Is there any subprogram similar to DBMS_METADATA.GET_DDL that can actually export the table data as INSERT statements?
For example, using DBMS_METADATA.GET_DDL('TABLE', 'MYTABLE', 'MYOWNER') will export the CREATE TABLE script for MYOWNER.MYTABLE. Any such things to generate all data from MYOWNER.MYTABLE as INSERT statements?
I know that for instance TOAD Oracle or SQL Developer can export as INSERT statements pretty fast but I need a more programmatically way for doing it. Also I cannot create any procedures or functions in the database I'm working.
Thanks.
As far as I know, there is no Oracle supplied package to do this. And I would be skeptical of any 3rd party tool that claims to accomplish this goal, because it's basically impossible.
I once wrote a package like this, and quickly regretted it. It's easy to get something that works 99% of the time, but that last 1% will kill you.
If you really need something like this, and need it to be very accurate, you must tightly control what data is allowed and what tools can be used to run the script. Below is a small fraction of the issues you will face:
Escaping
Single inserts are very slow (especially if it goes over a network)
Combining inserts is faster, but can run into some nasty parsing bugs when you start inserting hundreds of rows
There are many potential data types, including custom ones. You may only have NUMBER, VARCHAR2, and DATE now, but what happens if someone adds RAW, BLOB, BFILE, nested tables, etc.?
Storing LOBs requires breaking the data into chunks because of VARCHAR2 size limitations (4000 or 32767, depending on how you do it).
Character set issues - This will drive you ¿¿¿¿¿¿¿ insane.
Enviroment limitations - For example, SQL*Plus does not allow more than 2500 characters per line, and will drop whitespace at the end of your line.
Referential Integrity - You'll need to disable these constraints or insert data in the right order.
"Fake" columns - virtual columns, XML lobs, etc. - don't import these.
Missing partitions - If you're not using INTERVAL partitioning you may need to manually create them.
Novlidated data - Just about any constraint can be violated, so you may need to disable everything.
If you want your data to be accurate you just have to use the Oracle utilities, like data pump and export.
Why don't you use regular export ?
If you must you can generate the export script:
Let's assume a Table myTable(Name VARCHAR(30), AGE Number, Address VARCHAR(60)).
select 'INSERT INTO myTable values(''' || Name || ','|| AGE ||',''' || Address ||''');' from myTable
Oracle SQL Developer does that with it's Export feature. DDL as well as data itself.
Can be a bit unconvenient for huge tables and likely to cause issues with cases mentioned above, but works well 99% of the time.
Related
I'm using an external tool that scans tables in my database. It uses dba_objects.last_ddl_time to determine which tables have been scanned. Obviously, this strategy does not work if the table data is modified in between scans so sometimes I have to help it...
I need a way to "bump" the Last DDL time without actually changing anything.
I'm looking for the simplest possible instant DDL statement that can be executed on any table, knowing just the table name.
I have sysdba privileges.
Edit:
For example, I can use comment on table xxx is 'Boom'; but then I lose the original comment. I know how to fix this, but then it is no longer an small and easy statement I can quickly time in sql*plus
Changing LOGGING/NOLOGGING is pretty fast (though not instant).
If you set the LOGGING attribute back to itself, it will notch the LAST_DDL_TIME without making any real change to the table. This example below tries to touch every table except sys tabels (presumably you'd want more limits here)
BEGIN
FOR TABLE_POINTER IN (SELECT OWNER, TABLE_NAME, DECODE(LOGGING,'YES','LOGGING','NOLOGGING') DO_LOGGING
FROM DBA_TABLES WHERE OWNER NOT IN ('SYSTEM','SYS','SYSBACKUP','MDSYS' --etc. other restrictions here
))
LOOP
EXECUTE IMMEDIATE UTL_LMS.FORMAT_MESSAGE('ALTER TABLE %s.%s %s',TABLE_POINTER.OWNER, TABLE_POINTER.TABLE_NAME, TABLE_POINTER.DO_LOGGING);
END LOOP;
END;
/
EDIT: The above wouldn't work with temp tables. An alternative such as setting PCT_FREE to itself or another suitable attribute may be preferable. You may need to handle IOTs, Partitioned Tables, etc. differently than the rest of the tables as well.
We use Oracle 10g and Oracle 11g.
We also have a layer to automatically compose queries, from pseudo-SQL code written in .net (something like SqlAlchemy for Python).
Our layer currently wraps any string in single quotes ' and, if contains non-ANSI characters, it automatically compose the UNISTR with special characters written as unicode bytes (like \00E0).
Now we created a method for doing multiple inserts with the following construct:
INSERT INTO ... (...)
SELECT ... FROM DUAL
UNION ALL SELECT ... FROM DUAL
...
This algorithm could compose queries where the same string field is sometimes passed as 'my simple string' and sometimes wrapped as UNISTR('my string with special chars like \00E0').
The described condition causes a ORA-12704: character set mismatch.
One solution is to use the INSERT ALL construct but it is very slow compared to the one used now.
Another solution is to instruct our layer to put N in front of any string (except for the ones already wrapped with UNISTR). This is simple.
I just want to know if this could cause any side-effect on existing queries.
Note: all our fields on DB are either NCHAR or NVARCHAR2.
Oracle ref: http://docs.oracle.com/cd/B19306_01/server.102/b14225/ch7progrunicode.htm
Basicly what you are asking is, is there a difference between how a string is stored with or without the N function.
You can just check for yourself consider:
SQL> create table test (val nvarchar2(20));
Table TEST created.
SQL> insert into test select n'test' from dual;
1 row inserted.
SQL> insert into test select 'test' from dual;
1 row inserted.
SQL> select dump(val) from test;
DUMP(VAL)
--------------------------------------------------------------------------------
Typ=1 Len=8: 0,116,0,101,0,115,0,116
Typ=1 Len=8: 0,116,0,101,0,115,0,116
As you can see identical so no side effect.
The reason this works so beautifully is because of the elegance of unicode
If you are interested here is a nice video explaining it
https://www.youtube.com/watch?v=MijmeoH9LT4
I assume that you get an error "ORA-12704: character set mismatch" because your data inside quotes considered as char but your fields is nchar so char is collated using different charsets, one using NLS_CHARACTERSET, the other NLS_NCHAR_CHARACTERSET.
When you use an UNISTR function, it converts data from char to nchar (in any case that also converts encoded values into characters) as the Oracle docs say:
"UNISTR takes as its argument a text literal or an expression that
resolves to character data and returns it in the national character
set."
When you convert values explicitly using N or TO_NCHAR you only get values in NLS_NCHAR_CHARACTERSET without decoding. If you have some values encoded like this "\00E0" they will not be decoded and will be considered unchanged.
So if you have an insert such as:
insert into select N'my string with special chars like \00E0',
UNISTR('my string with special chars like \00E0') from dual ....
your data in the first inserting field will be: 'my string with special chars like \00E0' not 'my string with special chars like à'. This is the only side effect I'm aware of. Other queries should already use NLS_NCHAR_CHARACTERSET encoding, so it shouldn't be any problem using an explicit conversion.
And by the way, why not just insert all values as N'my string with special chars like à'? Just encode them into UTF-16 (I assume that you use UTF-16 for nchars) first if you use different encoding in 'upper level' software.
use of n function - you have answers already above.
If you have any chance to change the charset of the database, that would really make your life easier. I was working on huge production systems, and found the trend that because of storage space is cheap, simply everyone moves to AL32UTF8 and the hassle of internationalization slowly becomes the painful memories of the past.
I found the easiest thing is to use AL32UTF8 as the charset of the database instance, and simply use varchar2 everywhere. We're reading and writing standard Java unicode strings via JDBC as bind variables without any harm, and fiddle.
Your idea to construct a huge text of SQL inserts may not scale well for multiple reasons:
there is a fixed length of maximum allowed SQL statement - so it won't work with 10000 inserts
it is advised to use bind variables (and then you don't have the n'xxx' vs unistr mess either)
the idea to create a new SQL statement dynamically is very resource unfriedly. It does not allow Oracle to cache any execution plan for anything, and will make Oracle hard parse your looong statement at each call.
What you're trying to achieve is a mass insert. Use the JDBC batch mode of the Oracle driver to perform that at light-speed, see e.g.: http://viralpatel.net/blogs/batch-insert-in-java-jdbc/
Note that insert speed is also affected by triggers (which has to be executed) and foreign key constraints (which has to be validated). So if you're about to insert more than a few thousands of rows, consider disabling the triggers and foreign key constraints, and enable them after the insert. (You'll lose the trigger calls, but the constraint validation after insert can make an impact.)
Also consider the rollback segment size. If you're inserting a million of records, that will need a huge rollback segment, which likely will cause serious swapping on the storage media. It is a good rule of thumb to commit after each 1000 records.
(Oracle uses versioning instead of shared locks, therefore a table with uncommitted changes are consistently available for reading. The 1000 records commit rate means roughly 1 commit per second - slow enough to benefit of write buffers, but quick enough to not interfer with other humans willing to update the same table.)
Questions have been asked in the past that seems to handle pieces of my full question, but I'm not finding a totally good answer.
Here is the situation:
I'm importing data from an old, but operational and production, Oracle server.
One of the columns is created as LONG RAW.
I will not be able to convert the table to a BLOB.
I would like to use a global temporary table to pull out data each time I call to the server.
This feels like a good answer, from here: How to create a temporary table in Oracle
CREATE GLOBAL TEMPORARY TABLE newtable
ON COMMIT PRESERVE ROWS
AS SELECT
MYID, TO_LOB("LONGRAWDATA") "BLOBDATA"
FROM oldtable WHERE .....;
I do not want the table hanging around, and I'd only do a chunk of rows at a time, to pull out the old table in pieces, each time killing the table. Is it acceptable behavior to do the CREATE, then do the SELECT, then DROP?
Thanks..
--- EDIT ---
Just as a follow up, I decided to take an even different approach to this.
Branching the strong-oracle package, I was able to do what I originally hoped to do, which was to pull the data directly from the table without doing a conversion.
Here is the issue I've posted. If I am allowed to publish my code to a branch, I'll post a follow up here for completeness.
Oracle ODBC Driver Release 11.2.0.1.0 says that Prefetch for LONG RAW data types is supported, which is true.
One caveat is that LONG RAW can technically be up to 2GB in size. I had to set a hard max size of 10MB in the code, which is adequate for my use, so far at least. This probably could be a variable sent in to the connection.
This fix is a bit off original topic now however, but it might be useful to someone else.
With Oracle GTTs, it is not be necessary to drop and create each time, and you don't need to worry about data "hanging around." In fact, it's inadvisable to drop and re-create. The structure itself persists, but the data in it does not. The data only persists within each session. You can test this by opening up two separate clients, loading data with one, and you will notice it's not there in the second client.
In effect, each time you open a session, it's like you are reading a completely different table, which was just truncated.
If you want to empty the table within your stored procedure, you can always truncate it. Within a stored proc, you will need to execute immediate if you do this.
This is really handy, but it also can make debugging a bear if you are implementing GTTs through code.
Out of curiosity, why a chunk at a time and not the entire dataset? What kind of data volumes are you talking about?
-- EDIT --
Per our comments conversation, this is very raw and untested, but I hope it will give you an idea what I mean:
CREATE OR REPLACE PROCEDURE LOAD_DATA()
AS
TOTAL_ROWS number;
TABLE_ROWS number := 1;
ROWS_AT_A_TIME number := 100;
BEGIN
select count (*)
into TOTAL_ROWS
from oldtable;
WHILE TABLE_ROWS <= TOTAL_ROWS
LOOP
execute immediate 'truncate table MY_TEMP_TABLE';
insert into MY_TEMP_TABLE
SELECT
MYID, TO_LOB(LONGRAWDATA) as BLOBDATA
FROM oldtable
WHERE ROWNUM between TABLE_ROWS and TABLE_ROWS + ROWS_AT_A_TIME;
insert into MY_REMOTE_TABLE#MY_REMOTE_SERVER
select * from MY_TEMP_TABLE;
commit;
TABLE_ROWS := TABLE_ROWS + ROWS_AT_A_TIME;
END LOOP;
commit;
end LOAD_DATA;
I have a sql script that contains 700,000 Insert statements. I tried to run through Oracle Developer, but it failed to load the file itself. I tried to run from sqlplus but its taking quite a long time to execute such large file.
To speed up, I have deleted all the constraints on the table but there is no improvement.
I have looked for information, Please have been suggesting
Split the file into manageable sizes - which I will fall back as my last resort.
Sql Loader - as far as I understand SQL Loader is to export from the db in specific format and load them into the db with CTL.
Is there any better way to handle this scenario.
Convert the file from many SQL statements to a smaller number of PL/SQL blocks to reduce the round-trip overhead. This only requires a few minutes with a text editor and can improve performance by orders of magnitude, especially over slow networks.
Every 10,000 lines, add a begin and an end; to the file.
Change this:
insert into ...
insert into ...
insert into ...
...
To this:
begin
insert into ...
insert into ...
insert into ...
...
end;
/
Don't convert the entire file into one large PL/SQL block though. There is a limit to the size of anonymous PL/SQL blocks, you might get a parser error.
Agree with previous answer above, 700,000 inserts via script is going to be slow no matter what you do with it - load up your data as an external table or use SQL*Loader.
However
If you want to execute a large script with SQL Developer, don't OPEN the file - we have to open and parse and display the contents of that file. Ow.
Just do this in the worksheet
#script_name
That will execute the script.
To speed it up even further, hide or minimize the output area of the worksheet.
It's still not going to be super-fast with 700,000 inserts though.
AFAIK the option 2 is correct.
Use sed/awk/perl to convert the file into CSV (or fixed width) input file.
disable constraints, indexes, (possibly drop unique indexes)
create control file for your input file
exec sqlldr (turn direct path load on)
And this should finish withing few seconds.
You have also to check if you have some triggers in the table where you are doing the inserts. It can slow down the process if a lot of logic is coded behind.
The problem I am trying to solve:
I have a SAS dataset work.testData (in the work library) that contains 8 columns and around 1 million rows. All columns are in text (i.e. no numeric data). This SAS dataset is around 100 MB in file size. My objective is to have a step to parse this entire SAS dataset into Oracle. i.e. sort of like a "copy and paste" of the SAS dataset from the SAS platform to the Oracle platform. The rationale behind this is that on a daily basis, this table in Oracle gets "replaced" by the one in SAS which will enable downstream Oracle processes.
My approach to solve the problem:
One-off initial setup in Oracle:
In Oracle, I created a table called testData with a table structure pretty much identical to the SAS dataset testData. (i.e. Same table name, same number of columns, same column names, etc.).
On-going repeating process:
In SAS, do a SQL-pass through to truncate ora.testData (i.e. remove all rows whilst keeping the table structure). This ensure the ora.testData is empty before inserting from SAS.
In SAS, a LIBNAME statement to assign the Oracle database as a SAS library (called ora). So I can "see" what's in Oracle and perform read/update from SAS.
In SAS, a PROC SQL procedure to "insert" data from the SAS dataset work.testData into the Oracle table ora.testData.
Sample codes
One-off initial setup in Oracle:
Step 1: Run this Oracle SQL Script in Oracle SQL Developer (to create table structure for table testData. 0 rows of data to begin with.)
DROP TABLE testData;
CREATE TABLE testData
(
NODENAME VARCHAR2(64) NOT NULL,
STORAGE_NAME VARCHAR2(100) NOT NULL,
TS VARCHAR2(10) NOT NULL,
STORAGE_TYPE VARCHAR2(12) NOT NULL,
CAPACITY_MB VARCHAR2(11) NOT NULL,
MAX_UTIL_PCT VARCHAR2(12) NOT NULL,
AVG_UTIL_PCT VARCHAR2(12) NOT NULL,
JOBRUN_START_TIME VARCHAR2(19) NOT NULL
)
;
COMMIT;
On-going repeating process:
Step 2, 3 and 4: Run this SAS code in SAS
******************************************************;
******* On-going repeatable process starts here ******;
******************************************************;
*** Step 2: Trancate the temporary Oracle transaction dataset;
proc sql;
connect to oracle (user=XXX password=YYY path=ZZZ);
execute (
truncate table testData
) by oracle;
execute (
commit
) by oracle;
disconnect from oracle;
quit;
*** Step 3: Assign Oracle DB as a libname;
LIBNAME ora Oracle user=XXX password=YYY path=ZZZ dbcommit=100000;
*** Step 4: Insert data from SAS to Oracle;
PROC SQL;
insert into ora.testData
select NODENAME length=64,
STORAGE_NAME length=100,
TS length=10,
STORAGE_TYPE length=12,
CAPACITY_MB length=11,
MAX_UTIL_PCT length=12,
AVG_UTIL_PCT length=12,
JOBRUN_START_TIME length=19
from work.testData;
QUIT;
******************************************************;
**** On-going repeatable process ends here *****;
******************************************************;
The limitation / problem to my approach:
The Proc SQL step (that transfer 100 MB of data from SAS to Oracle) takes around 5 hours to perform - the job takes too long to run!
The Question:
Is there a more sensible way to perform data transfer from SAS to Oracle? (i.e. updating an Oracle table from SAS).
First off, you can do the drop/recreate from SAS if that's a necessity. I wouldn't drop and recreate each time - a truncate seems easier to get the same results - but if you have other reasons then that's fine; but either way you can use execute (truncate table xyz) from oracle or similar to drop, using a pass-through connection.
Second, assuming there are no constraints or indexes on the table - which seems likely given you are dropping and recreating it - you may not be able to improve this, because it may be based on network latency. However, there is one area you should look in the connection settings (which you don't provide): how often SAS commits the data.
There are two ways to control this, the DBCOMMMIT setting and the BULKLOAD setting. The former controls how frequently commits are executed (so if DBCOMMIT=100 then a commit is executed every 100 rows). More frequent commits = less data is lost if a random failure occurs, but much slower execution. DBCOMMIT defaults to 0 for PROC SQL INSERT, which means just make one commit (fastest option assuming no errors), so this is less likely to be helpful unless you're overriding this.
Bulkload is probably my recommendation; that uses SQLLDR to load your data, ie, it batches the whole bit over to Oracle and then says 'Load this please, thanks.' It only works with certain settings and certain kinds of queries, but it ought to work here (subject to other conditions - read the documentation page above).
If you're using BULKLOAD, then you may be up against network latency. 5 hours for 100 MB seems slow, but I've seen all sorts of things in my (relatively short) day. If BULKLOAD didn't work I would probably bring in the Oracle DBAs and have them troubleshoot this, starting from a .csv file and a SQL*LDR command file (which should be basically identical to what SAS is doing with BULKLOAD); they should know how to troubleshoot that and at least be able to monitor performance of the database itself. If there are constraints on other tables that are problematic here (ie, other tables that too-frequently recalculate themselves based on your inserts or whatever), they should be able to find out and recommend solutions.
You could look into PROC DBLOAD, which sometimes is faster than inserts in SQL (though all in all shouldn't really be, and is an 'older' procedure not used too much anymore). You could also look into whether you can avoid doing a complete flush and fill (ie, if there's a way to transfer less data across the network), or even simply shrinking the column sizes.