XML Insertion - Windows To Linux

I have a JScript executable that parses an input file containing data in PVL format and uses the DOM to create an XML record, which the script then writes to an output file. The script, input and output files all reside on a local Windows XP machine with Internet Explorer 6.0.
I need the script to insert the output file into a table that is defined against an XML schema. However, the 10.2 database resides on a Linux x86 box (OS is Red Hat Enterprise Linux Advanced Server), not the Windows client.
The script can successfully connect to the database (thanks to ODBC) and it is successfully inserting non-XML data into other tables. However, for the XML file, I'm trying to use the following insert statement (the db table's name is AIP_XMLTYPE and the XML output file is type1_aip.xml):
INSERT INTO AIP_XMLTYPE VALUES (XMLTYPE
(BFILENAME('LOCAL_XMLDIR','type1_aip.xml'),
nls_charset_id('AL32UTFS')));
LOCAL_XMLDIR is a directory defined in the database for the local C drive where the XML file resides. (I suspect this is the problem. Does a defined directory have to be on the database server?) The error the script returns is
[Oracle][ODBC][Ora]ORA-22288: file or LOB operation FILEOPEN failed. No such file or directory.
When I manually upload the XML file onto the Linux server and execute the insert statement in SQL*Plus (using a defined directory on the server, of course, instead of LOCAL_XMLDIR), it is successful, so I don't believe there is an issue with how the XML schema is defined nor with the format of what's in the XML file.
Is there another way I could insert the XML record? Is there another way to get the database to recognize the local directory? Could I have the script upload the XML file to the db server and then do the insertion?
Any suggestion is appreciated.

Hi,
You may can get some ideas from the case posted at http://www.oraclepoint.com/topic.php?filename=98&extra=page%3D1
Please register and then log in to get the case doc because it's only available for members.
By the way, I don't think the cross-platform issue will impact the implemtment of XML loading to database. I do hope the case study above helps.

Similar Messages

XML source file path (LINUX vs. WINDOWS)

Hello, we are sourcing data from XML for a project. RPD development is being done in Windows and then the RPD is getting deployed to the Linux env.
How can i make sure the path translate from Windows to Linux?
Windows path can be C:\myobieedata\region.xml
How can this be translated into Linux path?
Your help is appreciated.
Regards,

Just edit the connection pool to use the valid location of the files in Linux and copy that RPD over to the linux box. You can even edit it 'on-line' on the Linux environment and make the change there.
As an example I'd check the sample sales connection pool that was configured when you installed OBIEE on Linux and observe the difference to the sample sales connection pool when you installed on windows.

Transaction behavior changed with change in Windows to Linux Environment

Hi All,
I am facing a very weird problem.
I have a 2 BPEL Synchronous Services lets say A & B. A calls B in its flow and B inserts a row into database and then retrieves the inserted row (as the primary key is sequence generated through a Trigger).
This flow was working perfectly when:
·     It was on windows Development Box with SOA SUITE and DB Server on the same machine.
·     All the DB adapters had jca:Address information in it.
Problem:
·     After moving them to a Unix box with a remote DB Server, I am seeing the problem that retrieval is happening before commit of insert and failing.
Point to remember:
·     Insertion, selection, assignment all are happening in the different ‘SCOPE’
·     I have tried even putting them in a single scope but error remained same.
·     I have even tried putting a wait between them, so it’s a plain “Transaction not commited “issue
I am expecting it could be an issue with the way Data source has been created or something else.
Please suggest your pointers on this.
Thank you.

Hi Win
1. There is absolutely NO changes or specific requirements to deploy a portal EAR created from Windows and deploying on Unix or Linux Env. We do this very often. I export my entire portal applicatoin (EAR, WAR and dataSync if its there) from Workshop or Eclipse (10.3.2) IDE as .EAR file. Then we can deploy this .EAR file pretty much in any Portal Domain on any Env like Linux. So first I want to clarify that there is NO need to change any files or do steps to deploy a portal EAR from Windows to Linux Env.
2. Now coming to your error, the root cause seems to be this: java.lang.ClassNotFoundException: org.apache.beehive.netui.pageflow.PageFlowContextListener
a) On Windows did you test your EAR on a Portal Domain. I assume you may have already did this.
b) Is the domain that you created on the Linux Env and its version matches exactly with what you had in Windows Env. Say in Windows if you have WLP 10.3.1, then on Linux also you should have WLP 10.3.1. If not, you need to upgrade your portal EAR first in Windows to 10.3.x and then deploy.
c) Compare the config.xml file modules specially the portal framework shared modules. They should be same in both Windows and Unix except for the location ofcourse.
d) On Windows, by any chance did you deploy any extra additional modules in the portaldomain. If so do the same on Linux Env also.
e) Other then config.xml file, all the other details are within the .EAR file itself like weblogic.xml, weblogic-application.xml file etc etc.
f) So better compare you config.xml file in both the working and not working Envs.
g) Also on Linux, are you deploying on Single Server domain or cluster domain. If cluster, make sure you deploy your portalEar only to Cluster because none of hte shared modules are deployed to Admin. So deploy only to cluster.
Thanks
Ravi Jegga

Libraries on Windows to Linux SuSe

Hi everybody,
There are a thousand forms which have 4-5 attached libraries. These libraries are located in a library directory C:\LIBS.
I want to transfer these forms to Linux SuSe environment and libraries as well...
Except for the compilation and save of the forms and libs , how can I alter 'massively' - if it is possible - the location of the lib directory (let's say from C:\LIBS to /opt/libs).
I think of two ways:
1) transforming into xml , make the neccessary modification and return back to fmb
2) transforming into fmt , make the neccessary modification and return back to fmb (is this possible..?)
3)which else?
Which is the quickest way to alter the path of the forms 10.2. on Windows to Linux SuSe?
Thanks , a lot
Simon

Hi Francois,
I cannot , at least in Windows, convert the .pll files to .plx . Error FRM-91507 : Cannot create the lib and another message PDE-IIN005 : command insert is unknown appear.
That's why i have installed the .pll into a specific library directory C:\LIBS.
Then , imagine that i'll make the executable of the libs..
The files(.fmb) in Linux will still have the 'wrong' path C:\LIBS in their properties , won't they..?
Unfortunately , anyway i cannot test it right now....
Thanks a lot for your constant support and ideas....
Simon

File Upload Problem from Linux to Windows and Windows to Linux

Hi,
I am a newbie in the flex environment.
I have faced a problem for file upload from linux to windows and windows to linux
And I have put crossdomain.xml file in root folder, still this problem is appear.
Please help me, if anybody know the answer.
Thanks and Regards,
Senthil KUmar

Hi Pauley,
I thought you are moving from unix to windows. Now I got it.
Make sure that the folder is shared and have necessary permissions. Please see few suggestions here:
Re: NFS or FTP for file on Ip 10.x.y.z
Regards,
---Satish

Sqlplus in Windows and Linux display same data diversely

Database 10.2.0.1.0 on Linux.
NLS parameter from nls_database_parameters:
NLS_CHARACTERSET = AL32UTF8
NLS_NCHAR_CHARACTERSET = AL16UTF16
Environment variable:
NLS_LANG=ESTONIAN_ESTONIA.AL32UTF8
When I insert characters like öü from sqlplus they appear correctly.
But if I try to insert character ä (test is table with one varchar2(19) column)
insert into test values('ä');I recive:
ORA-ORA-01756: quoted string not properly terminated I don't understand why can't I insert that character.
When I use sqlplus on my Windows machine I see ?? instead of characters öü which I inserted successfully in Linux.
On Windows I'm able to insert characters öüä into table test. I can aslo quey them and see correctly. But if I query them from Linux sqlplus then I see just one starnge caracter ā instead of öüä.
On Windows I have NLS_LANG=ESTONIAN_ESTONIA.BLT8MSWIN1257
Can somebody help me with that issue. I'm very confused. I would like to be able to insert and see character öüä from both environments.

A          DUMP
öüä        Typ=1 Len=8 CharacterSet=AL32UTF8:
e2,80,9d,c2,81,e2,80,9e
k??k       Typ=1 Len=4 CharacterSet=AL32UTF8:
6b,f6,f6,6b
??         Typ=1 Len=2 CharacterSet=AL32UTF8: f6,fc
?          Typ=1 Len=1 CharacterSet=AL32UTF8:
f5Before you worry about the output/display aspect, you have illegal data stored in the db. E.g. there is no single-byte code points F5, F6 or FC in UTF-8.
This is probably due to having char set part of NLS_LANG incorrectly set. In the Linux system, are you sure it is setup to produce utf-8? I think the Linux system uses a single-byte Iso Latin encoding.
What code page are you using on the Windows side?
Also you have (in the first row above) special Windows characters right double quote marks (U+201D) and double low quote marks (U+201E), plus a control "character", not defined in Windows-1257. The Windows special characters does not exist in the closest-matching Iso char set (Latin-4?).
How are you inserting the data? What does the users use to enter the data?
Message was edited by:
orafad

Copy of a CD (created by Disk Utility) not recognized by Windows or Linux

I created a copy of AutoCAD 2005 (not pirating it... just wanted a backup copy of the install CD).
I used Disk Utility (File > New > Disk Image From Folder) to successfully create a DMG file on my MacBook. I could mount the DMG and browse all the files on it.
I then burned a CD (Disk Utility > select DMG > click burn), which mounts fine on OSX. However, my Windows and Linux boxes will not recognize it (while they recognize the original CD just fine).
I tried again, this time setting the file format to DVD/CD master. It made a CDR file, which I burned to CD. Got the same result: Windows and Linux don't recognize it while OSX does.
So instead I burned the files to a CD ala drag and drop using Finder. This time the CD was recognized by Windows and I used it to install AutoCAD.
However, my understanding is that doing it this way won't work for a bootable CD, so I'd like to figure out why Disk Utility wasn't able to make an exact copy.
Any thoughts? Problem with copy protection? Problem with the CD's file system?
Details about the optical drive in my MacBook (1st gen 2.0GHz, 2GB RAM):
MATSHITA DVD-R UJ-857:
Firmware Revision: HBE4
Interconnect: ATAPI
Burn Support: Yes (Apple Shipped/Supported)
Cache: 2048 KB
Reads DVD: Yes
CD-Write: -R, -RW
DVD-Write: -R, -RW, +R, +RW
Burn Underrun Protection CD: Yes
Burn Underrun Protection DVD: Yes
Write Strategies: CD-TAO, CD-SAO, DVD-DAO
Media: No
MacBook 1st gen CoreDuo 2.0GHz, 2GB RAM Mac OS X (10.4.8)

Yeah OS X's Disk Utility doesn't burn using the ISO 9660 file system while Finder does (I could swear I made a copy of an MS Office CD with Disk Utility, but maybe I'm remembering wrong). Anyway, Disco (a program I got in a MacHeist bundle) will make an exact copy of CDs. Anyone looking for a free solution, check out LiquidCD and burn, which may have the ability to copy Windows CDs (appears to be on their lists of features, but I haven't tried them myself).

Sync iPod to Mac, but store files on Windows and Linux

I am an advanced user on both mac and widows system. I had my old 3rd Gen ipod where it was in FAT32 format and would sync to a Win machine. Then I could use the ipod to transfer data (ala external harddrive) on windows and mac platforms.
So now I have a 5th Gen Ipod and Itunes on my G5. Today I wanted to use the ipod to move some files between my work windows machine and home to my mac. To make a long story short the new ipod is formatted to be hfs+ journeled system, which windows can't read.
So I figure the easy solution is to reformat the ipod to be win-dos format, let it put the ipod software on, dump the library back and I should have a 5th gen with FAT32 that can be read by windows (or linux). So here's the rub
I used Disk Utility in OSX to partition the ipod several times. the first time I just specified ms-dos format (or whatever the disk utility calls a windows partition). All went well until the updater could not put the software on the new ipod claiming that it had to be in mac format. Then I tried to split the drive and parition into a large mac/hfs partition (for all the itunes files) and a small windows partition for the ipod (for data file storage). The problem is that the updater kept selecting the fat32 partition for installation (because I had to choose master boot record in order to be able to put the windows parition on there. at least this is what I suspect was happening). Two partitions woudl mount on the desktop, but only one would control the ipod.
So why not just let the windows parition be the one partition that had the ipod control software on it, you say? Because then I'm back to the problem that the iPod Updater won't deal with updating the windows partition.
Any ideas on a work around?

My concern is that when I have to update the iPod (in the future), will the windows based updater software be required?
Only if you want to continue to use it on Windows machines.
If you want to use the Mac Upddater, it will not be read by Windows.
If I do what you are suggesting, will the ipod return to being a resident mac based sycing device even though the first update/install of the ipod software was on a windows machine?
No. You can use it on both/either Mac and WIndows.
This is why I thougth some sort of partition with space set aside for the windows partition would be the workaround?
The iPod Updater formats the entire disk, not a partition.

How to best partition the HDD for dual boot: Windows 8 & Linux

Hi,
I'm a newbie in Windows 8, linux, Partition....
I would like to use my new laptop with windows 8 pre-installed to the following:
Create a Dual Boot Windows 8 / Ubuntu
Create on my C drive (unique drive on my laptop):
*a partition for the system
*one for Windows 8
*one for Data and
*last one for Ubuntu OS
My aim is :
-to install VmWare on Windows 8 and run Checkpoint Splat with virtual machines to perform CCSA labs.
to install GNS3 on Ubuntu for CCNP labs but with GNS3 accessing files in the Data partition
I have a Toshiba Satellite with 750Gb and 8 Go RAM.
According to you what's the best size for my partitions?
Is there a better way to partition my C drive regarding my aim?
Best Regards
***I don't know if it's the right section, sorry for that. Feel free to move this thread in the appropriate section***

Hi,
Thank you all,
Isn't it too much 200G for Windows as i will on ly install VMWARE Workstation 9 for lab purposes and possibly a few softwares.
VMWARE installed on Windows will require 2 GB + 1 GB per host (i will use approximatively 15 hosts).
Checkpoint products will be installed on VMWARE that is installed in Windows :
- disk space needed for 3 Checkpoint gateways: 3*37 GB
- disk space needed for 2 Checkpoint gateways management servers: 2*10 GB
On Linux I'll install GNS3 with 10 GB disk space required and a few softwares.
I've got a Toshiba stellite computer with Windows 8 pre-installed.
Processor: Intel(R) Core(TM) i5-3210M CPU @2.5 Ghz
RAM: 8 Go
64 bits OS, processor x64
See below the screenshot
I've taken screenshot of my partition (below).
* Don't take into account the disk 1 (E) partition
Can u confirm that in Disk o, partitions with 450 Mo, 260 Mo and 9.48 Go are for recovery system?
I am planning to let the recovery partition as it is and divide the TI30985500A (C) partition (688 Gb) into:
- 200 Gb windows partition
-100 Gb Linux partition
-388 Gb data partition
What do you think about this partionning regarding the informations i've provided above?
Greets

Executing WINWORD from both Windows and linux using Java

Dear All,
I have a problem when trying to use Runtime class to execute a winword document. My code is as follows.
Runtime r= Runtime.getRuntime();
try{
r.exec("rundll32"+" "+"url.dll,FileProtocolHandler"+" "+"WINWORD"+ " "+"file://C://welkom.doc");
catch(Exception ex){
System.out.println("Unable to open the doc file:"+ex);
When I execute this swing application with the above, the program is unable to open the specific file. And it says "Problem with the short cut, unable to open the file". When I run the program without the file name, it opens a new doc file.
Attn: what I want to do here is I need to open the winword file from both windows and linux. So I cannot use the relative path of the WINWORD.exe file.
Can anyone help me?
Thanks in advance.
Regards,
Bala

If you take a look at the api, there's an exec(String[]) method. This is the one you should use when calling a command with multiple parameters.
so something like...
String[] cmd = new String[] {"rundll32", "url.dll,FileProtocolHandler", "WINWORD", "file://C://welkom.doc"}
r.exec(cmd);

10gR2 to 11gR2 from Windows to Linux with 6TB Database.

Hi,
I need to do migration from 10gR2 to 11.2.0.2.0 and from Windows to Linux for 6TB database Size. As both Endian format same, I did the migration using RMAN Convert database on UAT but for same 10gR2 version.(Both Source on Target Oracle Version was 10.2.01.0). Now I have to migrate from 10gR2 to 11.2.0.2.0. So it's 2 way process.
is it possible to club these 2 way migration process ???
Are following steps right or any modification is required ??
1) Install 11.2.0.2.0 on Linux OS
2) Open Source Database (Windows) in Read Only. Generate TRANSPORT.SQL, CONVERTSCRIPT.RMAN & Pfile.
3) Copy Source datafiles in temporary location on Target OS. Also copy above files.
4) Create dummy controlfile.
5) Startup nomount with Changed pfile
6) Mount database with Dummy Controlfile.
7) Run CONVERTSCRIPT.RMAN to convert actual datafiles from Windows Platform to Linux Platform.
8) Shutdown the instance and delele Dummy Controlfiles.
9) Edit TRANSPORT.SQL and do Appropriate Changes like redolog files, datafiles path...
Now my question is, can I add ?/rdbms/admin/catupgrd.sql in TRANSPORT.SQL script after Startup upgarade command so that Complete Upgradation to 11.2.0.2.0 would be done.
10) Then run TRANSPORT.SQL on Target System (11.2.0.2.0)
Tell me if more inputs are required from my side...
Regards,
Manish

This forum deals with migrations from foreign databases to Oracle.
You might better post your request in this forum:
General Database Discussions

Migration 10g to 11gR2 from Windows to Linux with 6TB DB Size in one step

Hi,
I need to do migration from 10gR2 to 11.2.0.2.0 and from Windows to Linux for 6TB database Size. As both Endian format same, I did the migration using RMAN Convert database on UAT but for same 10gR2 version.(Both Source on Target Oracle Version was 10.2.01.0). Now I have to migrate from 10gR2 to 11.2.0.2.0. So it's 2 way process.
is it possible to club these 2 way migration process ???
Are following steps right or any modification is required ??
1) Install 11.2.0.2.0 on Linux OS
2) Open Source Database (Windows) in Read Only. Generate TRANSPORT.SQL, CONVERTSCRIPT.RMAN & Pfile.
3) Copy Source datafiles in temporary location on Target OS. Also copy above files.
4) Create dummy controlfile.
5) Startup nomount with Changed pfile
6) Mount database with Dummy Controlfile.
7) Run CONVERTSCRIPT.RMAN to convert actual datafiles from Windows Platform to Linux Platform.
8) Shutdown the instance and delele Dummy Controlfiles.
9) Edit TRANSPORT.SQL and do Appropriate Changes like redolog files, datafiles path...
Now my question is, can I add ?/rdbms/admin/catupgrd.sql in TRANSPORT.SQL script after Startup upgarade command so that Complete Upgradation to 11.2.0.2.0 would be done.
10) Then run TRANSPORT.SQL on Target System (11.2.0.2.0)
Tell me if more inputs are required from my side...
Regards,
Manish

The directory structure is different...that's why you need to create the tablespaces in Linux. e.g. export dump says create tablespace in D:\..., there is no such thing in Linux. You can just export it, install software, create the db, tablespaces in Linux and import into the 10g in Linux.

File Path for Windows and Linux

hi can any one give solution how to implement code for odi file copy .
i am using Same Code for Windows and Linux
OdiFileCopy "-FILE=#Var_Lookup_File_Path \ Employee.csv" "-TOFILE=#Var_Lookup_File_Path_Backup\Employee.csv" "-RECURSE=NO" "-OVERWRITE=YES" "-CASESENS=NO"
the problem of above code is
it is working in windows but my Prod is Linux environment.
in linux \ not works.. / will work
please suggest how to concatnate variable and file without' \' '/'

Thanks Phani
i think in OS command also we may give path like #variable\File Name.
So how it will replace \ with / in Linux.
Could you give code for my Scenario..
Thanks for helping

Jdev 11g pre3 Preforms on Windows contra Linux

Have been working a lot with the Jdev 11g, think it is nice and easy to work with.
Except after a little time (or bigger project) it gets really slow, takes like 30sec, when changing a property and tabbing out. (If you look a the memory usage it runs around 2-4 GB, with only Jdev running plus standard Windows stuff)
Have now been trying on:
XP prod on a Dual Core Pc with 2 GB ram
Win 2003 Server – server with 8 gb ram
Now I tried to work a little with it on a Kubuntu (running on a virtual Pc 2007 (1 GB ram), and here it seem to a much small issue with the slowdown and with the memory usage.
Should there be a different between running Jdev 11g on a Window or Linux PC?
Have anyone tried to work a longer period on a Linux PC with JDev?

Hi,
note that the IDE performance issue is mentioned in the release notes of JDeveloper 11. It also mentions that this is an area that developers are looking at in a release version TP3. Performance should not vary too much between OS.
Frank

Native WIndows or Linux on XServer Hardware

Does anyone know if I can run Windows or Linux on the XServe natively without Leopard?
-Thanks

Hello Test Bed Fred, and welcome to the AppleBoards,
Assuming you are asking about Intel based current or recent Xserves you can run Windows using virtualization like Parallels Server BME, but you can't simply install it on the machine. You can also install and run a number of *nix flavors virtualized if you want to do that.
I'm sure there are *nix installs that can be run directly on the metal - the EFI boot is the only real technical hurdle that needs to be overcome. All that being said unless the machines are free I wouldn't generally choose to do this because you're paying Apple for software you won't use to get the hardware, but you can do whatever you want.
HTH,
=Tod

XML Insertion - Windows To Linux

Similar Messages

Maybe you are looking for