oracle Identity Management Component is Oracle Identity Manager (OIM). Earlier this product was called Xellerate Provisioning (by a company called Thor technologies). OIM product provides a central repository to store user and group information for any organization. One of the important features of OIM is it can integrate with various target systems available (such as Peoplesoft HRMS, SAP, Active Directory, Siebel etc). Also, various other Oracle products such as JD Edwards, EBS and Oracle Retail  have connectors as well.
I like the OIM Connectors Page at the Oracle Website. You should visit once. There are connectors for most commonly used products in the market (such as Sun Java Directory, Novell eDirectory, SAP products, Databases, Siebel etc). In this post, I want to explore the Peoplesoft Connectors and how can we deploy these connectors in an enterprise implementing OIM. I am going to provide a conceptual understanding only, for more details on the Connectors; you should refer the connector documentation (Search for “oracle identity manager connector documentation” to visit the Connector Documentation page). Also, other products (that has no connectors) can be integrated with OIM using Generic Technology Connectors (GTC) which is delivered as part of OIM product. We will talk more about GTC in later posts.

Integrating Peoplesoft HRMS system with OIM

Peoplesoft HRM (or HRMS) Systems are ERP systems deployed in many enterprises across the world. Hexaware supports many such Peoplesoft HRMS systems implementation and support across the globe. There are two Peoplesoft connectors available for OIM product.  They are:
  • PSFT Employee Reconciliation Connector
  • PSFT User Management Connector
These two connectors are used for different purposes in a Peoplesoft based environment. Let’s explore the use of these connectors using an Architecture diagram. I created the following diagram to show the integration and the use of PSFT connectors.
In this High-level Architecture, I used an existing Peoplesoft HRMS System as a trusted source for OIM. OIM will play a role of central repository to store user and group information. The User Provisioning will be happening to multiple target systems mentioned in the diagram.
PSFT Employee Reconciliation Connector is used to perform trusted source reconciliation with Peoplesoft HRMS system. In this scenario, Peoplesoft HRMS system is the source for all the user or employee related information during the entire user management lifecycle (user add, user delete, user modification etc). There are two versions of the PSFT Employee Recon Connector.
  • Version 9.0.4.x
  • Version 9.1.x
If you are in Peopletools 8.48 or earlier releases, then you should opt for 9.0.4. For detailed list of supported releases, you can refer the connector documentation.
Both Version 9.0.4.x and Version 9.1.x use Integration Broker Architecture for integrating with OIM. As you are aware, the IB architecture is considerably changed starting with Peopletools 8.48. There are new features added in Peopletools 8.49. For Version 9.1.x, the Supported Peoplesoft HRMS systems are 8.9, 9.0 and 9.1 with Peopletools 8.49 and 8.50.
Let’s explore these two Peoplesoft Connectors for OIM in future posts. I really like to share and learn more about these connectors, mainly for two reasons. I worked as Peoplesoft Admin for so many years and I also learned some basics of OIM recently. Let’s meet in next post. Until then

Tuesday, 6 July 2010

Oracle Internet Directory LDAP Relpica States in Fusion Middleware 11g


Oracle Internet Directory LDAP Relpica States in Fusion Middleware 11g (11.1.1)

In the Oracle Fusion Middleware 11g Documentation (I think I was referring to Version 11.1.1 of the doco), you can find OID Administrator’s Guide. As the name suggests, this is the top most important and valuable guide for Oracle Internet Directory Administrators. I think I have read most of this guide already. However I still refer this guide, since there is a lot of information provided in this guide (and it is a reference guide too).
I want to write about the LDAP Replica states mentioned in the Appendix D (How Replication Works) of this Guide. In Fusion middleware, Oracle provides lot of details about Oracle Internet Directory Replication. Earlier this information was scattered around the Oracle Support Website and was difficult to find. Now, I think Oracle collected most of this information in this guide.
If you are working or supporting or planning to implement an OID Replication High Availability environment, then you should be familiar with this section of the topic in the guide. This replica states information will be useful if you are running LDAP Based Replica (Just to refresh your memory, there are two types of Replication possible, ASR based and LDAP Based – ASR is based on Ddatabase Links, while LDAP based replication uses a LDAP Client process).

orclReplicaState Attribute

orclReplicaState attribute stores the Replication State for the LDAP Based Replication Replica. You can check the current Replica State of the OID using the ldapsearch command. (In a Live System that uses LDAP based replication, it will be set to the numeric value of 1 – which means it is in online state).
You need to run the following LDAPSEARCH and check the orclreplicastate attribute as shown below. Please make sure to replace values for the arguments specific to your site, I just gave an example.
ldapsearch -h localhost -p 389 -D cn=orcladmin -w password -b “orclreplicaid=local_replica_ID, cn=replication configuration” -s sub objectclass=*
You need to check the value of the orclreplicastate in the output. Alternatively, you can get the orclreplicastate attribute value directly as shown below example:
ldapsearch -h localhost -p 389 -D cn=orcladmin -w password -b “orclreplicaid=local_replica_ID, cn=replication configuration” -s sub objectclass=* orclreplicastate
The local_replica_ID is specific to your installation, normally it is machine_database. You can check the value using a ldapsearch query as shown below.
Ldapsearch Argument Description:
Argument Description
-h Hostname or IP Address of the LDAP Directory ServerI used localhost since I am running this command on the same server where OID is running.
-p Port Number for the LDAP Directory, default LDAP port is 389, LDAPS port is 636.If you use the port 636, then you should define the –U argument.
-D Bind DN – LDAP DN for connecting to LDAP Directory
-w Password for the Bind DN – It is site specific.
-b Base DN for the search – here it starts from the top.
-s base Search Scope is base (other values are sub and one)

orclReplicaState possible values in 11g

There are 9 LDAP Replica States mentioned in this guide (In 10g OID, there are only 7 LDAP Replica states, it looks like Oracle added two more LDAP Replica states in 11.1.1). As I mentioned earlier, in a normal production system which uses LDAP based replication, the orclreplicastate will be set to the value of 1 automatically during the start of the replication server first time.
Let’s list the LDAP replica states:
LDAP Replica State Description
0 Bootstrap - This is one of the important Value. You can setup a new LDAP based consumer replica using this value. Lets talk about it in next blog
1 Online – For regular replication processing.
2 Offline
3 Bootstrap in progress
4 Bootstrap in progress + cn=orclcontext completed
5 Bootstrap completed with failures
6 Database based
7 Sync Schema only (Not Data)
8 Bootstrap without schema sync (Only Data)
In a LDAP replication with high-availability environment, it is a must that you should understand these values and their significance. Let’s talk about these values and how we can exploit this attribute and their values in the coming blogs. Until then

Monday, 17 May 2010

Fusion Middleware: New features in Oracle Internet Directory


Going forward, I am planning to write more about Fusion Middleware 10g, Fusion Middleware 11g and Oracle Database 11g. These are the areas that I am developing more interest now-a-days. I am currently working on Fusion Middleware 10g. First of all, I am learning these new software. And when I write here I feel my knowledge level increases. The first Fusion Middleware component that I am going to write about is Oracle Directory Server and Oracle Internet Directory. Both are LDAP Directories from Oracle and part of Fusion Middleware (Why two LDAP Directories as part of Fusion Middleware? – Think about it).
I worked in multiple LDAP Directories during the last few years. A LDAP Directory is software that stores information or entries or data in a tree like format for easy access, it is based on a standard. As per my experience with LDAP Directories, these are the major LDAP directories:
  • Oracle Directory Server (earlier Sun Java/Iplanet Directory)
  • Novell’s eDirectory (earlier NDS)
  • Microsoft’s Active Directory (AD)
  • Oracle Internet Directory (OID)
  • openLDAP
Among these, I like Sun Java Directory (now, Oracle Directory Server) the most. It is because I worked on it first and it was from Sun, standards based, and works well in heterogeneous environments. There are other various reasons, but we will talk more about Oracle Internet Directory here.

What is Oracle Internet Directory?

Oracle Internet Directory is a LDAP Version 3 Compliant Directory Server from Oracle Corporation. Oracle Internet Directory (OID) is used in most of the Oracle Components (such as Oracle Single Sign On) and is one of the primary components delivered as part of the Fusion Middleware.
OID is used to integrate Oracle Middleware and applications and mainly used with Oracle Applications. Oracle Internet Directory stores its data in an Oracle Database. The directory store is an Oracle Database. Oracle Database is a required component to run Oracle Internet Directory. This is one of the major differences among the remaining four major LDAP directory servers.

New Features of OID in Fusion Middleware 11g

OID is delivered by Oracle for the use of Oracle Identity Management. This was part of the Oracle Application Server “Application Infrastructure” Component. So, Oracle Internet Directory is not a new component that is delivered as part of Fusion Middleware.  It was already there in Version 10g as well.
I am currently working on Oracle Internet Directory Version 10.1.4.2.0. Fusion middleware version of Oracle Internet Directory is called 11.1.1. There are few improvements between these two versions. I noticed that the improvements lie on these lines

1. Manageability Features

Oracle Directory Services Manager and integration with Weblogic Admin Server are the major changes in the OID Version 11.1.1.  Fusion Middleware is Weblogic-Centric. So it is time to learn Weblogic again. Oracle Process Manager and Notification Server (OPMN) is still used in Fusion Middleware for managing OID, as well as other components.
ODSM (Oracle Directory Services Manager) is replacing Oracle Directory Manager (oidadmin). ODSM is a new web-based management tool for managing Oracle Internet Directory in Fusion Middleware 11g.

2. Replication Features

One of the important features that you can setup is a multi-master replication using LDAP based replica model. In earlier versions, it was not possible. Earlier you need to use ASR based replication to setup a multi-master replication. Now it is possible to setup multi-master replication using LDAP based replication.

3. Instance Configuration

There are changes in configset information. Now every instance can have a separate rootDSE information. This was one of the major issue in earlier version. I need to explore this option more. I will write more about this later.
A last important note is: why Oracle is delivering two separate LDAP Directories now as part of Fusion Middleware 11g or as part of its Directory Services Offerings. Why Oracle supports Oracle Directory Server and Oracle Internet Directory? This is because; Oracle Applications are tightly integrated with Oracle Internet Directory. For Example, Oracle Single Sign On needs Oracle Internet Directory. This is one of the reasons Oracle is unable to move to Oracle Directory Server. Let’s hope this will soon change.
Let’s talk more about OID in coming weeks.  Until then

Read More about Fusion Middleware

Thursday, 29 April 2010

Moving to Oracle Server Technologies


Believe me; life is not easy when you are working with a Vendor Company, such as Hexaware Technologies that I work for (Hexaware is an Oracle Platinum Partner as well). I have to learn all the new things with the little time you get, sometimes you have to learn lot of things in less than few hours. For a person like me, this is exactly what I want and like to do. Learn new things all the time!!! That is my motto!

One thing I like the most here is, I have the freedom to move to other IT technologies that I have little or no experience with. However that was not easy for a person like me or anybody for that matter. You have to keep learning and understand new things that come up.

As you are already aware (or if you are reading my blogs for first time), I started my IT career as the Web Developer with Apache and Perl CGI development (really old technologies!!). After couple of years, I got bored with Web Development. Then, I moved into Unix System Administration, mainly worked on Solaris and HP-UX and related hardware and software. And again, I got bored with UNIX Administration and moved to Peoplesoft Infrastructure and Administrator positions.

I was a happy person (I am still happy!!) for almost 7 years working with Peoplesoft Infrastructure for many clients. Now, I got an opportunity to work in Oracle Server Technologies here, especially Oracle Database, Oracle Identity Management and Oracle Fusion Middleware technologies.

If you are in the IT industry, you have to know one thing for sure. Keep learning. We have to develop a mentality like kids have. They are always curious to learn new things and all the time. This is an important quality you have to develop if you want to excel in IT Technical career. You have to develop curiousness to learn new things (from internet, from other blogs, from collogues, from peers, from managers and almost everywhere!).

I started working in Oracle Server Technologies (Oracle Database, Oracle Application Server, Oracle Fusion middleware, Enterprise Manager etc ) less than a year ago. However, before starting, I had an fundamental understanding of what they are and why do we need them. You cannot build this in one day. You should be aware of other technologies. One major thing that helped me was, my UNIX skills. I am able to solve almost any problems if that runs on UNIX.

Two things you have to understand in UNIX World. Everything is handled as a file and everything runs in the server is a process. If you are able to make these two simple facts, then I am sure you will be able to fix any servers, anything that runs in UNIX/Linux.
Okay, I think we are going off topic. Other than books and internet, I use two simple ways of learning.

a) Blogging
b) Teaching/Mentoring

Both of these are not easy for me. I have to really develop mastery to some level before start teaching someone. Believe me, it is not easy to teach, especially in IT industry, it is difficult with all the new things popping up almost every second. That is why I wanted to start blogging more often and conduct more mentoring classes in Hexaware.

And now, within last one year, I have got quite an expertise on Oracle Server Technologies. During this time, my experience with UNIX, Web Development and Peoplesoft really helped a lot in understanding the architecture of the Oracle Server Technologies. I am still learning new things everyday (that is why I want to write here, at least I can use them later!).

I want to use this new blog site to start sharing knowledge, write about errors or failures and how we handle them (lessons learned) etc. I will start with a new topic here soon. Until then.