(1) Normally there is no mess about where the base folder is left and it will always work. The installation always makes use of the environment variables ORACLE_BASE, ORACLE_HOME and others.(2) It's a terrible idea to install /homeWell, everything will be subject to user permissions and you're gonna pass it by solving errors caused by this. For https://en.wikipedia.org/wiki/Filesystem_Hierarchy_Standard any third party software (not to come from an official repo, not to follow the conventions of the packages officially installed) should go on /opt.(3) The healthiest thing would be to backup your schemes and recreate the partitions of the album, giving more space to / (and over there right to /optand installing there. But that would be my opinion.

The steps to solve this problem have been: Check that the PHP and Instant Client architecture was the same.Rest in php.ini the lines extension=oci8_12c and extension=pdo_oci.Back to download the DLL file from PECL https://pecl.php.net/package/oci8/2.2.0/windows ).Decompress in the same folder both https://www.oracle.com/database/technologies/instant-client.html .Add the route where Instant Client has decompressed PATH.Restart the server.After these steps it worked properly.

To be able to determine whether or not there is a column in ORACLE you can support yourself in the view of the so-called data dictionary DBA_TAB_COLS, this view is composed of a lot of interesting information, and above all that you need, from this view you need these fields: In this https://docs.oracle.com/database/121/REFRN/GUID-857C32FD-AE30-4AB9-811B-AC3A7B91A04D.htm#REFRN23276 you can consult the official document for your version ORACLE 12cNow, with this information in hand, we will create the procedure you need. We use the view of the data dictionary and integrate it into a procedure as follows:DECLARE v_col_exists NUMBER BEGIN SELECT count(*) INTO v_col_exists FROM DBA_TAB_COLS WHERE column_name = 'nombre_columna' AND table_name = 'nombre_tabla'; IF (v_col_exists = 0) THEN EXECUTE IMMEDIATE 'ALTER TABLE nombre_tabla ADD nombre_columna NUMBER(10)'; ELSE DBMS_OUTPUT.PUT_LINE('La columna nombre_columna si existe'); END IF; END; /

You can use CASE.SELECT CASE WHEN ugp.ugp_rolename= 'john_doe' THEN 'rol es john doe' WHEN ugp.ugp_rolename= 'derp' THEN 'rol es derp' ELSE '' END AS "role name", usr_udf_emplid as "ID" FROM ...

Normally, an export/import of a newer version should not be made to an older version. The problem is that the export of the newest version can include objects that do not even exist in the older version.It gives me the impression that you're already aware of this problem, and that's probably why you tried to add the parameter. version during the import.However, it is not during the import you need to specify the parameter version. Rather, you should use the parameter version during export (expdb) for the file dmp do not include devices specific to version 12c.That's why it starts by running again export using the parameter VERSION=11 Just like you got it. And then you can try the import again.

Not estimated, the fact that you change a predetermined program does not affect your compilation or use of other programs at all while maintaining its extension. Now if you want to change the default programs you can use the following https://support.microsoft.com/es-cl/help/4028161/windows-10-change-default-programs as a guide. EDIT I also leave a method of how to remove a default program from your files.How do we delete an association of a file type with a program in Windows?Open Record Editor (regedit) then go HKEY_CURRENT_USER\Software\Microsoft\Windows\CurrentVersion\Explorer\FileExtsto the left, under the key FileExts, identifying subclave sites and their file extensions for those who want to delete default association and remove them Then go HKEY_CLASSES_ROOT and clean if there are subclave sites with extensions those files.Rest and ready. Greetings

You can correct it by doing:first to validate the route and the nameselect name from v$datafile; Now we're looking for the route temp01.dbf the name can vary and we copy it Now if it were C:\oracle\app\oradata\zoom\temp01.dbf We'll add a second datafile with C:\oracle\app\oradata\zoom\temp02.dbfalter tablespace temp add tempfile 'C:\oracle\app\oradata\zoom\temp02.dbf' size 1g;

Dude, you have to put the Oracle version on your question.The error is due to the fact that, from Oracle 12c, there is an installation mode in which a CDB (a container) and multiple PDB (pluggable DB) is created. Container (CDB) users should have the structure you quote in the comment (c##xxxxxx, with only ASCII characters) and these users NO are used to create databases you use in your application. They are to manage the container, the operating characteristics of the BD, logs, boot parameters, etc.To create a user in whose scheme you will create tables and others as you normally did, connect to the PDB (by default it is to serviceName orclpdbnot to confuse with the SID orcl which is the CDB) and create your user normally.I recommend you go to the documentation to better understand the changes, as you also see that you went from one version to another assuming they were the same.

I share the query to get the optional result:DECLARE @columnas VARCHAR(max) SET @columnas = '' SELECT @columnas = coalesce(@columnas + '[' + cast(producto as varchar(12)) + '],', '') FROM (select distinct producto from productos) as DTM set @columnas = left(@columnas,LEN(@columnas)-1) DECLARE @SQLString nvarchar(500); set @SQLString = N' SELECT * FROM #productos PIVOT ( sum(valor) FOR producto IN (' + @columnas + ') ) AS pivot_producto;' EXECUTE sp_executesql @SQLString

I don't see the way you can figure it out. regexp_replace(), the problem is that a variable pattern is detected in length to replace it with a fixed number of characters #. It seems to me much easier, to solve it this way:select lpad(substr(identification,-4), length(identification), '#') from clientes;

In PL/SQL, you cannot assign to a variable the result of a subquery the way you're doing it, you might come from SQL-Server (TSQL) or another engine where you can.In Oracle, what you can do is select intoLike this:Select sum(Monto_Total) into Total from TablaC where id = Amount.id_camp;

The sentence https://docs.oracle.com/cd/B19306_01/server.102/b14200/statements_7003.htm creates a new https://selectasterisco.wordpress.com/2013/07/02/conceptos-y-manejo-basico-de-tablespaces-en-oracle/ which according to the definition of the linked article is:a logical store of database files. Each tablespace owns one or more files (datafiles) where it stores all the information; these files must have a logical structure. When creating a database, you have to create at least one tablespacewhich by default is SYSTEM.In these spaces is where the information of the objects of the database is stored.About your sentence:SIPROJWEB2 is the name of tablespaceLOGGING indicates that the database engine will include in the log of the database the modifications made in the data files of tablespace./u01/app/oracle/product/12.2.0/dbhome_1/dbs/ dbb.DBF is the name of the only data file that will have the new tablespace.The file will have an initial size of 500MB.EXTENT MANAGEMENT LOCAL indicates that tablespace will be administered locally. SEGMENT SPACE MANAGEMENT AUTO indicates that the motor will self-manage the free space within tablespace using a bit map, and the tablespace it will grow automatically if necessary.

In the end I don't know what happened, I did your tips to run administrator, and I followed the installation steps and nothing for which I chose to format the notebook as if I didn't progress I wouldn't be able to continue my tasks. As a conclusion now that I'm thinking, I don't know if it's been a port error or that SQL Server took care of something that Oracle was occupying.

AlejandroThis is a way to get the first record of the set in Oracle (It's not the only one), the key is to order it in the field you require (date in your case) and then nest the set within another and with ROWNUM get a single record (if you want more ROWNUM records).SELECT * FROM MITABLA WHERE ROWNUM=1 ORDER BY MICAMPO In your particular case it would be:SELECT * FROM ( SELECT CASO.IDCASO, CASO.FECHACREACION, GTALT.IDUSUARIO, GTALT.FECHACREACION FROM CASO LEFT JOIN GRUPODETRABAJO GTALT ON CASO.CODUSUARIOALTA = GTALT.IDUSUARIO AND GTALT.FECHACREACION <= CASO.FECHACREACION ORDER BY GTALT.FECHACREACION DESC ) WHERE ROWNUM = 1

TL;DR: To access defined resources on a JEE server, use JNDI.Although concepts are not complicated, technical details have their details and I don't have to define these things for a long time. In any case I think (see that there are no other answers) better give you an approximate answer that may fail in some detail that there is no answer.As I commented, your Datasource will be defined in the JNDI. He'll usually be inside. namespace java: (e.g. java:miDataSourceor java:jboss/1. Through the server management console you can consult the JNDI elements.To get the data, two options:JNDI pure and hard.(This also serves to access any JNDI resource)When you deploy/creas the datasource on the server, you are assigned a JNDI name and accessible by this service:javax.naming.Context ctx = new javax.naming.InitialContext(); DataSource ds = (DataSource) ctx.lookup("java:miDatasource"); The question of names is somewhat smooth; if I remember correctly you can access the name "global" (java:) but that can cause problems; if you have two applications that use the same name for Datasources that are different, you will not be able to have them on the same server.So in your application/webapp, what you do is indicate that Datasource uses and assigns a "local" name to the application; for example in the application jboss-web.xml:<resource-ref> <res-ref-name>jdbc/miDataSourceLocal</res-ref-name><!-- nombre local --> <res-type>javax.sql.DataSource</res-type> <jndi-name>java:/miDataSource</jndi-name><!-- nombre global --> </resource-ref> This creates a JNDI input java:comp/env/jdbc/miDataSourceLocal which is the one you use in your code:DataSource ds = (DataSource) ctx.lookup("java:comp/env/jdbc/miDataSourceLocal"); @ResourceIn instances managed by the server (that you do not new, but create the server, like Servlets and EJBs, you can use @Resource and let the server inject the Datasource:@Resource(name="miDataSourceLocal") private javax.sql.DataSource myDB This simply automates access to JNDI, so the part of managing names to avoid collisions is also applied here.1 If I remember correctly, JBoss forces you to this.

The short answer is: you cannot grant privileges to a user of another database In oracle.Longer this answer:This does not mean that someone connected to a database cannot modify information that resides in another database, simply that you should use a controlled authentication mechanism through what is known as db links. In essence, the db-link defines the connection parameters to the database and how to authenticate it, which can be by a user and password that are burned in the db link; i.e. all users of database A, when they appear as the same user when performing operations in database B, or may be the CURRENT USER. The easiest is the first, for example, in the host where your database is Orcl2, you set a TNS alias with the parameters Orcl1 and execute this instruction:CREATE DATABASE LINK rOrcl1 CONNECT TO Orcl2User IDENTIFIED BY Orcl2Password USING 'Orcl1'; In the Orcl1 database there must be the user Orcl2User and your password Orcl2Password.From there, all Orcl2 users can access Orcl1 objects with the permissions of Orcl2User, for example, perform:select * from juan.Inventario@rOrcl1; If that's not enough for you, then let's talk about it. database for the CURRENT USER:In this case, users must be global users that must exist in both databases (LDAP).The creation of the database link is this way:CREATE DATABASE LINK rOrcl1 CONNECT TO CURRENT_USER USING 'Orcl1'; The use is the same, but this time, each user will have their own privileges (which are locally granted in each database, since users exist in both).For more information, visit the documentation of the clause https://docs.oracle.com/cd/B19306_01/server.102/b14200/statements_8003.htm#sthref8228 of create usercomplemented by the example https://docs.oracle.com/cd/B19306_01/server.102/b14200/statements_5005.htm#i2143859 documentation CREATE DATABASE LINK

I found that to make it work by calling from php it should take into account the size of buffer for each variable in or out of the procedure, which I still don't understand is Why does the inserts with internal search not return the last id?

To format the number in ORACLE SQL you have different elements that help you handle the symbols of the coins. In this case'L' que retorna el símbolo de moneda por defecto definido en NLS_CURRENCY 'G' que retorna el símbolo entero por defecto definido en NLS_NUMERIC_CHARACTER 'D' que retorna el símbolo decimal por defecto definido en NLS_NUMERIC_CHARACTER In your case, the format would TO_CHAR(sueldo_emp, 'L9G999G999''NLS_NUMERIC_CHARACTERS=".,"') In given case your numbers are integers

Seeing the code of the current implementation of TFDQuery, I see clearly that by using the method OpenOrExecute There is no way to collect the value of the affected rows. The method Just try open the result of the command, as if it were a sentence select and ignores the exception that occurs if the sentence does not produce at least one set of data as a result.The underlying way to get the number of rows affected by a sentence insert/update/delete It's different.One way you could work on the-render of this limitation is to prepare the command and, depending on the type of sentence that FireDAC's own driver determines that it is to open or execute it, in which case the property RowsAffected is updated correctly.The basic idea is:Join the SQLPrepare the sentenceTake a look at the value of the property CommandKind property Command of TFDQuery, that will have one of the values skSelect, skInsert, skUpdateamong others.Depending on that value, you decide which method to execute (Open, ExecSQL, OpenOrExecute)If executed Open or ExecSQL, the property RowsAffected will show you correctly the number of affected rows.The code I've tried is:begin FDQuery1.SQL.Text := 'tu sentencia aquí'; FDQuery1.Prepare; if Assigned(FDQuery1.Command) then case FDQuery1.Command.CommandKind of skNotResolved, skUnknown: FDQuery1.OpenOrExecute; skSelect, skSelectForLock, skSelectForUnLock: FDQuery1.Open; skDelete, skInsert, skMerge, skUpdate, skCreate, skAlter, skDrop, skStoredProc, skStoredProcWithCrs, skStoredProcNoCrs, skExecute, skStartTransaction, skCommit, skRollback, skSet, skSetSchema, skOther: FDQuery1.ExecSQL; end else FDQuery1.OpenOrExecute; Label2.Caption := IntToStr(FDQuery1.RowsAffected); end; I've tested SQL Server and this has worked me reasonably well. I suggest that before you implement it you also try it with Oracle (since it depends on the specific driver of the database), and also that you make sure to perform the correct operation for each type of sentence (I have done a quick test, but I cannot guarantee the result in cases like skSelectForLock or equally exotic values.

1-You will never compile because the "xxsumen" table does not exist at the time of the compilation.2-why do the temporary tables be created within the procedure?It is not a good practice and compromises the security of the database.The best thing is to create them previously.