Htmldb 1.6 synonym / DbLink issue

Hi,
1. I've tested the following issue in 3 version 1.5 , 1.5.1 , 1.6 and it seems issue is yet not solved.
2. When you create a publich synonym for a remote table or even use the dblink directly , you can't based anythy form / report etc only using a free query method.
3. Is there any option to create this not using manual selects / handling ...
Thanks
Etay G.

To create the views:
SELECT 'CREATE OR REPLACE VIEW '||Lower(object_name)||' AS SELECT * FROM '||Lower(object_name)||'@database_link;' FROM user_objects@database_link
WHERE object_Type IN ('TABLE','VIEW');
Just replace 'database_link' with your database link.
Sequence number generators are tricky because you cannot create a view to select the nextval of a sequence number generator in another database over a database link. You have to just append the database link to the sequence reference or create a view which calls a function in the remote database which in turn returns the nextval of the sequence.

Similar Messages

DB Links vs. Public Synonyms Security issue

I have been debating on using either a public synonym or a db link for my purpose. I have a dev, test, and prod database. I have applications that have been developed using a public synonym. I know that if I were to switch to db links I would have to go back and change the applications to have access to the tables. I was wondering if it would be more secure to just have the public synonyms with select privilege or have db links. I need to decide if I should use db links for various users in the same DB (e.g. prod) and to use db links from one DB to another. Can someone explain what the security risks are between the two and which would be safer to use?
Thanks

I appreciate that you have taken the time to read my post. The version of the database I am using is 10g. I guess to clarify my post I am asking if someone can provide me advantages and disadvantages of using either public synonyms or db links. I don't know if I am clear, but when I refer to a DB Link I am referring to the following type of access where a client in a database A can access information in a remote database B (e.g. schema1.table2@databaseB). I am not sure what you mean by "Database links are two entirely different technologies that do two entirely different things". But I would appreciate if someone with DBA experience can provide some insight regarding security issues associated with using public synonyms and db links.
Thank you,

DBLINK ISSUE

Hi
Please advise on the following scnerio regarding the performance when handle with millions of records
1.Using dblink to select data from database and compare against another table
2.createb temporary table,insert data into temporary table using dblink and compare against another table
Thanks and Regards,
MR

Like every other problem I don't think we can make a generic statement saying A performs better than B or vice-versa. It depends on many factors. Well to start with it would be ideal if you could post atleast a draft version of what you are trying to do and preferably an explain plan. Then hopefully you will be able to get some suitable answers rather making a wild guess.
Regards
Raj

DBlink issue from 9i to 11g

I created Public DB Link from 9.2.0.8 to 11g database. And getting following error:
ORA-01017: invalid username/password; logon denied
ORA-02063: preceding line from TEST11
I tested sqlplus connection and it works fine.
Any suggestion?

Right. Because Oracle always stores unquoted identifiers (in this case the password) in upper case. In the past, it didn't matter because the destination database didn't care about the password's case. Now that the destination database does care about the case of the password (assuming a default 11g install), you can no longer get away with using case-insensitive passwords in your database link definitions. If you create database links to non-Oracle databases (where passwords are generally case sensitive), you've always had to specify the password in double-quotes.
This is an expected outgrowth of the move to enhance security in 11g by making passwords case sensitive.
Justin

Synonym required to query custom tables from Oracle Apps database

When creating an Answers report that queries from a table owned by a custom schema (not APPS), I cannot query the table and get error. Once I add a Synonym for APPS on this table, then OBIEE works normally.
Since OBIEE is logging in as the APPS user, are Synonyms required for all non-APPS tables/views or is there another feature of the Administrator where it can store the table alias (our typical owner of custom tables is XXFI)?

OK, well "compare" can mean one of two things: Compare structure or compare contents. Here is a quick script to compare column structures of two tables on one database:
(select COLUMN_NAME,
DATA_TYPE,
DATA_LENGTH,
DATA_PRECISION,
DATA_SCALE,
NULLABLE
from dba_Tab_columns
where owner=:OWNR1
and table_name = :tablename
minus
select COLUMN_NAME,
DATA_TYPE,
DATA_LENGTH,
DATA_PRECISION,
DATA_SCALE,
NULLABLE
from dba_Tab_columns
where owner=:ownr2
and table_name = :tablename)
union all
(select COLUMN_NAME,
DATA_TYPE,
DATA_LENGTH,
DATA_PRECISION,
DATA_SCALE,
NULLABLE
from dba_Tab_columns
where owner=:ownr2
and table_name = :tablename
minus
select COLUMN_NAME,
DATA_TYPE,
DATA_LENGTH,
DATA_PRECISION,
DATA_SCALE,
NULLABLE
from dba_Tab_columns
where owner=:ownr1
and table_name = :tablename)
If this query returns any rows, then these indicate that there are structural differences between the tables. We do a minus in both directions to ensure that an additional column in either schema will be returned in the query.
If you are going across dblinks to remote tables then you have to amend the "dba_tab_columns" to "sys.dba_tab_columns@yourdblink"
If you also want to compare indexes, triggers, etc then do the same sort of thing for the associated dba_ views for those objects.
And if you want to compare table contents, then often the fastest way is also to check minuses in both directions, if the tables are not too big:
e.g.
(select * from schema1.table@dblink1
minus
select * from schema2.table@dblink2)
union all
(select * from schema2.table@dblink2
minus
select * from schema1.table@dblink1)
local tables, of course, simply omit the "@dblink" issues.
There are tools to help in such things. TOAD, for example, has a pretty good schema comparison tool, and there are plenty of other options out there. But if you need to script this yourself then the logic I've shown is a good starting point.
Cheers,
Mike

Error while executing procedure in TOAD which contains parameter and dblink

Hi all,
I am getting error while executing the following code in TOAD and APEX.
It contains parameter( start_date and end_date) and dblink(@TO_FUT).
When I hardcode the parameter values, It works fine.
When I dont give hardcoded values in the procedure, the error comes.
When I comment the lines whereever parameter comes, still the error comes.
I am not sure whether it is parameter issue or dblink issue or some other issue
Is anybody having any clue about this error?
Your help is very much appreciated.
The code:
CREATE OR REPLACE PROCEDURE PROC_ENR_ACC(START_DATE DATE,END_DATE DATE)
AS
BEGIN
EXECUTE IMMEDIATE 'TRUNCATE TABLE OUTPUT_TABLE';
INSERT INTO OUTPUT_TABLE
(FIRST_NAME ,LAST_NAME,HID ,EFF_DT,CID,PLAN, FLAG)
SELECT
TEMP1.FIRST_NAME,
TEMP1.LAST_NAME,
TEMP1.HID,
TEMP1.EFF_DT,
TEMP1.CID,
TEMP1.PLAN,
NVL((SELECT 'Y'
FROM FADMIN.MEMI_TABLE@TO_FUT MEMI
WHERE MEMI.MEME_CK = TEMP1.MEME_CK
AND MEMI.TXN_EFF_DT = TEMP1.EFF_DT
AND MEMI.PBP = TEMP1.PLAN
AND MEMI.MEMI_TXN_CD IN ('AC','RE')
AND MEMI.MEMI_SEQ_NO = (SELECT MAX (MI.MEMI_SEQ_NO)
FROM FADMIN.MEMI_TABLE@TO_FUT MI
WHERE MI.MEME_CK = MEMI.MEME_CK
AND MI.TXN_EFF_DT = MEMI.TXN_EFF_DT
AND MI.PBP = MEMI.PBP
AND MI.MEMI_TXN_CD IN ('AC','RE')
AND MI.MEMI_INTF_STS = 03 )),'N') AS FLAG
FROM
( SELECT
M1.FIRST_NAME ,
M1.LAST_NAME ,
M1.HID ,
M1.EFF_DT,
M1.CID,
M1.PLAN ,
(SELECT MEME.MEME_CK
FROM FADMIN.MEME_TABLE@TO_FUT MEME, FADMIN.GRGR_TABLE@TO_FUT GRGR
WHERE MEME.M_HIN = M1.HID
AND MEME.GRGR_CK = GRGR.GRGR_CK
AND GRGR.GRGR_ID = M1.GRGR_ID
)AS MEME_CK
FROM FADMIN.METR_TABLE@TO_FUT M1,
( SELECT M3.HID, M3.EFF_DT FROM FADMIN.METR_TABLE@TO_FUT M3
WHERE M3.TYPE IN ('60','61')
AND M3.REPLY IN ('117','118')
AND TRUNC (M3.METR_TRANS_DT) BETWEEN START_DATE AND END_DATE
AND M3.SYIN_INST = (SELECT MAX (M4.SYIN_INST)
FROM FADMIN.METR_TABLE@TO_FUT M4
WHERE M4.HID = M3.HID
AND M4.TYPE IN ('60','61')
AND M4.REPLY IN ('117','118')
AND TRUNC (M4.EFF_DT) = TRUNC (M3.EFF_DT))
) TEMP
WHERE M1.TYPE IN ('60','61')
AND M1.REPLY ='011'
AND TRUNC (M1.METR_TRANS_DT) BETWEEN START_DATE AND END_DATE
AND M1.SYIN_INST = (SELECT MAX (M2.SYIN_INST)
FROM FADMIN.METR_TABLE@TO_FUT M2
WHERE M2.HID = M1.HID
AND M2.TYPE IN ('60','61')
AND M2.REPLY ='011'
AND TRUNC (M2.EFF_DT) = TRUNC (M1.EFF_DT))
AND M1.HID = TEMP.HID (+)
AND M1.EFF_DT = TEMP.EFF_DT (+)
AND TEMP.HID IS NULL
UNION
SELECT
M1.FIRST_NAME ,
M1.LAST_NAME ,
M1.HID ,
M1.EFF_DT,
M1.CID,
M1.PLAN ,
(SELECT MEME.MEME_CK
FROM FADMIN.MEME_TABLE@TO_FUT MEME, FADMIN.GRGR_TABLE@TO_FUT GRGR
WHERE MEME.M_HIN = M1.HID
AND MEME.GRGR_CK = GRGR.GRGR_CK
AND GRGR.GRGR_ID = M1.GRGR_ID
)AS MEME_CK
FROM FADMIN.METR_TABLE@TO_FUT M1
WHERE M1.TYPE IN ('60','61')
AND M1.REPLY IN ('016', '017', '022', '023')
AND TRUNC (M1.METR_TRANS_DT) BETWEEN START_DATE AND END_DATE
AND M1.SYIN_INST = (SELECT MAX (M2.SYIN_INST)
FROM FADMIN.METR_TABLE@TO_FUT M2
WHERE M2.HID = M1.HID
AND M2.TYPE IN ('60','61')
AND M2.REPLY IN ('016', '017', '022', '023')
AND TRUNC (M2.EFF_DT) = TRUNC (M1.EFF_DT))
)TEMP1 ;
END;
The error:
ORA-00918: column ambiguously defined
ORA-02063: preceding line from TO_FUT
ORA-06512: at "PROC_ENR_ACC", line 5
ORA-06512: at line 5

Okay
Here's the code formatted
CREATE OR REPLACE PROCEDURE PROC_ENR_ACC (START_DATE DATE, END_DATE DATE)
AS
BEGIN
   EXECUTE IMMEDIATE 'TRUNCATE TABLE OUTPUT_TABLE';
   INSERT INTO OUTPUT_TABLE (FIRST_NAME,
                             LAST_NAME,
                             HID,
                             EFF_DT,
                             CID,
                             PLAN,
                             FLAG)
   SELECT   TEMP1.FIRST_NAME,
            TEMP1.LAST_NAME,
            TEMP1.HID,
            TEMP1.EFF_DT,
            TEMP1.CID,
            TEMP1.PLAN,
            NVL (
                 (SELECT   'Y'
                  FROM   MEMI_TABLE MEMI
                  WHERE       MEMI.MEME_CK = TEMP1.MEME_CK
                  AND MEMI.TXN_EFF_DT = TEMP1.EFF_DT
                  AND MEMI.PBP = TEMP1.PLAN
                  AND MEMI.MEMI_TXN_CD IN ('AC', 'RE')
                  AND MEMI.MEMI_SEQ_NO = (SELECT   MAX (MI.MEMI_SEQ_NO)
                                          FROM   MEMI_TABLE MI
                                          WHERE   MI.MEME_CK = MEMI.MEME_CK
                                          AND MI.TXN_EFF_DT = MEMI.TXN_EFF_DT
                                          AND MI.PBP = MEMI.PBP
                                          AND MI.MEMI_TXN_CD IN ('AC', 'RE')
                                          AND MI.MEMI_INTF_STS = 03)
                  ),'N') FLAG
   FROM (SELECT   M1.FIRST_NAME,
                  M1.LAST_NAME,
                  M1.HID,
                  M1.EFF_DT,
                  M1.CID,
                  M1.PLAN,
                  (SELECT   MEME.MEME_CK
                   FROM   MEME_TABLE MEME, GRGR_TABLE GRGR
                   WHERE       MEME.M_HIN = M1.HID
                   AND MEME.GRGR_CK = GRGR.GRGR_CK
                   AND GRGR.GRGR_ID = M1.GRGR_ID) MEME_CK
         FROM   METR_TABLE M1,
                (SELECT   M3.HID, M3.EFF_DT
                 FROM   METR_TABLE M3
                 WHERE   M3.TYPE IN ('60', '61')
                 AND M3.REPLY IN ('117', '118')
                 AND TRUNC (M3.METR_TRANS_DT) BETWEEN START_DATE AND END_DATE
                 AND M3.SYIN_INST = (SELECT   MAX (M4.SYIN_INST)
                                     FROM   METR_TABLE M4
                                     WHERE   M4.HID = M3.HID
                                     AND M4.TYPE IN ('60', '61')
                                     AND M4.REPLY IN ('117', '118')
                                     AND TRUNC (M4.EFF_DT) = TRUNC (M3.EFF_DT))
                 ) TEMP
         WHERE   M1.TYPE IN ('60', '61') AND M1.REPLY = '011'
         AND TRUNC (M1.METR_TRANS_DT) BETWEEN START_DATE AND END_DATE
         AND M1.SYIN_INST = (SELECT   MAX (M2.SYIN_INST)
                             FROM   METR_TABLE M2
                             WHERE       M2.HID = M1.HID
                             AND M2.TYPE IN ('60', '61')
                             AND M2.REPLY = '011'
                             AND TRUNC (M2.EFF_DT) = TRUNC (M1.EFF_DT))
        AND M1.HID = TEMP.HID(+)
        AND M1.EFF_DT = TEMP.EFF_DT(+)
        AND TEMP.HID IS NULL
        UNION
        SELECT   M1.FIRST_NAME,
                 M1.LAST_NAME,
                 M1.HID,
                 M1.EFF_DT,
                 M1.CID,
                 M1.PLAN,
                 (SELECT   MEME.MEME_CK
                  FROM   MEME_TABLE MEME, GRGR_TABLE GRGR
                  WHERE       MEME.M_HIN = M1.HID
                  AND MEME.GRGR_CK = GRGR.GRGR_CK
                  AND GRGR.GRGR_ID = M1.GRGR_ID) MEME_CK
        FROM   METR_TABLE M1
        WHERE   M1.TYPE IN ('60', '61')
        AND M1.REPLY IN ('016', '017', '022', '023')
        AND TRUNC (M1.METR_TRANS_DT) BETWEEN START_DATE AND END_DATE
        AND M1.SYIN_INST = (SELECT   MAX (M2.SYIN_INST)
                            FROM   METR_TABLE M2
                            WHERE   M2.HID = M1.HID
                            AND M2.TYPE IN ('60', '61')
                            AND M2.REPLY IN ('016', '017', '022', '023')
                            AND TRUNC (M2.EFF_DT) = TRUNC (M1.EFF_DT))
        ) TEMP1;
END;
/When posting code if you put {noformat}{noformat} (with the curly brackets and in lower case) above and below it you can preserve formatting.
To your original question...
Is the link name supposed to be variable?
What do you mean by 'hard coding' the values, do you mean in the procedure itself so you'd actually PROC_ENR_ACC();?
What happens when you run this?DECLARE
start_date DATE;
end_date DATE;
BEGIN
start_date := '01-JAN-2009';
end_date := '01-FEB-2009';
PROC_ENR_ACC(start_date, end_date);
END;
I need to know exactly how to recreate your error...
Cheers
Ben
http://www.munkyben.wordpress.com
+Don't forget to mark replies helpful or correct+ ;)

Issue Tracker App Demo

As I'm a newby to htmldb, I'm building the Issue Tracker app demo. Did anyone who has already built the demo have trouble with create/edit Person information on page 4? When I submit, I get the following error:
Error ERR-1025 Error processing PLSQL expression. (:p5_person_role in ('CEO','Manager') and :p5_assigned_project = '%||'null%') or (:p5_person_role in ('Lead','Member') and :p5_assigned_project != '%'||'null%')
OK
ORA-06550: line 2, column 29: PLS-00103: Encountered the symbol "NULL" when expecting one of the following: . ( ) , * @ % & - + / at mod remainder rem and or || multiset
Any assistance would greatly appreciated

Glenn,
Looks like a syntax error in a condition: a missing quote: :p5_assigned_project = '%||'null%')
should be
:p5_assigned_project = '%'||'null%')You'll have to poke around to find where that condition exists (on a region, an item, etc.). Or you could search for it in the export file.
Scott

Dblink poor performance with varchar(4000) after upgrade 11g

Hi
Since a long time we connected from a 10g database via dblink to another 10g database to copy a simple table with four columns. Tablesize at about 500MB
column1 (varchar(100))
column2/3/4 (varchar(4000)).
After the upgrade of the source database to 11g the dblink performance is poor with the big varchar columns. If I copy (select column1 from ) only column1 then I get the data within minutes. If I want to copy the whole table (select column1, column2, column3, column4 from...) then the performance is poor. It didn't finish within days.
Does anyone know about dblink issues with 11g and big varchar columns?
Thank you very much

Use DBlink to pull data from table(s) using IMPDP:
#1 Create DBlink in the Target database
create database link sourceDB1_DBLINK connect to system identified by password10 using 'SourceDB1';
[update tnsnames.ora appropriately]
#2
select DIRECTORY_PATH from dba_directories where DIRECTORY_NAME=’DATA_PUMP_DIR’;
create DATA_PUMP_DIR directory if not exists
#3
impdp system/pwdxxx SCHEMAS=SCOTT NETWORK_LINK=ORCL10R2 JOB_NAME=scott_import LOGFILE=data_pump_dir:network_imp_scott.log
That's all

Objects using dblinks become invalid

I am using numerous functions / procedures / synonyms which point to other databases via dblinks. All was well when the dbLinks / synonyms were public. Last week I was asked to put up a test schema on the same database so I duly made all the public dblinks & synonyms private. All seemed well, everything compiled so I made it live.
Now we get intermittent problems whereby the procedures using the synonyms & dblinks suddenly become invalid. They refuse to automatically compile, causing errors in the applications using the database. SOmetimes they automatically recompile after a time, sometimes they refuse to compile until I do them manually in SQL developer.
anybody have any idea why this should be happening?
Main database is Oracle Enterprise 11.1.0.7 64 bit, connecting to one of same version, one of 10.2.0.4 64 bit. All running on AIX 6.1.

Hi,
Same thing is happening here.
DB A with PROC_SP (Strore procedure)
DB B with 2 users; user TEST1 referencing PROC_SP@A and user TEST2 referencing also PROC_SP@A.
When compiling code on user TEST1 that uses PROC_SP@A, the code on TEST2 becames invalid, and vice-versa.
Did you solve your problem ?
My Database version is 11.2.0.2.0, both DB (A and B) use the same version (Exadata with 11R2)
Thanks in advance.

Issues with Dates not displaying

Hello! I've built a small form with HTMLDB to handle operational tickets/issues/incidents. I've been able to display the data in one page, graph statistics in another but my page where the incidents can be edited has a small issue.
When I click on an edit icon (on pg1) to edit the row, it brings me to page 2 and all the data from that row is displayed EXCEPT for the dates. I've chosen a date picker for that item but in the case where that column was already filled with a date beforehand, it now displays null. All the other data (numbers and varchar2s) display without issue. I can edit the fields and submit the changes and all the changes update correctly (even the dates once new ones are picked) but in some cases I do not want to change that date but display it for informational purposes.
I'm using the TIMESTAMP datatype for those dates by the way.
Thanks,
Thom

Indeed! I used the wizard and had the dates as TIMESTAMP datatypes. I changed the datatypes back to DATE for now and it works perfectly. Another satisfied customer! ;-)
Thanks!

Integrate JSpell with HTMLDB

Hi,
Has any one tried integrate JSpell HTML 2004 version with HTMLDB. I have few issues.
1) Can we setup HTMLDB to use Jar files.
2) JSpell is installed in a OC4J instance on mount point /jspellhtml2k4 on same server. How to access full application with in HTMLDB.
I added a virtual path /jspellhtml2k4 in apache httpd.conf but not successful.
Thanks
S.Anupindi

What is JSpell? If it's a spell checker there is already a built-in spell checker in HTML DB that you can use with HTML textareas.
If it's just a spell checker this sounds like a lot of trouble to use something different from what's already built-in.
Is there any reason that you need to use JSpell?
Earl

Two HTMLDB instances with single Database

please tell me what if I just want to configure and connect an additional apache server to my existing Database. SO that i can get another way for connecting to same databse. ( Mirroring). If this is feasible then please tell me how to configure the apache server. I have tried to configure dad but i think i am not doing it the right way.I am on Oracle 10g with applications made in HTMLDB, I tried to use the dad file for making the connection. During the installation the error message states that " FLOWS_010500 user already exists if you want to continue the installation, please drop the user or specify a different database and try again.. Can any one tell me how can i connect to the Databse with another instance of HTMLDB.
Thank you.

Hi,
Issue Clerifications
1. You have one DB Node and you want install two htmldb application nodes.
2. Which apache servers running on that DB Node.
3. Which listiner ports are there on that DB Node.
4. Are you try to install HTMLDB, What error messages are you facing.
Regards,
Upendran .A
DBA Team

ExecuteUpdate on synonym for table on remote DB

I am trying to insert rows into an Oracle table remotely. The main DB and the remote DB are both Oracle, and we are connecting from the remote to the main DB through a private link. There is a private synonym on the remote DB. I can execute selects against the private synonym without issue. Inserts however, don't work. And, it doesn't throw an exception. It simply doesn't insert the rows.

Yes, that's the weird thing. If I execute the exact same query in SQL plus it works fine. I haven't turned autocommit off. So, it should be commiting. I could try add an explicit commit() to the method. But, my understanding from both the API documentation and the O'Reilly JDBC book is that I shouldn't have to do that.

Select list works for some users not for others

Apex 3.2, Oracle 10g
I have a page with a select list. (other select lists work fine) This view is a based on a view over a dblink with the chain being (public synonym -> dblink view -> dblink real view )
One user logs into the Apex app and the select list populates.
I log in to the app, either through the developer login or the app login, navigate to the same page and the select list does not populate.
What can cause this in Apex or is it a database issue.
Someone said it may be a rights issue, but I thought all Apex users appeared as the same 'user' to the database.
Miscellaneous points: The app authentication is via LDAP/OID and I can get the select list code work in SQL Developer.
Thanks,
Sam

not sure what you are asking
It is a "Select List"
The sql is
select x d, x r
from view
wher x is not null
order by 1
The view is a synonym in the application schema, points to a view on the same database instance (different user) which points to the real view via dblink to a different database instance.

LOVs in Print Friendly Mode

Hi,
I have a page item of type Display as Text (based on LOV, saves state). The page shows the display value ("Pending" in this case). However, when I launch the page in Print Friendly mode, it shows the return value ("2"), instead of the display value ("Pending").
This also occurs for any other select lists on my pages if the select list is read only.
Any ideas or workarounds?
Thanks,
Chaunaci

yes, this does appear to happen for items of type "Display as Text(based on LOV, saves state)", but things seem to work fine for my select list items in the production version of htmldb. i'll log the issue in a bit, but two ways to work around it for now would be to:
1) add an extra item to your page that would render correctly in printer friendly mode. use our built-in condition types of "Current Page Is in Printer Friendly Mode" and "Current page is NOT in Printer Friendly Mode" to determine which/when of your two items should be displayed.
2) tweak your LOV query a little bit to pay attention to the api variable, htmldb_application.g_printer_friendly. it's a boolean that returns true if your app is in printer friendly mode.
if you really are seeing this behavior with select list items, could you please provide more specifics as to how those items were set up?
thanks,
raj
[Edited by: rmattama on Apr 1, 2004 1:26 PM]
i now see the issue w/the select lists occurs in read-only mode. gotcha. didn't read that right the first time. i'll add this to the bug rpt.

Htmldb 1.6 synonym / DbLink issue

Similar Messages

Maybe you are looking for