Interview Questions for Oracle Apps DBA’s (11i/R12) – Part 9
Few more Interview questions for Apps DBA’s
Q. I am applying database patch using opatch. I don’t have oraInventory, can I apply database patch ?
No, you can’t apply a database patch using opatch if Inventory is missing.
Register ORACLE_HOME with global Inventory (in case inventory is missing) using steps mentioned here “3.6.2 Creating the Central Inventory”
Q. How many concurrent mangers can be used ? what is the limit ?
There is no such documented limit on number of concurrent managers, as long as you have enough operating system resource (memory).
Q. Suppose my concurrent manager status is showing Inactive ? what is the reason ?
This could be because of multiple reasons, to name few
a) Processes value for Concurrent Manager is 0 (System Administrator > Concurrent > Manager > Define > search for Manager > Click on the Work Shifts button and check value of Processes)
b) Concurrent Manager was shut down by some one (Check using System Administrator > Concurrent > Manager > Administer)
c) There is no workshift defined against Concurrent Manager (or workshift removed from Manager)
.
Q. What is transaction manager ? what is the significance ?
Normal/Standard concurrent manager execute long running, data-intensive application program asynchronously, where as transaction manager support “synchronous” processing of particular requests.
For more info and significance of transaction manager Check Oracle Applications System Administrator’s Guide – Configuration (zipped) Here (Page 176)
Q. My cloning is successfully completed but my apps listener is not getting up ? What is the problem ?
First check log file of apps listener to find root cause at
1. $OAD_TOP/admin/log/$CONTEXT_NAME/adalnctl.txt
2. $ORACLE_HOME/network/admin/apps_$SID.log (on Middle Tier)
To name few
1. Failure to create proper apps listener file via adgentns.sh during Rapid Clone process
2. FNDSM (Service Manager issues)
Q. How frequently do you preform patching ? (For work ex < 1.5 yrs)
Patching is done mainly to fix existing problem (bug) or for new features (enhancements). Frequency of patching vary from company to company and implementation stage.
Q. How frequently you perform cloning ? (For work ex < 1.5 yrs)
This again vary from company to company. Frequency ranging from every few days or every few weeks to every few months.
Q. How much time it will take to complete cloning ?
Cloning time depends on
– Network speed to copy various tops (APPL_TOP, COMMON_TOP.., DATA_TOP)
– Customizations (printer setting and any post clone issue)
– Topology (Number of middle tier, is system integrated with OID & SSO, Database is RAC or not, DMZ setup)
Time for Rapid clone for single node
Approx. 30 Minutes for pre-clone
Approx. 1-2 hours for post clone on database tier
Approx. 1-2 hours for Middle tier
+
Time to copy files from source instance to target Instance
Q. What is the purpose of the appsutil folder in ORACLE_HOME in database tier ?
appsutil directory in database ORACLE_HOME is place holder for Autoconfig and Rapid Clone related files on database Node.
Q. After cloning my oraInventory got corrupted ? can I apply my Opatch?
No, you can’t apply database patch using opatch (Though you can apply apps patch using adpatch). Recreate inventory as mentioned above.
Q. what command is used to kill the process which is using largest memory ?
Vary as per operating system if you wish to kill o.s. process (kill -9 <pid>)
Q. I am applying patch can I open another session and run adadmin ? (Thanks to Amit Pancholia for pointing this out)
Yes & No (depends on stage of patching), you can run adadmin if FND_INSTALL_PROCESSES is NOT locked (in use) by adpatch session. If this table is in use by adpatch session then you can’t run adadmin at same time.
Q. I am applying patch can I open another session in another node and run adpatch ?
Well, question is tricky. If you are using shared appl_top (application tier) and using distributed patching (a.k.a. Distributed AD) feature then yes you can start patch from one node giving total number of workers and start X number of workers from node1 , then start another adpatch session from second node mentioning remaining workers to start from second node. (More on distributed patching in shared application tier environment coming soon …).
Q. I am not able to start concurrent manager ? What is the problem ?
This could again be for various reasons, to check root cause of problem check log file at
$APPLCSF/$APPLLOG/$SID_MMDD.mgr (where MM is month & DD is Date on which Concurrent Manager was started)