Question about sql

Hello. I have this task:

View department_id, job (job_id), and max salary for this job where deaprtment_id = 50. Can sort by salary Max you help me to do?

Hello

Try this query (but untested):

SELECT DISTINCT d.department_id, j.job_title,

Max (e.salary) OVER (PARTITION BY e.job_id) max_salary

J employment, employee e, Department d

WHERE d.department_id = 50

AND e.department_id = d.department_id

AND j.job_id = e.job_id

ORDER BY max_salary;

Tags: Database

Similar Questions

question about SQL and VMDK places

Hello world. Im kinda new here so please be gentle
I'm confused on the locations of the files for sql.
At this moment I have 1 with 3 vmdk files sql server executes a muscular 100 GB sql database.
My disks are as follows
1 OPERATING SYSTEM
2 DB
3 newspapers
They are all thick and they are all stored with the virtual machine in the virtual machine configuration.
It lies in a cluster HA/DRS and we use them both.
The question boils down to performance.
If my VM is stored on DATASTORE_A and the VM folder on this data store has all the VMDK files in it, it would be safe to say that they are all on the same disk as much as vsphere is concerned.
Now to get best performance, I want to separate my VMDK files to separate data (with faster operation) warehouses to avoid conflicts of drive on my current data store. What is the right way to go about this?
How separate you your VMDK to other data warehouses, and enjoy other performance data store? WE do not use DTS at all so im sure it although we use HA/DRS in our group... and yet, I don't think that it is important in this case... just more of an FYI.
What do you do to get your SQL servers to run in vmware?
Seems like a basic question.
The answer seems easy. Separate disks so that they do not fight.
'S virtually done it with VMDK but the data store underlying is 1 speed and they all exist on this subject, so... How to separate them at that level for best performance?
Also... once I keep them separated, how can I keep my sanity by knowing my VMDK are not with the virtual machine? This kind of made a mess . Of course the performance takes a front seat for my mental health.
Thank you all

1. If I sVmotion my other vmdk files via the advanced options of one in another data store, it will create the files folder structure based on my VM name and place the vmdk inside? (I guess I could just test it and see).

Yes, the VM configuration file (.) VMX) will be modified to point to the new location of the virtual disk. You can use the virtual machine property editor to check for this, once the Storage vMotion.

2. with regard to the disk controller, I assume you mean a virtual controller by VMDK in the vmx / correct the settings of the virtual machine and not a phyical on the hardware controller? (IM pretty sure I know the answer here but I just want to be clear).

Yes, the virtual controller is correct.

If so, what type of controller? LSI Logic or it would benefit all a paravirtual adapter?

As long as your operating system is supported, go with the PVSCSI every time - especially for newspapers and the controller file data.

General question about sql DB to vcenter

Basically, we need to know how many connections SQL DB to go from one instance of vCenter 5?
It is dependent on host?
It is dependent on comments?
Is there a way we can we estimate how many connections?
DBA SQL told us we have currently a large number of connections, but we do not think that this could be cauisng questions.

vCenter uses a connection pool (10 by default as he answered already). This is usually done by web application to manage db connections required by multiple login sessions (no need to reopen a db for each session connection... open a db connection sometimes takes more time that select it dumb that they do).

Question about SQL * connect more in the syntax
Good evening

I can connect successfully from the client to the host using the following command:
```
0 19:40 [Oracle] [23G] [569M] E:\Client>sqlplus scott/abc123@dbca

SQL*Plus: Release 11.2.0.1.0 Production on Sun Aug 22 20:01:56 2010
```
But I wanted to omit the password, so I tried:
```
0 20:03 [Oracle] [23G] [590M] E:\Client>sqlplus scott/@dbca
```
thinking that SQL * more tempt me the password but, if it does at all. I tried in vain a few variations. I read the specifications of command line for SQL * more and it seems that this command should work.

Questions: What is the problem with the command I typed and what is the correct syntax?

Thank you for your help,

John.
To omit the password using connect scott@dbca , but do not connect scott/@dbca of sqlplus.
or sqlplus scott@dbca

Question about SQL server 2008

I can't remove the Default constraint in Management Studio. But in the query, I can drop Default constraint

For assistance please ask for help in the appropriate Microsoft TechNet SQL Server Forums

Thank you & good luck.

A question about the order of COLUMN sqlplus and PL/SQL
Hello everyone.

I would like to ask a question about sqlplus command COLUMN that I use in order to define the length of my queries of output columns. So, for example, I run something like:
```
COLUMN market FORMAT a20
```
That is why any column with the name 'the market' will be composed by sqlplus as 20 characters.

Currently I'm working on a PL/SQL script that makes this automatic procedure, in other words, instead of manually write for each column in the above COLUMN order, it uses the DBMS_SQL. He read first as input the name of the table and then inside a loop, it detects the maximum length.

For those interested, here is the Oracle documentation page that I found (example 8)
http://download.Oracle.com/docs/CD/B19306_01/AppDev.102/b14258/d_sql.htm

If everything worked well enough I managed to recover the maximum of each column in my table length, except that I don't know how to use the COLUMN sqlplus command in my PL/SQL block. I would like to do something like that (it's just a pseudo-code)
```
FOR colIndex in 1 .. numberOfColumns LOOP
        currentColumnName = getColumnName(tableDescriptionObject, colIndex);
        currentColumnLength = getColumnLength(tableDescriptionObject, colIndex);
        
        COLUMN  currentColumnName FORMAT a -- and here after 'FORMAT a' I have to write currentColumnLenth but I don't know how to proceed.

END LOOP;
```
I would like to know is there a command in PL/SQL to define the appropriate column typeset in sqlplus length?

Thanks in advance,
Dariyoosh
Hi, Dariyoosh,

Sory, I don't think that there is a simple way to publish SQL * most orders fom PL/SQL.
PL/SQL runs in the database (back-end); There isn't any direct contact with SQL * more at the front end.

The best way I know is that PL/SQL script (myfile.sql) with orders of the COLUMN, and then have SQL * more run the script (@myfile).

Question about ticks in time

I have a question about the number of ticks.
CREATE function abc (netTicks in number)
RETURNS THE TIMESTAMP
as
DATE date1: = '' 2010-11-01;
v_ticks_for_2010_11_01 NUMBER (19.0): = 634241664000000000; -I did not understand how this number came
v_ticks_in_a_second NUMBER (19.0): = 10000000; -THERE ARE 10 MILLION OF TICKS IN ONE SECOND
Start
utils. DateAdd ('SECOND', ((netTicks-v_ticks_for_2010_11_01) / v_ticks_in_a_second), date1); - what do what is this function?
end
/
Thank you

Hello

NewLearner wrote:

Hi, I have a data type of bigint type in SQL server. During the migration I changed it to number (19.0) in oracle. But how to have a variable equal to bigint in oracle that is coming as an input parameter. I can not give number (19.0) for the input parameter. His gift of syntax error. How to make this support in oracle?

In the function declaration, you give just the type of data (e.g., NUMBER). Sizes, as NUMBER (19, 0) , are given to variables that pass to the function. For example:

DECLARE

x_ticks NUMBER (19, 0): =...;

x_ts TIMESTAMP

BEGIN

x_tx: = abc (x_ticks);

...

When writing to abc, suffice to say that it will be passed a NUMBER. When you call with a certain NUMBER (19, 0), then it will use this NUMEBR (19, 0) value.

Also in the Sub function return to_timestamp (' 2014-01-01') + interval '10' sec; instead of having fixed the value of 10 for the interval. I want to spend a dynamic value. example: something like v_interval: = (a - b) / c; return to_timestamp ('2010-11-01') + v_interval second interval; Help, please

TO_TIMESTAMP call with 1 single argument is just begging for trouble. Always use a 2nd argument, for example

TO_TIMESTAMP ('2014-01-01', "YYYY-MM-DD")

Rather than use a literal INTERVAL here, you could use a variable INTERVAL, use NUMTODSINTERVAL to convert a variable NUMBER of INTERVAL.

A question about the insertion of the localTimeStamp returned the value into a timestamp with time zone column

Hello
Oracle version: Enterprise Edition Release 12.1.0.1.0 - 64 bit
OS: CentOS 4.6 X86_64
I have a question about the localTimeStamp function. According to the documentation for this function is
the same thing as current_TimeStamp except that the returned value does not include the time zone.
Consider the following example:
SQL > create table tmptab (colval timestamp with time zone not null);
Table created.
SQL > insert into tmptab (colval) values (localTimeStamp);
1 line of creation.
SQL > select t1.colval, tmptab from t1;
COLVAL
---------------------------------------------------------------------------
24 JANUARY 14 09.45.42.253732 H + 01:00
SQL >
Why introduce the foregoing did not fail? the data type of the column in my table colval expects a timestamp with time zone
during each which inserts as I understand (correct me if I'm wrong) is not provided by the localTimeStamp function.
Could someone kindly tell me what I misunderstood?
Thanks in advance,

dariyoosh wrote:

Hello

Oracle version: Enterprise Edition Release 12.1.0.1.0 - 64 bit

OS: CentOS 4.6 X86_64

I have a question about the localTimeStamp function. According to the documentation for this function is
the same thing as current_TimeStamp except that the returned value does not include the time zone.

Consider the following example:

SQL > create table tmptab (colval timestamp with time zone not null);

Table created.

SQL > insert into tmptab (colval) values (localTimeStamp);

1 line of creation.

SQL > select t1.colval, tmptab from t1;

COLVAL

---------------------------------------------------------------------------

24 JANUARY 14 09.45.42.253732 H + 01:00

SQL >

Why introduce the foregoing did not fail? the data type of the column in my table colval expects a timestamp with time zone

during each which inserts as I understand (correct me if I'm wrong) is not provided by the localTimeStamp function.

Could someone kindly tell me what I misunderstood?

Thanks in advance,

Then, when you do

create table t (neck DATE);

Insert into t (to_date ('1st January 2014', ' mon-dd-yyyy "");)

Are you surprised that it works? After all a DATE column contains a time component, you do not provide: but he

works and gives you 00:00:00 as the component "hour".

Similarly, the timezone component is being developed to automatically with, I think, time zone of your system.

Question about business rules

Hello
I have a question about business rules. We have a lot of business rules for our planning application, and I wanted to know if there is a way to export all sort of business rule names in an Excel type file?
Please, share your ideas.
Thank you
~ Hervé

You can run a query against the relational tables planning and get the names of BR from there.

SELECT b.object_name

Hsp_calc_mgr_rules has

hsp_object b

WHERE a.id = b.object_id;

Cameron a quite handy planning SQL queries here http://camerons-blog-for-essbase-hackers.blogspot.com/search/label/stupid%20Planning%20queries

I have a few http://www.orahyplabs.com/search/label/sql here (all plan not specific but)

Concerning

Celvin

http://www.orahyplabs.com

A question about external authentication with PHP OCI8 using a portfolio store

Hello

SQL> SELECT * FROM v$version;

BANNER
--------------------------------------------------------------------------------
Oracle Database 11g Enterprise Edition Release 11.2.0.1.0 - 64bit Production
PL/SQL Release 11.2.0.1.0 - Production
CORE     11.2.0.1.0     Production
TNS for Linux: Version 11.2.0.1.0 - Production
NLSRTL Version 11.2.0.1.0 - Production

SQL>

OS: Linux Fedora Core 17 (x86_64)
PHP Version: 5.4.14
Apache version (httpd): Apache/2.2.23 (Fedora)

I have a question about OCI8 connection to the Oracle server via portfolio, which I understand is a good practice because we do not have to hard code the users password. I'm trying to read pages 117 and 118 of the book The Underground PHP and Oracle manual online to learn how to set up a portfolio. I proceeded by steps as explained in these two pages (except that at the end I connect, as a normal user, not a privileged user) yet, apparently PHP does not detect the user.

Here's what I write in my script

$connection = 
                oci_connect 
                    (
                        "training", 
                        "", 
                        "php_ora_usr", 
                        "AL32UTF8"
                    );

the training is the name of the user that I use in my PHP scripts to connect to the oracle database and the password is mypassword (is not true, just for the purpose of this thread :)) the code above gives me the following error:)

Warning: oci_connect(): ORA-01005: null password given; logon denied in /var/www/html/myscript.php on line 91

Fatal error: Connection step: ORA-01005: null password given; logon denied in /var/www/html/myscript.php on line 13

But if I also provide the password which is,

$connection = 
                oci_connect 
                    (
                        "training", 
                        "mypassword", 
                        "php_ora_usr", 
                        "AL32UTF8"
                    );

Obviously it works, but at the same time, this means that my wallet is not operational.

So I'll write here, step by step, how I did according to the instructions contained in the pages 117 and 118 of the above mentioned book and I would be grateful if you could kindly tell where I made the mistake (s)

The following text was performed under my oracle linux (so the dbagroup)

Step 01:_
I created and provided the password for the directory of portfolio

mkstore -wrl /home/oracle/wallet_dir -create

Step 02:_
I created an entry for my oracle user that will be used in my PHP scripts to connect to oracle
user: training
password: mypassword

mkstore -wrl "/home/oracle/wallet_dir" -createCredential php_ora_usr training mypassword

I also checked that the entry was actually created for my user in the portfolio

$ mkstore -wrl "/home/oracle/wallet_dir" -listCredential
Oracle Secret Store Tool : Version 11.2.0.1.0 - Production
Copyright (c) 2004, 2009, Oracle and/or its affiliates. All rights reserved.

Enter wallet password:                
   
List credential (index: connect_string username)
1: php_ora_usr training
$

Step 03:_
I added the following to my $TNS_ADMIN/tnsnames.ora

.  .  .
php_ora_usr = 
  (DESCRIPTION =
    (ADDRESS = (PROTOCOL = TCP)(HOST = localhost)(PORT = 1521))
    (CONNECT_DATA =
      (SERVER = DEDICATED)
      (SERVICE_NAME = db02)
    )
  )
.  .  .

Step 04_
I added the following to my $TNS_ADMIN/sqlnet.ora

# sqlnet.ora Network Configuration File: /u01/app/oracle/product/11.2.0/db_1/network/admin/sqlnet.ora
# Generated by Oracle configuration tools.

NAMES.DIRECTORY_PATH= (TNSNAMES, EZCONNECT)

ADR_BASE = /u01/app/oracle

WALLET_LOCATION =
    (SOURCE =
        (METHOD = FILE)
        (METHOD_DATA =
            (DIRECTORY = /home/oracle/wallet_dir)
        )
    )
SQLNET.WALLET_OVERRIDE = TRUE
SSL_CLIENT_AUTHENTICATION = FALSE
SSL_VERSION = 0

_ Step 05
The TNS_ADMIN environment variable is already set correctly in * / etc/sysconfig/httpd * so I didn't need to do something for this step

Step 06_
I gave to access Apache (as root) in the directory of portfolio (in fact, it seems that the user oracle (dba) of linux also has the required privileges to grant this access to the apache)

# setfacl -m u:apache:rx /home/oracle/wallet_dir
# setfacl -m u:apache:r /home/oracle/wallet_dir/{cwallet.sso,ewallet.p12}

Step 07_
Restart Apache

# service httpd restart

Step 08_
Finally make a test connection using the connection string php_ora_usr which the entry was created earlier. And the test was not successful, as indicated above

$connection = 
                oci_connect 
                    (
                        "training", 
                        "", 
                        "php_ora_usr", 
                        "AL32UTF8"
                    );

The error message:

Warning: oci_connect(): ORA-01005: null password given; logon denied in /var/www/html/myscript.php on line 91

Fatal error: Connection step: ORA-01005: null password given; logon denied in /var/www/html/myscript.php on line 13

I would appreciate it if you could kindly shed some light.

Thanks in advance,
Dariyoosh

That gives a good clue to begin with. With your platform/Apache/DB, you might need different ACLs. Test directory and permissions. Check what euid your Apache runs as. Good luck.

EBS 12.1.3: a few questions about cloning
Hello

We have EBS 12.1.3 on AIX 7.1.

A few questions about the cloning of level shared environment of application layer non-shared application environment. We followed 'Cloning Oracle Applications V12 with rapid Clone [ID 406982.1]'

Environment (shared application layer) source:
-Two nodes in the active application and LBS
-A database of active and a passive database configuration with Oracle Data Guard.
-Terminate SSL at the load balancer
-Updated CFP implementation

Environment target (level app not shared):
-A single node of the application active
-One data node active
-Without SSL
-No PCP

After the cloning was completed, we noticed the following problems in the target system:

(1) home page is not available because application $CONTEXT_FILE layer always contains the endpoint "https" and ssl settings.

Workaround solution:

Change (with text editor) layer Application $CONTEXT_FILE as follows.
< webentryurlprotocol oa_var = "s_webentryurlprotocol" custom = "yes" > http < / webentryurlprotocol >
< sslterminator oa_var = "s_enable_sslterminator" > # < / sslterminator >
< login_page = "s_login_page" custom = 'yes' oa_var > http://myhost.mydomain.com:8000/OA_HTML/AppsLogin < / login_page >
< externURL oa_var = "s_external_url" custom = "yes" > http://myhost.mydomain.com:8000 < / externURL >

Run the automatic configuration service and application layer.

Is there a better solution and supported for this?

(2) simultaneous managers do not work properly, because they still have old names of node (used in the CFP in the source system). The solution is used:

1. connect to the EBS GUI web and delete all nodes of all concurrent managers.
2 shutdown application layer. Check that all concurrent managers have been extinguished. If not, kill the process manually
3. as a user apps run cmclean.sql
4 restart the application layer

Is there a better solution for this?

(3) database ARCHIVELOG mode nonarchivelog mode changes in the target system. Is anyway to prevent this?

(4) in addition, we would like to make a full clone (app + db levels) of production without closing the application. Is this possible? What would be the best documents to describe it? In Production, we have Oracle Data Guard physical standby installation as per "Note technique Oracle: Business Continuity for Oracle E-Business release 12 Using Oracle 11g Physical Standby Database [1070033.1 ID]. This can be used when executing a clone of warm for the database, right? What application layer? The steps in 406982.1 include AutoConfig in what application level should be running down. Must the application be broken when running "to maintain the Snapshot information?"

(5) one last question (not related to cloning). Oracle Enterprise Manager console webUI has recently stopped working in IE8. For example, https://myhost.mydomain.com:5501 / em gives "cannot display the Page. It was working properly last week. In Firefox, the same link works fine. Any ideas how to solve this problem?

Thanks for your replies in advance.

BR,
TH
(1) home page is not available because application $CONTEXT_FILE layer always contains the endpoint "https" and ssl settings.

Workaround solution:

Change (with text editor) layer Application $CONTEXT_FILE as follows.
http
#
http://myhost.mydomain.com:8000/OA_HTML/AppsLogin
http://myhost.mydomain.com:8000

Run the automatic configuration service and application layer.

Is there a better solution and supported for this?

No, you must modify the context manually and run the automatic configuration.

(2) simultaneous managers do not work properly, because they still have old names of node (used in the CFP in the source system). The solution is used:

1. connect to the EBS GUI web and delete all nodes of all concurrent managers.
2 shutdown application layer. Check that all concurrent managers have been extinguished. If not, kill the process manually
3. as a user apps run cmclean.sql
4 restart the application layer

Is there a better solution for this?

You follow the best approach. You could update the details of the server node if you wish.

(3) database ARCHIVELOG mode nonarchivelog mode changes in the target system. Is anyway to prevent this?

By default, the quick Clone will create the target in noarchivelog mode database.

(4) in addition, we would like to make a full clone (app + db levels) of production without closing the application. Is this possible? What would be the best documents to describe it? In Production, we have Oracle Data Guard physical standby installation as per "Note technique Oracle: Business Continuity for Oracle E-Business release 12 Using Oracle 11g Physical Standby Database [1070033.1 ID]. This can be used when executing a clone of warm for the database, right?

Fix.

What application layer? The steps in 406982.1 include AutoConfig in what application level should be running down. Must the application be broken when running "to maintain the Snapshot information?"

You can copy the files from the application tier node, while demand is rising. And, there is no need to stop the application to run "to maintain the Snapshot information."

(5) one last question (not related to cloning). Oracle Enterprise Manager console webUI has recently stopped working in IE8. For example, https://myhost.mydomain.com:5501 / em gives "cannot display the Page. It was working properly last week. In Firefox, the same link works fine. Any ideas how to solve this problem?

If it works from a browser, then it should not be a problem of EM. Have you tried another client and see if it works? Have you changed a setting in your Internet Explorer browser? Please make sure you add the URL of the EM to the list of trusted sites.

Thank you
Hussein

Question about snapshots and virtual machines

Hi all
I have a small question about snapshots. I use VMware for about 6 months and I read about snapshots and how they work but still have a question to make sure that I understand how it works and it will work the way I think it works.
I have a Windows 2003 server, which is a virtual machine. I have an application that uses MSDE database. I want to upgrade the database to MS SQL 2005, but to do this I need to back up the MSDE database, uninstall MSDE, and then install SQL 2005 and move the SQL database. This happens on the same VM.
What I was thinking if I well Snapshots correctly is to snap turned the virtual machine before doing anything. Then do my upgrade steps and see if everything works. If its all messed up so I should be able to return to my shot and everything should be the way he had. What is the good? One of the main reasons I moved the server VM was because of this upgrade.
After reading the documentation for the snapshots, it's how I undestood it. Pretty much any change could be reversed with snapshots. I wish just they had put in a few examples in the PDF file. Also it would be the same for all type of grades of the OS? Could I snapshot a Windows 2000 Server, and then "in the upgrade of the square" to Windows 2003? I've cloned a VM to do this but was curious about whether snapshots would be just as easy.
Thanks for your help.

Hello.

What I was thinking if I well Snapshots correctly is to snap turned the virtual machine before doing anything. Then do my upgrade steps and see if everything works. If its all messed up so I should be able to return to my shot and everything should be the way he had. What is the good?

That is right.

After reading the documentation for the snapshots, it's how I undestood it. Pretty much any change could be reversed with snapshots. I wish just they had put in a few examples in the PDF file. Also it would be the same for all type of grades of the OS? Could I snapshot a Windows 2000 Server, and then "in the upgrade of the square" to Windows 2003? I've cloned a VM to do this but was curious about whether snapshots would be just as easy.

Exactly the same thing for the upgrades of the OS.

Perhaps the most important thing with the snapshots is to keep an eye on them and don't let them sit too long or become too big. You have described the use cases are very good uses for snapshots. Just make sure you have a solid plan with a schedule set, so that you can come back or validate changes before the snapshot becomes too large.

Good luck!

Question about the database to use with ESX and Labmanager

Hello
Finally my company is buying a permit for Labmanager and ESX server using Labmanager to our tests.
We had a question about the database that we use for the actual installation.
Could we use MY - SQL? Or that we have the olbigation to install a SQL or Oracle DB for the whole system to work?
We prefer to use MY_SQL because it is open-source and will cost less to use for us.
Thanks in advance for the answer and please excuse my bad English, I'm french spoke first.
Lafa91
Montreal.

Lab Manager installs SQL Express as part of the installation and use. If you install also Virtual Center as part of your deployment of Lab Manager (do not use an existing VC server), you can use the database SQL Express is included for small installations of ESX, but MY SQL is not an option. You can search the databases supported in the installation guide for what version you deploy.

Question about *.aud in audit_file_dest fields

A node 3 RAC 11 g 2 on Linux 5.2, we recently had a full disk because of too many files in audit_file_dest. The parameter audit database is such

SQL> show parameter audit

NAME                                 TYPE        VALUE
------------------------------------ ----------- ------------------------------
audit_file_dest                      string      /u01/app/oracle/admin/prdcch/adump
audit_sys_operations                 boolean     FALSE
audit_syslog_level                   string
audit_trail                          string      DB

With audit_sys_operations = FALSE, I don't expect operating SYS written to audit_file_dest. But in fact we still see a lot. Here are my questions about sys audit
(1) how really turn off sys audit
(2) I see much usefule sys files in audit_file_dest of the audit. A typical such file contained as (removded header)

Sat Jul  2 19:15:22 2011 -05:00
LENGTH : '155'
ACTION :[7] 'CONNECT'
DATABASE USER:[1] '/'
PRIVILEGE :[6] 'SYSDBA'
CLIENT USER:[6] 'oracle'
CLIENT TERMINAL:[0] ''
STATUS:[1] '0'
DBID:[10] '2362038599'

From what I understand, it tells me the time and duration of the session SYS. It don't me didn't say where machine (which can help to identify who connected in a SYS), not the activities

 I find this info is useless. I must missed something important, hope some one can shine some light here.
3) If cannot stop the SYS audit, I would like to redrect the audit_file_dest to a differnt drive with larger space. Is there anything thing need special attention when doing it in RAC. Can each instance has differnt audit_file_dest?

Thanks

Although audit_sys_operations is false, some actions are audited by default in 11g. See also DBA_AUDIT_TRAIL.
(1) you can't turn it off
(2) TERMINAL CLIENT is empty, it means that the connection of the server itself. In your example someone logged on locally as "/ as sysdba".
(3) audit_file_dest is instance specific. You can put each in their own directory or put them all in a share. If the files are very small, according to the activity of DB there could be millions of them. In my environment remove us periodically as they age (more than 3 months).

Question about the Oracle database resource management
Good evening
I have a question about Oracle database Resource Management (DRM) for Oracle Database 10 g. I had an incident which, in a development/UAT environment (different instances, same box, don't ask), a developer wrote a wrong SQL statement that consumed 99% of the CPU. Currently, there is no limit of resources on two cases of db, and the developers don't want limits. However, my manager wants a method to prevent something like this does not happen. If I use the Oracle database resources management, create a consumer group (called "DevEnv") at 40% CPU usage, what will happen if a user of DevEnv runs a bad SQL statement that normally consumes 90% CPU again? Anyone would take a different approach? Thank you.

Sincerely,
Sho Fukamachi
782718 wrote:
If users in the case of "db_uat" use 60% CPU would be users who belongs to DevEnv pending 'db_test' cannot be used more than 40% CPU, so their queries become slower at their end? I would like to know if this behavior works the same in an environment of multiple instances. Thank you.

Yes, because the available processor would be 40%, they use more, while.

Question about sql

Similar Questions

Maybe you are looking for