Oracle 9i DB taking forever to start up

Hey all

Just a quick Question. I have cloned databases in the past with no problem at all. But I have cloned a DB in this instance and a problem as arisen.

When teh Machine starts up I enter command prompt and enter the follwoing

set oracle_sid=rrprod

sqlplus /nlog

connect / as sysdba

it then says connected to and idle instance. It is supposed to startup automatically.

so i then run

startup

and from here it takes about 5 min to come up with this

ORACLE instance started.

Total System Global Area 101785428 bytes
Fixed Size 454484 bytes
Variable Size 75497472 bytes
Database Buffers 25165824 bytes
Redo Buffers 667648 bytes
Database mounted.

and then another 8/9 minutes to actually open the database.
Its not a huge problem as i very rarely shut this machine down and once the database is open it runs fine.

Anyone know what could be the prolem here? If you have a solution or a suggection could you please explain to me exactly what to do as Im and Oracle Newbie!

Cheers

Graham

Hi

Did you notice anything unusual in the alert log when you start the instance .

Can you post the alert log here as a attachement

regards
Hrishy

Hey

Thank you for your reply!

Fogive me, but as I said I am a total oracle newbie. how do access the alert log?

are we talking about the event log that can be accessed through windows in admin tools or something else

Cheers

Graham

Hi
Your are working in which platform linux or windows.
If u dont know the path of the alert log file.
the alert lof file is in the bdump directory.
to get the path of the bdump directory path.
issue the command in the sql prompt
SQL>show parameter background_
then inside the bdump directory you can see the alert log sid_alert.log
gopu

Hey

I have managed to fix the problem due to your help.

when I cloned the Database I did not copy the bdump folder over in the admin folder.

with the result that there was no log file. I have created the bdump folder and the log file and shut down and started up the database again to fill the log file with data for you.

but because the log file was there this time the Database started up properly

thank a million buddy i really appriciate this :)

Graham

I'm sorry but I don't think that was the case.

If background_dump_dest is not available Oracle would try to write the *sid*alert.log file into $ORACLE_HOME/rdbms/log

My guess is database started faster this time just because of it shutdown nice and clean prompting a clean startup.

On the other hand, you did the right thing by creating the bdump folder.

Quote:

---

Originally Posted by PAVB

I'm sorry but I don't think that was the case.

If background_dump_dest is not available Oracle would try to write the *sid*alert.log file into $ORACLE_HOME/rdbms/log

My guess is database started faster this time just because of it shutdown nice and clean prompting a clean startup.

On the other hand, you did the right thing by creating the bdump folder.

---

This is what is suspected and was interested in looking at the alert log for possible rollback that was happening while the database was being opened.

regards
Hrishy

FYI

there is no log file or folder under D:\oracle\ora92\rdbms on my machine
where else would the log file be stored?
I want to find a copy of the log file to show it to you guys!

AND

in the past shutting down the DB manually (clean) and then starting it back up again did not fix the problem. it still took ages to start back up.

Graham

please open sqlplus session and do...
show parameter background_dump_dest

What's the location Oracle points you?

Please check first and last entries on alert.log and correlate with the timing of your startup/shutdown activity.

Quote:

---

My guess is database started faster this time just because of it shutdown nice and clean prompting a clean startup.

---

Why would an unclean shutdown make the instance to take about 5 minutes to start?I think the problem is rooted deeper than it looks.

Quote:

---

Originally Posted by simply_dba

Why would an unclean shutdown make the instance to take about 5 minutes to start?I think the problem is rooted deeper than it looks.

---

why? WHY? how about database recovery? :rolleyes:

Quote:

---

Originally Posted by PAVB

why? WHY? how about database recovery? :rolleyes:

---

My dear PAVB,

database recovery would start after database is mounted ,not when the instance is starting.As the user says it takes abt 5 minutes to start the instance and few more to mount and open. So 5 minutes to allocate memory and start the background processes, that is what i am worrying about

Quote:

---

Originally Posted by simply_dba

As the user says it takes abt 5 minutes to start the instance and few more to mount and open. So 5 minutes to allocate memory and start the background processes, that is what i am worrying about

---

I get it now, you are troubled about the 5 minutes nomount stage and just don't care about the 8 to 9 minutes it takes to mount and open.

Happy you!

Quote:

---

Originally Posted by PAVB

I get it now, you are troubled about the 5 minutes nomount stage and just don't care about the 8 to 9 minutes it takes to mount and open.

Happy you!

---

I do care, But I do not wrongfully assume database recovery taking place , as you have, without having a look at the alert log.

Hi,

If there is a problem of having badsector in disk,one may suffer memory allocation -and also problem in DBopen too. This may be his problem.
Since i had suffered the same prob when i had lot of badsector.