Some dba and OS related confusions

Question

Hi All,
I have few doubts related to database and OS perspective. Could pelase someone let remove my these doubts and clarify below questions:
Qus1: Can we have two listener.ora file with same default port number 1521?
Qus2: Why do we create two groups "dba" and "oinstall" while installing oracle database?
Qus3: If my database is in archive log mode and in case archiver process is killed then redo will be archived or not otherwise what will happen?
Qus4: If I have catalog database for RMAN metadata. And in case catalog database is crashed. Will I be able to use RMAN backupsets for restore the database?
Qus5: What is needed to run "some_path/root.sh" script while installing the oracle database at last step?
Qus6: How can we get to know what are the database running in my server from OS level itself means if any database is shutdown or crashed, can we get to know about those database from OS level itself? If yes then please specify the command?
Qus7: How can we configure UDP protocol for my listener?
Qus: If I am updating one table and update query is going on. Inbetween OS gets shutdown then when it is back my transaction will rolledback. Please let me know how it happenes means is there any process to do this or something else?
Please clarify above questions.
Regards,
Michel

Answer 1

Q1: Can we have two listener.ora file with same default port number 1521?
A1. If on the same machine then only if they are listening on different virtual IPs, otherwise not on the same machine. As previously stated, you can try this out for yourself.
Q2: Why do we create two groups "dba" and "oinstall" while installing oracle database?
A2. For role separation such that if you multiple software installations and/or databases on the same machine, all will have access to the central inventory owned by the primary "oinstall" group, while the other installations/databases can be owned by other secondary groups (i.e. "dba", "dba1", and so on). Taken a step further you can have different oracle software owners with the same primary "oinstall" group but different secondary groups.
Q3: If my database is in archive log mode and in case archiver process is killed then redo will be archived or not otherwise what will happen?
A3. I like the idea of testing this out for yourself so I'll leave this to you ;-)
Q4: If I have catalog database for RMAN metadata. And in case catalog database is crashed. Will I be able to use RMAN backupsets for restore the database?
A4. This depends on a few things but in the simplest case, assuming you have the control file, only need to recover a datafile, and the backup data is still available (as specified by the CONTROL_FILE_RECORD_KEEP_TIME parameter), then yes.
Q5: What is needed to run "some_path/root.sh" script while installing the oracle database at last step?
A5. You must have access to the root account (directly or indirectly via sudo as an example) at the command line.
Q6: How can we get to know what are the database running in my server from OS level itself means if any database is shutdown or crashed, can we get to know about those database from OS level itself? If yes then please specify the command?
A6. There are two different questions here, namely check for running databases, and check database status. To check for running databases from the OS there are some options, the most basic is likely the command:
ps -eaf | grep ora_pmon | grep -v grep
The above checks for running processes, pipes that output through a 'grep' filter for commands having 'ora_pmon' (Oracle PMON process), which in turn is filtering out the command used to do the check itself (i.e. having 'grep' in the command).
To check the status you'll need to scrap the alert log file for certain strings. You could also use 'srvctl status database -d <dbname>' if that is available, but it will only give online/offline status.
Q7: How can we configure UDP protocol for my listener?
A7. Don't worry about this, it's taken care of auto-magically. Read the Oracle networking documentation for information.
Q8: If I am updating one table and update query is going on, in between OS gets shutdown then when it is back my transaction will be rolled back. Please let me know how it happens meaning is there any process to do this or something else?
A9. There's nothing that you need to do, it's handled internally by the database. I highly recommend reading through the concepts documentation for full understanding of this, as well as how things generally work inside the Oracle database.
Hope this helps.