In a server how do we find Production, test and development databases. is there any command.
Please tell me.
Thank you!
In a server how do we find Production, test and development databases. is there any command.
No, you need to know or need to know is TEST/DEV/PROD.
In most cases if the database is in Archivelog mode, then you can assume it as production. Check if all RMAN tasks configured, it could PROD. Same job RMAN will be scheduled on DEV/TEST according to criticality.
Can someone give me an idea of database load tests. Suppose that testers are performing on any application load test, and we check the charge (bottlenecks) side database, so what are the areas that we need to observe on the database side.
Version of Oracle - 10.2.0.3
Thanks in advance
Check if the server is ok with memmory and CPU usage and swap...
SIDE database, check the active sessions long running where they expect and suggest something so that these applications can accelerate. You can propose the amendment of the code as the new index, advice or contours etc... After checking waiting for them and various slow sql statemenets sql plans
Pay attention to the TYPE of repository, that you want to create.
I recommend you to perform the following steps for a basic deployment environment. As you improve with ODI try different configurations.
(1) a single master repository work for your DEV environment.
(2) use the same Master repository and create a kind of execution referential work that you will use this as your Test environment.
(3) create a new repository repositories master and execution in another server and configuration in your PROD environment. It's your safe and you deploy only when everything is working fine. In addition, you will have a regular backup to repository strategy.
Now just set three new connections in ODI Studio.
A DEV using information from repository of 1.
A TST uses 2 repository information.
One for PROD uses 3 repository information.
You're good to go. Develop you project in DEV, generate deployment scenarios to test in TCT and if all goes well. Deploy the topology and scenarios needed to PROD.
I have a planning application in Dev ('X') which is now obsolete and is not synchronized to the top with the planning application in Prod ('Y'). X and that are the same applications that the names are different in the environments.
I want to know what is the best way to synchronize to the top of the 'X' planning application which is in dev with a planning application 'Y' which is in production to ensure that sound closer to production now? What are the things that I have to keep in mind while doing this?
If I had to use LCM for it is necessary that both applications should have the same name of application and database?
It would be really helpful if someone could suggest the best and effective to do and throw some light on this topic.
Thank you -Krrish
If hypadmin is a native user and manually created it means that it has a different SID between environments, this value is stored in the HSP_USERS table, if it is the default Administrator account value does not, in any case.
Go to the base tables of data for the application in environment planning target and look at the HSP_USERS table, store the value of the SID 50001 user, if the user who created the app is the owner of the application, then the role should be 3
Stop planning.
Copy the relational database from source to target, completely replace the existing one, how can depend on either the database Oracle or SQL Server, to be some drop/recreate/import
Go to the HSP_USERS table and update the SID of the user 50001 id you have stored as it will be updated by moving the relational database.
Start planning
If everything is done correctly, you should be able to connect to the upsized planning application.
I work in an environment where there are about 50 cases (of flavours - 8.0.4 to 11.2.0.2).
My question relates to the password that are specific application - custom schema passwords and passwords apps...
Do you provide the passwords above your team of developers/EBS Production support in your case TEST/development? (does not reference to PROD here)
If in all these cases, each has its own user account with all the privileges of the user APPS (select any etc...), what is the justification for a developer who need the APPS password?
It is advisable to give the APPS password in TEST/development cases?
Thank you
Do you provide the passwords above your team of developers/EBS Production support in your case TEST/development? (does not reference to PROD here)
We do this in all environments except TEST (which is used for QA and check the deployment scripts), UAT and PROD.
If in all these cases, each has its own user account with all the privileges of the user APPS (select any etc...), what is the justification for a developer who need the APPS password?
If you already have users of database created for the users with all the privileges and password apps should not be shared.
It is advisable to give the APPS password in TEST/development cases?
Please see my response above. In addition, please see:
1. I'm confused here because I don't know if I'm Port already in use error if I use the method later as I will reuse the existing port numbers.
2. in addition, it would be great if you let me know how to get to the database and applications of port numbers. I know there is a rule of thumb to calculate the same.
Kind regards
Mahesh
1. the question of whether I can use existing xml target file method or direct method when cloning a post?
You can do both. If you do not plan to change the DEV instantiate details (port number, directory names, SID,... etc) and then use the existing context files.
2. because I'm refreshing will the node (both source & target in the same server), I get in a 'Port already in use question' I need to reuse already configured port numbers.
Not as you want to stop the DEV instance until you refresh it, if none of the ports will be in use.
3. how to Port serial numbers? I suppose that there is a rule of thumb and a Metalink note for this but can't remember the ID. MetaLink
I'm pretty new to oracle and I would mess around with it some more, I created my database (MVF) and created a test tablespace. Now it's what would be the best way to fill the space of storage of data? can I just insert a few lines in the tablespace? or are there ways more effective to do this? Perhaps someone who could provide me with a pl/sql script for it? or a procedure which would make this (insert x the amount of dummy data) on a regular basis?
Thanks in advance!
(ps this is my first post so please excuse me if my post is not according to the regular lines)
Published by: 994141 on 15-mrt-2013 12:38
Hello
Welcome to the forum!
994141 wrote: Hi all
I'm pretty new to oracle and I would mess around with it some more, I created my database (MVF) and created a test tablespace. Now it's what would be the best way to fill the space of storage of data? can I just insert a few lines in the tablespace?
This is the usual approach. You have some situations you want to test: unique compared to duplicate values, certain combinations of values, some missing columns, the values limits extremes, even the questionable data or invalid (for example birth_date in the future). Create individual INSERT statements for these cases. Write a script to do this, you can use the features of the Editor (cut/paste, search/replace) allows you to easily create statements or re - run them if you get something wrong the first time.
or are there ways more effective to do this?
If you want a large amount of data, you may copy it to an existing table or view. Even if you created the database, you probably have hundreds of lines in views of data, such as object dictionary. For example, if you have 100 rows in an employee table, you could say:
CREATE TABLE employee
( employee_id NUMBER PRIMARY KEY
, last_name VARCHAR2 (20)
, first_name VARCHAR2 (20)
, hire_date DATE
);
INSERT INTO employee
( employee_id
, last_name
, first_name
, hire_date
)
SELECT ROWNUM
, SUBSTR (object_name, 1, 20)
, SUBSTR (owner, 1, 20)
, created
FROM all_objects
WHERE ROWNUM <= 100
;
You might find yourself with a large number of employees 'SYS '.
Here is another way to generate data. Say, you want a table with 60 lines, one line per day during the last 60 days:
CREATE TABLE test_dates
AS
SELECT LEVEL AS dt_id
, SYSDATE - LEVEL AS dt
FROM dual
CONNECT BY LEVEL <= 60
;
Perhaps someone who could provide me with a pl/sql script for it?
As you can see, a lot of things can be done without PL/SQL, just using plain old SQL. Of course, PL/SQL is available if you really need.
I need to be able to recreate the same database with the same content of data files, several series of stress tests. Is it possible to simply save all the Oracle Home Directory, the first series of tests and then restore the directory to return to the initial state?
As seems unlikely, which would be the best approach to "restore" the database to its physical state initial quickly? For example, if the data files got extended due to the large amount of data loaded in the first test, the next round should begin with the initial configuration at all.
Restore the directory to return to the initial state?
Yes, but the Oracle data files generally are NOT reside under ORACLE_HOME.
You need to restore only the data files & not the tree of the software itself
Hello members, What is the best way to clone a solaris db about 80 GB to Windows 2003 server production? Can do datapump exports while the production db is online?
Let's take a step back. What are the requirements and what do mean by the word "clone"?
"Clone" usually involves a physical copy of the database, so the physical structures are identical on the source and the cloned systems. This would require RMAN, if this can be problematic, because you're on 10.2 and the endianness of a Solaris (not x 86 Solaris) and a box is, in my opinion, different Windows.
You can make a logical copy using DataPump. This is a point (actually several points in time by default) copy of the data, but the physical structures will change (index will be built fresh, the data will be physically reorganized, etc.). A logical copy means that the destination database is completely separate from the source, so no hope of the application archive logs, etc.. And generating export requires some input/output relatively heavy on the production database server while it is up, which is also problematic.
How can I create a simulated IO vi. I tried to follow the tutorial, but it did not work. Do I need an array of pre filled with data for the business running or I can generate with one express vi? My IO vi does not seem to enter the case of race ever. What is the process of running this VI? He is running once or whenever the main vi trying some samples?
Hi Bartekluk
What tutorial did you follow? Please post the link and I'll take a look.
I used the following commands: (as of 10 gr 2) expdp DIRECTORY = DATA_PUMP_DIR SCHEMAS = EXCLUDE MDLOG = DUMPFILE = mdlogMETADATA.dmp STATISTICS include happy tablespace = metadata_only =
Then: (in 11 GR 2) Impdp DIRECTORY = DATA_PUMP_DIR DUMPFILE = mdlogMETADATA.dmp include TABLESPACE sqlfile = c.sql =
instead I get ORA-39002: invalid operation ORA-39168: TABLESPACE object path was not found.
Now I'm looking for the incompatible options between (10 gr 2) expdp and impdp (11 GR 2)... and if include = TABLESPACE should be replaced by choice...
During this time... is there something that can tell me if I make mistakes?
Tanks
Hello
The expdp command you listed in your first post may not work. You have exclude and include in the same order.
Exclude said excluding these items but includes everything.
Include says include only these objects and exclude everything else.
If you use the command expdp of your second post, there is no tablespace objects in a schema export. Include = tablespace will only recreate the tablespaces and not objects in storage. If you want to do this, you must delete the schema = MDLOG and add the full = Y. Tablespace definitions are included in a full export.
What is your ultimate goal? You want just the definitions of tablespace moved? If so
complete expdp = include = tablespace directory =...
If you want all the storage space and the objects in these storage spaces, so I think it would be 2 steps:
I'll put up of backups rman for our databases of test & dev. These are cloned manually using commands in the o/s. As all of these databases are clones of prod, they have the same db_id. So I've discovered that I need to use the utility 'nest' change the db_id to save in the rman catalog. However, if I create a new clone of prod and give it a new db_id, older backups for the previous db_id (i.e. the previous clone that this replaced) are still in the recovery catalog.
I see this info in the recovery catalog views, i.e.
----- Select * from rc_database DB_KEY DBID NAME RESETLOGS_CHANGE DBINC_KEY # RESETLOGS_TIME 1 701 3620254146 HOUTEST 3309633197328 21/04/2009 13:14:05 341 342 3688071086 RMANTEST 1 2008-12-22 11:40:30 919 920 3699186203 RMANTEST 3209104 27/04/2009 15:19:52 ----- In this example, db_id 3688071086 no longer exists and I'm unable to realize by using rman commands.
So the problem is, how can we remove these old outdated backups? I tried the recovery window setting to a low value and running "remove obsolete", but the extent of the rman is concerned, it will not erase records for the old db_id and that no longer exists.
For now we are just rman configuration on our test systems, this is why it is not yet in production. Of course, Miss me something fundamental here, maybe someone could point me in the right direction?
The platform is Sun Solaris 10 with Oracle 10.2.0.3
Thank you
See scenario,
1. you have A DBID 1111 data say. 2. register cela in the catalog and took the backup. 3. now you have changed the DBID of 1111 to 2222 for A database. 4. now you must save new data in the recovery catalog as in the recovery catalog database are identified by the DBID. So now, recovery catalog will take this A database as a new database. 5. now, when do you the backup list, it will not display the backup for old DBID. Even if you try to connect to the database as old DBID, it will not be possible as long as changing the DBID of a database means logically that old database has been deleted. It's what will examine RMAN. But the information for the old database DBID will remain in the recovery catalog schema. Giving orders of overlap or remove obsolete commands does not remove this backup from old DB of the disk or tape as the old database is not recognized by the RMAN now. So, you need to delete information on the old database to the catalog of recovery manually from the catalog schema then remove physical backups manually from tape or disk.
I was asked to describe a method to paralyze all servers (physical and virtual) in our datacenter as fast so that possible, but in a particular order. In regards to the MV I have stop a lot over the weekend with a script that runs get - vm '< my server name >' | stop-vmguest-confirm: $false against a list of servers 10. But that was the extent of my PowerCLI script to date. And never change these virtual machines in this list.
Can someone point me in the right direction on the commissioning of the following maybe
Nightly run a script that will meet the list of all virtual machines in my environment (to capture all newly created or recently deleted VMs) and put all the virtual machines in a CSV file or something in an order based on test in the section notes so that for example all the application of test and development servers are grouped to be arrested front of databases of test/dev servers.
Then have a script just wait, the runtime will begin to close all virtual machines in the most recent CSV in the right order.
The reason for this, if you are interested is that 2 weeks ago we had a power failure so we have no generator. Our UPS allows the data center to operate for about 30 minutes. Of course, because these things can never happen at night, at the time that someone was here to start closing things they had only minutes to disable the physical key and virutal machine and they obviously ran out of time. So now we are looking for a logon and perform a 1 click solution to start stop everything.
We will be having a blackout week next to replace some electrical equipment and would like to have the script ready right here.
Anyone who is willing to give me a hand or point me in the right direction?
Thank you
Hello
Nothing really wrong with the approach, but why bother with CSV if you can do hot?
I guess you something in the notes which indicate that it is (testDB, prodDB, testApp). You can also do this if you have a Server naming convention that indicates what it is.
About: config is the parameter "network. IDN.blacklist_chars' that contains what looks like garbage graphics characters. It's a bit weird - as it is corrupt. Is this normal?
My sound has kept silent for while now. I didn't know about the volume control sneaky hidden on the side of my laptop, so I've done all sorts... Today, I was so desperate to hear certain sounds (a little knowledge is a dangerous thing) that I played
I have a HP Pavilion Media Center TV PC. It works fine until last Friday, 10/02/2012. The machine became unbootable plue. On the eve of the machine stopped properly. (We always shut down the computer at night.) Now when I turn on the machine, I hear
I am new to the use of Cisco Networking products. I gave me a mission to determine if our company 5510 and 5505 IPS/IDS. In doing my research I discoververed 5505 have no IPS/IDS, but you can buy a card and 5510 have modules IPS/IDS. How can I determ