C++ Call Interface :: OCCI And Valgrind - Memory Leak
Aug 21, 2012
I wrote a simple OCCI program and ran valgrind with it. I got some memory leak in the program. I tried to search around but couldn't get any answers. I have tried to compile the program using gcc version 4.5, and icpc version 12.0.4 and they both gave the same errors. The OS I am using is RedHat Linux 6.1.
int main() {
oracle::occi::Environment *theOCCIEnvPtr = NULL;
[Code] .......
and one of the leaks is :
==9744== 32 bytes in 1 blocks are definitely lost in loss record 61 of 224
==9744== at 0x4A07279: malloc (vg_replace_malloc.c:263)
==9744== by 0x62D9147: slzsetevar (in /apps/oracle_client/libclntsh.so.11.1)...
We have a occi application which is linked dynamically to occi lib on Solaris. Now during build we linked it with occi of Oracle 10.2.0.4 client but we want the same build to be able to run under Oracle 10G client as well as Oracle 11G client. But it seems that occi is not forward not backward compatible as name of dynamic occi lib has a version in it.
I install Oracle instant client 11g in my Windows 7(32bit), and use MSVC6++ to test program occidml.cpp, compile is no problem, but when I run it, it crash on create Connection. I have tried to link with different version of OCCI library ( v8,v9), but all crash on create Connection.
I'm dealing with an ORA-1000 error in a Pro*C application where all the cursors are correctly closed (or so it seems to me).
Here is the code for a simple program which reproduces the problem:
Each cursor is opened in a PL/SQL package:
CREATE OR REPLACE PACKAGE emp_demo_pkg AS TYPE emp_cur_type IS REF CURSOR; PROCEDURE open_cur(curs IN OUT emp_cur_type, dept_num IN NUMBER); END emp_demo_pkg;
[Code]....
While testing the initialization parameter open_cursors is set to 50.
It's my understanding that Oracle doesn't close the cursors until it needs the space for another cursor, which in my test case seems to happen when I enter a value of 50 or bigger for "number of loops". To see how oracle is reusing the cursors, while the test program is running I run SQL*Plus and query v$sesstat for the session that's running the test with the following sentence:
select name, value from v$sesstat s, v$statname n where s.statistic# = n.statistic# and sid = 7 and name like '%cursor%';
Even before I enter a value for number of loops I can see that the session opened 4 cursors and closed 2 of them:
NAME VALUE ---------------------------------------------------------------- ---------- opened cursors cumulative 4 opened cursors current 2
Entering a value of 5 for number of loops yields
NAME VALUE ---------------------------------------------------------------- ---------- opened cursors cumulative 11 <----- 7+ opened cursors current 8 <----- 6+
With a value of 30
NAME VALUE ---------------------------------------------------------------- ---------- opened cursors cumulative 36 <----- 25+ (apparently, Oracle reused at least 5 cursors) opened cursors current 33 <----- 25+
With a value of 47
NAME VALUE ---------------------------------------------------------------- ---------- opened cursors cumulative 53 <----- 17+ opened cursors current 50 <----- 17+
Now I reached the upper limit set by the initialization parameter open_cursors.
Entering a value of 48, I get the ORA-1000 error.
ORA-01000: maximum open cursors exceeded ORA-06512: at "SCOTT.EMP_DEMO
Since I open and close the cursor in the same loop iteration, I expect to find in every iterarion 1 explicit cursor and a number of implicit cursors (the PL/SQL call along with the so-called recursive cursors), but I don't expect the sum of all of them to be greater than 50. If my understanding is correct Oracle should be reusing the 50 cursors previously marked as "closeable", not raising the ORA-1000 error.
I am completely new to anything with Oracle or OCI,
Also, basic information:
Language: C Oracle DB version: 11g
Trying to use OCI
I am trying to write a client that connects to an Oracle database and gets some information (obviously). On the OCILogon() step, I need the username, password, and db address. However, someone said there will be a tnsnames.ora file, and that should be enough.I've searched online and I have not been able to find anything that relates OCILogon() to tnsnames.ora, so here are my questions:
1) Can OCILogon() use tnsnames.ora at all? 2) Is there any other function that I should use instead of OCILogon() that would work with the tnsnames.ora
Actually i am not using OCI directly. I have a Delphi application in which I would need to clone an open db connection. Not only username, password, and target database (this is easy and i already do that).
What I would need is also inherit context variables and global temporary tables content status at the moment in which the cloning operation takes place. It seems a very difficult, if not impossible, thing to achieve.
I am using OCI and instant_client (11.2.0.3.0) in my C++ program (on Windows 7) to connect remotely to 11G database. While doing a connection using server name as well as servername:port/service_name as connection strings, I am getting error of ORA-12504 TNS:listener was not given the SERVICE_NAME in CONNECT_DATA. According to instant client FAQ & white paper at [URL]..., no tnsnames.ora is required for instant client. Then why could I be getting this error? I ensured that only the instant client directory is in Path environment variable. There was a ODBC driver installed on this same host.
We have code that occasionally errors with ORA-12151, with tracing indicating some funk surrounding OCI calls to OCIPing(), OCIBreak(), and OCIReset(). What the code is trying to do here is release a connection from a pool (our implementation, not OCIPool) and before releasing it trying to make sure nothing is actively running on the connection etc. The environment is created in THREADED mode (ie not non-blocking).
Reading the documentation its not really clear to me if OCIBreak() and OCIReset() are really intended to be called while in THREADED mode.
Is it a bad idea to call break/reset as part of a general connection-reset routine?
The basic control flow (minus error handling etc) is:
I am trying to fetch the data from system table "FLOWS_020100.WWV_FLOW_ACTIVITY_LOG1$" in Oracle Express edition 10.2.0.3 database, using the OCI library on Windows with C++
This table has a TIMESTAMP column of date type.
I've the following query as below:
select TIME_STAMP from FLOWS_020100.WWV_FLOW_ACTIVITY_LOG1$ where TIME_STAMP > to_date('31/OCT/12 23:59:59', 'DD-MON-YY HH24:MI:SS') order by TIME_STAMP asc;
For this, first I am preparing the query as below:
select TIME_STAMP from FLOWS_020100.WWV_FLOW_ACTIVITY_LOG1$ where TIME_STAMP > :PKVAL order by TIME_STAMP asc;
Before calling OCIExecute(), I am setting the PKVAL buffer to following value:
I want to monitor the stream activity with strmmon tool. How to compile the strmmon.mk? I was seeing with visual studio command prompt but it does not work properly.
In Oracle 10.2 installation there are OCI headers in rdbms/public directory: 10.2.0/rdbms/public/ocidem.h 10.2.0/rdbms/public/ocidfn.h 10.2.0/rdbms/public/ocikpr.h 10.2.0/rdbms/public/oci.h 10.2.0/rdbms/public/oci1.h ...
What should be installed additionally to have OCI headers in rdbms/public in Oracle 9.2.0?
I am having an issue to connect my C++ application to my Oracle DB. When my code calls
oracle::occi::Environment::createEnvironment(),
I get an error "ORA-01804: failure to initialize timezone information".
The application works on other people's Windows machine with their Oracle DB. But it doesn't work from my machine connecting to their Oracle instance. So I guess the issue does not come from my instance but rather from my Oracle dlls maybe?
I am in the process of converting my skill in oracle and this time PRO*C from Windows to Linux. I have oracle 11g R2 installed on a UBUNTU (12.04) server and have installed the instant client as described in
[URL]
This has a query which will run in SQLPLUS but fails with PRO*C
:~/Projects/proc/proctest1$ proc INAME=proctest.pc SQLCHECK=NONE ONAME=proctest.c LNAME=proctest.lis LTYPE=long Pro*C/C++: Release 11.2.0.3.0 - Production on Mon Apr 22 21:00:18 2013 Copyright (c) 1982, 2011, Oracle and/or its affiliates. All rights reserved. System default option values taken from: /home/neill.rutherford/instantclient_11_2/precomp/admin/pcscfg.cfg PCC-W-02109, SQLCHECK=NONE is no longer supported.. using SYNTAX
I work with oracle 10g, on windows 7. sometime when i try to start my sqlplus in prompt command this message display.
C:UsersSony>sqlplus SP2-1503: Unable to initialize Oracle call interface SP2-0152: ORACLE may not be functioning properly
But the solution people have proposed into the forum are:
1 - check the environment variable Oracle_home and modify if it ending by ""
2 - or try tho run my promt command as Administrator.
I try the both, but only the second works. while now i don't understand why it works when i start the promt command as Administrator and why it no works when i don't run it like administrator. Another thing is i am the only user of my computer, i have only one session (administor session).
In my application, when I passed a Type Array as parameter to the stored procedure and when the Type Array is large enough(normally 3000 type objects， each objects has about 15 attributes), the OCIObjectSetAttr performance deteriorates large enough from below 1 millisecond to 1~3 miliseconds.
when i try to connect with sqlplus or toad then i get an error
Microsoft Windows [Version 6.1.7600] Copyright (c) 2009 Microsoft Corporation. All rights reserved. C:Users>sqlplus SP2-1503: Unable to initialize Oracle call interface SP2-0152: ORACLE may not be functioning properly C:Users>
and i want to tell what i do before it-my all files were be converted in .FLT then i shorted out this problem after that when i tried to connect with sqlplus or toad then i get above error.and i have restarted my system but throwing same error.
I'm using an existing stored procedure in my C code. The stored procedure in question has been compiled and is proven to work without any errors. However, when I use the same in my C code, its failing with the above error.
The Store procedure definition looks like :
CREATE OR REPLACE FUNCTION SP( srq_id integer , unid IN SPkg.arr_parmid, parm_typ IN SPkg.arr_parm_typ,
[code].....
Type definitions
TYPE arr_parm_typ IS TABLE OF char INDEX BY BINARY_INTEGER; TYPE arr_parmid IS TABLE OF tbl_parm.UNID%TYPE INDEX BY BINARY_INTEGER; TYPE arr_parm_lbl IS TABLE OF tbl_parm.PARM_LBL%TYPE INDEX BY BINARY_INTEGER; TYPE arr_parm_vlu IS TABLE OF tbl_parm.PARM_VLU%TYPE INDEX BY BINARY_INTEGER; TYPE arr_vlu_hint IS TABLE OF tbl_parm.VLU_HINT%TYPE INDEX BY BINARY_INTEGER;
NAME TYPE VALUE ------------------------------------ ----------- ------------------------------ lock_sga boolean FALSE pre_page_sga boolean FALSE sga_max_size big integer 3G sga_target big integer 2G
from what I read I beleive this will initially grab 2GB of memory on startup and will grab up to to 3GB of memory total for the SGA. The "total" memory can be allocated to different peices of the SGA when needed but will never exceed 3GB. Is this correct or would these settings infringe on any available memory on a system that is already tight on memory?
Secondly, what happens if both these values are set to the same value?
Can i know the internal process of initialization of DB into memory in timesten , when a new connection is establishing? Will timesten create tables and indexes in RAM when first connection is established if the RAM policy is default?
want to know the internal functional flow of timesten when any command is fired against it.
I have a confusion with MEMORY_TARGET and MEMORY_MAX_TARGET parameter. if i set SGA_TARGET, SGA_MAX_SIZE along with MEMORY_TARGET and MEMORY_MAX_TARGET then how oracle will manage the memory? Because as per my understanding if we set MEM
We have implemented PeopleSoft with an Oracle database. What I need to do is provide clients with access to information stored in that database. Some additional information may come from other external databases, however for right now I am just trying to crawl, not run.
I am working on getting table information from the PS provider, so that I can understand what information is available. I am also working with the clients to determine if there are pieces of information that they need access to which are not currently native to the PS implementation. I need to identify these for step 2 of the process. My superiors were looking at using an SOA architecture using SOAP to move the data. I am open to suggestions and since the project is in its infancy, I have time to alter perceptions before it is too late.
I am doing a research for our next project.MIGRATING oracle forms6i to java/apex/adfi've gather documentations to present to my boss.but she wants an actual(hands on) migration without purchasing yet.
application that i can test?Is FORMS to .XML to JAVA? is that possible ?where topic my message belongs to.(that's why i've put in GENERAL topic)
have created or used a JavaBeans can interface with OpenOffice?
I modified the java classes and finally managed to create a document. Odt OpenOffice Writer and launch from Eclipse. If I am not using the libraries oooBeans but there are 2 problems now:
1. For known issues I had to set the Java class path of the folder where you have OpenOffice executables and should be changed ...
2. I created a. Jar file executable, but I do not know how to integrate Oracle Forms and use it as JavaBeans. Perhaps this step you're more prepared
I have a problem in oracle forms 10g, as a following: Arabic language show as special character for example(????,!!!!!!,....),I use now charcterset AMERICAN_AMERICA.WE8ISO8859P1 ,I cannot change this charcterset on my database Because I am afraid that the old data may be lost
Note: I am use now oracle form 6i without no problem in Arabic language and when I intended to converted to oracle form 10g ,the language problem has come back, and I was try to change nls_lang to several charcterset charcterset(Arabic_Egypt.AR8MSWIN1256 ,......) but still show in special character character like (????,!!!!!!,....)
