Paths multiple problem with PowerVault MD3200i + Oracle Enterprise Linux (5,7)

Hello

I'm trying to configure MPIO on a host (running Oracle Enterprise Linux Server 5.7) connected to an Powervault MD3200i. I have a few unexpected problems with/dev/dm-0 missing after reboot.

I configured 3 virtual disks (RAID 1) with MPIO. 3 devices have been created (/ dev/dm-0, / dev/dm-1 and/dev/dm-2).  If I mount them manually, they work very well, so I've added to the fstab. After a reboot, I checked that/dev/dm-1 and/dev/dm-2 have been automatically but is not/dev/dm-0. In fact, the/dev/dm-0 has disappeared from the directory of the device (!)

If I rinse 'multipath f' and ' multipath - v2 "the device gets recreated, but disappears on the"
next reboot again. How can I resolve this behavior? Can anyone help?

Thanks in advance.

See you soon

I thought about it... the problem is that I had the wrong specified dev in fstab.

Should I use the devices in/dev/map /... it worked immediately.

Tags: Dell Tech

Similar Questions

  • Can I use file of udev rules for RAC on Oracle Enterprise Linux 6 11.2.0.3?

    Version of the grid: 11.2.0.3
    Platfomrm: The last Oracle Enterprise Linux 6 (Unbreakable Enterprise Kernel)


    Yes guys Oracle sales have convinced our Manager to go OEL of AIX.
    We plan to install RAC on Oracle Enterprise Linux 6 (UEK) 11.2.0.3.

    I read this thread on the asmlib and udev rules file.
    List of rules for CAR 10.2 udev

    As a benefit to ASMLib, periera Levi said (quotes)
    '+ A good async IO interface to the database, all the I/o interface is based on a model of performance + ASYNC'

    Oracle is not out for 6 OEL or RHEL 6 AsmLib. I think that they will not release asmlib more.

    So, we will use only the udev rules file. Without ASMLib, we will have an ASYNCHRONOUS i/o or other storage related issues?

    Someone at - it successfully installed 11.2.0.3 RAC OEL 6 with udev rules file?

    I use +/etc/udev/rules.d/40-multipath.rules+ udev file to configure the permissions of device of multiple devices on path to e/s for use by o/s user grid.

    It is as follows:

    # modded for setting mpath device permissions
    
    SUBSYSTEM!="block", GOTO="end_mpath"
    
    KERNEL!="dm-[0-9]*", GOTO="end_mpath"
    
    PROGRAM!="/sbin/mpath_wait %M %m", GOTO="end_mpath"
    
    ACTION=="add", RUN+="/sbin/dmsetup ls --target multipath --exec '/sbin/kpartx -a -p p' -j %M -m %m"
    
    # goto labels added
    PROGRAM=="/sbin/dmsetup ls --target multipath --exec /bin/basename -j %M -m %m", RESULT=="?*", NAME="%k", SYMLINK="mpath/%c", GOTO="check_cluster_devs"
    PROGRAM!="/bin/bash -c '/sbin/dmsetup info -c --noheadings -j %M -m %m | /bin/grep -q .*:.*:.*:.*:.*:.*:.*:part[0-9]*-mpath-'", GOTO="check_cluster_devs"
    PROGRAM=="/sbin/dmsetup ls --target linear --exec /bin/basename -j %M -m %m", NAME="%k", RESULT=="?*", SYMLINK="mpath/%c", GOTO="check_cluster_devs"
    
    GOTO="end_mpath"
    
    # set device permissions for Grid Infrastructure
    LABEL="check_cluster_devs"
    RESULT=="*", GROUP="oinstall", MODE="660"
    RESULT=="*", OWNER="grid", GROUP="oinstall", MODE="660"
    OPTIONS="last_rule"
    
    LABEL="end_mpath"
    
    # eof
    

    I understand that it is no longer necessary that can be configured in +/etc/multipath.conf+ as follows:

    ..
    multipaths {
        multipath {
            wwid  SSCST_BIOscst1_sdb_757d30d0
            alias scst1_sdb
            mode 0660
            uid oinstall
            gid grid
          }
    .. etc.
    }
    

    However, has never work properly myself--must do something wrong that I saw a few sources Internet showing it works.

    As far as I KNOW, it is also with OL6 (as it is with OL5 and earlier versions), not a question of AsmLib or udev/MPIO. Udev is a Mapper devices used by the kernel. The decision is whether to add another layer called AsmLib. Personally, I have yet to see any convincing evidence that why should I increase the complexity of the kernel driver stack, adding AsmLib. We also had a few serious incompatibility with AsmLib issues in the past (where he with, caused compatible I/O corruption at the logical level).

    If you look at the clusters in the www.top500.org list, many of them have shared storage accessible via udev/multichemin Pentabytes and nothing else. MPIO has its roots as a driver of kernel for these groups. One of the reasons why we stopped years ago to use EMC Powerpath for our CARS that are wired to EMC San to shared storage.

  • memory usage of Oracle on Oracle Enterprise Linux 5.2

    Hello

    I installed a copy of Oracle 9.2.0.4 db server software (Enterprise edition) on Oracle Enterprise linux 5.2 (32 bit)
    After installing the software, I created two instances db with DBCA.

    Now the question is that, after I start the instance db, after awhile, I see more than 95% of the 8 GB of memory on this server in use.
    The size of sga max of these instances collectively is not more than 800 MB. In fact, even after closing one of the instances of memory
    consumption do not go down significantly because the instance is not assigned too much memory in the first place.

    With the help of TOP does not give a clear picture at all. If I keep the db instance down after restarting the server, consumption of memory is quite normal - less than 500 MB for all internal servers OS treats. SHMMAX is a default value of 4 GB and swap memory occupies about 16 GB. Another point that I want to stress is that I don't see most of the 16 GB memory swap are allocated. Nearly 90% of swap remains free.

    I was wondering how I can solve this problem on a new server.

    Hi usr1298

    I've installed a lot of servers Oracle Linux 2.1, 3,..., 5, and in each of them, I noted that the same behavior. At first when I saw this behavior, I thought something was wrong, but it's the behavior of memory on Linux when Oracle works. I think that default Oracle software reserves all this memory (90%) and that it uses on request but really the software does not use all the memory in a real way in the proceeding does not heavy operations. As I wrote, do not worry about this and worry on this subject when the swap space began to be used more than 20% despite the SGA_MAX_SIZE and the sum of the other parameters of memory does not receive nearly the total amount of RAM.

    Joel Perez
    DBA Oracle
    Members AS

  • Oracle 10g on Oracle Enterprise Linux 6 using VMWare

    I am trying to install Oracle 10 g on a Linux server. The reason is that the system, which I am working 10g and want a test environment on my computer.
    I installed the latest version of Oracle Enterprise Linux 6 on VMWare Player.
    So far so good.
    Now I want to install 10g on the system and have read both Expert Oracle 10 g database Administration and the fast for 10g Release 2 Setup Guide (10.2) for Linux x 86
    I seems to be now stuck in the situation to get all loaded packages that are recommended. Some of the packages are not on the cd, so I'm starting to wonder if it is a valid version for installation.
    Because the version is releast Red Hat Enterprise Linux Server 6.0, I had that with
    # cat/etc/redhat-release

    So 10g is not supported on this version?
    If so do I download the previous version of Linux?
    If so what version and if you can direct me in the sense I would be really happy.

    Currently 6 Linux is not certified to any version of database Oracle.

    Operating system for 10 g 2 (which is what I assume you are using) is listed in the Release Notes - http://download.oracle.com/docs/cd/B19306_01/relnotes.102/b15659/toc.htm#CHDFHIEA

    Srini

  • Application specialist Oracle Enterprise Linux - 1z0-402

    I saw [http://education.oracle.com/pls/web_prod-plq-dad/db_pages.getpage?page_id=351] and [http://education.oracle.com/pls/web_prod-plq-dad/db_pages.getpage?page_id=144] I think I notice a change.

    After having passed 1z0-402 (just now) to get this (Oracle Enterprise Linux application specialist) certification?


    Or certification is available to those within the Oracle partner network?


    Page: [http://education.oracle.com/pls/web_prod-plq-dad/db_pages.getpage?page_id=146] is not clarified this.


    Some certifications and existing reviews seem to be spring upward with this logo attached, clarification would be useful.



    I confess that I feel the pass mark on this review associated with 57% is too low to justify the title of the certification associated without the title of devaluation. I wouldn't think of someone who passed this exam with 58% being sent to me as a specialist of Oracle partner Oracle Enterprise Linux network implementation. Probably a pass mark of 65 to 70% would be reasonable, and I would plump for 67%.

    Published by: bigdelboy on November 29, 2009 06:37
    -A bit of history is on {message identifier: = 3728606}
    -In addition the long awaited ISBN: 978-0071615440 is now scheduled for February 5, 2010. The availability of this book would be the success rate of the exam 1z0-402 too low?

    If you took the review 402 previously, you should automatically receive implementation specialist Oracle Enterprise Linux certification. This certification is available for all candidates and not OPN members only. If you do not receive this certificate within 4 to 6 weeks, please contact [email protected] customer service and they will help with the reception of this certificate.

    With regard to the test score. Results of the review reflect the level of difficulty of the examination. More scrutiny is difficult, more passing grade. This is a very simplified explanation for everything that will determine a test score. There are a lot of mechanics that go into determining the points of overtaking, but a decisive factor is the level of difficulty of the examination.

    Kind regards
    Brandye Barrington
    Certification Forum Moderator
    Manager certification program

  • DB installation on oracle enterprise linux

    Hi guyes,

    in fact I Normand to do some experiments and need to install oracle database 10 G R2 on any version of linux.by the time I heard about oracle enterprise linux.

    I would like to know, what is the difference between red hat enterprise linux 5.2 and oracle enterprise linux to install the oracle database?

    What is the adapted version to install the database? It is in the oracle compatibility matrix?

    Thanks and greetings
    VD
    (forgive me for posting this thread in the wrong place)

    Published by: vikrant dixit on November 19, 2008 20:57

    Everything you did for RHEL5.2 will work perfectly in OEL5.2

    You raise a very good point: the unfortunate fact is that there is * nothing * to all to OEL a house "better" or "easier" for that plain old RHEL Oracle databases. There is no core set environment variable user parameters or pre configured, for example, it would be just for an Oracle straight out of the box installation: you must always you fix all exactly as if it were of RHEL.

    This is normal: OEL was and is intended to be binary compatible with RHEL and therefore no settings of Oracle or parameters that are not in RHEL can be included in the OEL.

    So yes, if it's complicated on RHEL, it is exactly as complicated as on OEL.

    (I'm biased though: If you use my script to doris, you will find that you can install Oracle on virtually any Linux distribution as easily as any other.) See http://dizwell.com/2008/05/01/doris-redux, and if you want to watch a video of how to do to use the script: http://dizwell.com/2008/09/29/proving-a-point/)

  • VMware p2v oracle enterprise linux

    Hi people,

    I need to perform a p2v of oracle enterprise Linux server.

    Can someone please tell me the correct procedure (if no document is available, that would be great ), thanks for the pointers.

    This was better than I expected, I downloaded converter 5.5.1 and it went smoothly, no reconfiguration needed, it installs GRUB, patch mount point entries in fstab and re-create the virtual disk original (all automatically).

    Bravo from VMWare.

  • Oracle Enterprise Linux: Required Package for 11.2.0.3 CARS

    I downloaded Oracle Enterprise Linux 64 Bit in November 2012 (Oracle Linux Release 6, update 3)

    Once I have installed OEL, how can I check if all needed packages required for installation of grid Infrastructure 11.2.0.3 smooth is installed?

    Hello

    the new name of "oracle-validated" under OL6 is oracle-RDBMS-server-11 GR 2-preinstall:

    https://blogs.Oracle.com/Linux/entry/oracle_rdbms_server_11gr2_pre

    Concerning
    Sebastian

  • Installing Oracle 11g R2 on Oracle Enterprise Linux 5.1

    Hello
    May I know if Oracle 11 g R2 can be installed on OEL5.1, if yes, oracle is certified or not.

    Concerning
    Bobby

    Edited by: user12119634 (bobs) 30 Sep, 2010 22:07

    Hi user;

    Its certified

    Linux x 86 platform
    Platform Version Oracle Enterprise Linux 5/Oracle VM
    Product version 11 GR 2
    Certified certification status

    Certified on Oracle Enterprise Linux 5 Update 2 or higher
    See Oracle Database Installation Guide 11g Release 2 (11.2) for Linux for the minimum requirements of the operating system.

    Respect of
    HELIOS

  • Oracle enterprise linux download

    Hi pl tell me where I can download oracle enterprise linux-also oracle for linux

    Hello

    Thanks, but I tried E delivery of oracle previously.
    I am able to check notify his "you will receive mail..." »
    despiting waiting for two days, I have received no acknowledgement of delivery e oracle.

    Your credentials have been selected and sent to download e-Delivery Web site software. If you get an e-mail within 2-3 days after the date of the registration, you can contact "Oracle E-delivery of Customer Service" by email ([email protected]).

    Kind regards
    Hussein

  • Oracle enterprise linux and oracle 10g?

    Hey;

    I am fairly new to try to download oracle and get installed on a linux machine. I made a few attempts on linux fedora core 10 and home edition of windows vista that does not go very far. Since I just passed my exams for ocp via a boot camp (yay for me!), I'm hoping to get a couple of linux boxes operational so that down the concepts by the practice.

    From what I read, it seems that I can download enterprise linux, put it on a system and then download oracle 10g and install this w/little or no difficulty (at least not expected...)

    Am I reading that right or did I miss something? Any info greatly apprecaited.

    Doug O'Leary

    Hello

    You can download enterprise linux and 10g and go there. Pre-packaged for the components required to install the oracle Enterprise linux software, here is a link for further reference see cross-platform compatibility.

    http://www.Oracle.com/technology/support/MetaLink/index.html

    Concerning
    OrionNet

    Published by: OrionNet on June 8, 2009 17:00

  • Oracle Application Server 10g R2 install Oracle Enterprise Linux 5

    Hello...

    I want to install Oracle Application Server 10g R2 on Oracle Enterprise Linux 5. Can I get an installation instructions (or not) why?
    Oracle installation guide mentions Suse 9 & 8 and Red hat 3 & 2. There is nothing related to Oracle Enterprise Linux 5 or 4?

    Thank you

    See Note: 564174.1 on MetaLink.

    Thank you
    Shail

  • EBS 11i install on Oracle Enterprise Linux 4

    Hi, friends:

    I want to install EBS 11i of Oracle Enterprise Linux 4, and I am looking for the installation guide for that, Cloud someone give me some advice?

    Thanks in advance.

    Ronnie

    PL see ML notes 316806.1

    Srini

  • Impdp problem with disconnection of Oracle (without recovery)

    Hi all

    I am trying to import a data dump generated from an Oracle Enterprise Manager 10.2.0.4.0 on a server unix with the following command:
    expdp mc/XXX DIRECTORY = DUMPFILE EXPORT_DIR = SLTN-MCEXPORT-2010-07 - 22.dmp PATTERNS = MC LOGFILE = SLTN-MCEXPORT-2010-07 - 22.log

    I want to import into oracle XE 10 g (10.2.0.1.0) on a windows server. That's why I do the following:

    Impdp mc/xxx DIRECTORY = DUMPFILE EXPORT_DIR = SLTN-MCEXPORT-2010-07 - 22.dmp PATTERNS = MC LOGFILE = SLTN-MCIMPORT-2010-07 - 22.log

    I had to manually create TABLESPACE MC_DATA: 'MC_DATA' of CREATE TABLESPACE DATAFILE 'D:\OracleHisRuns\Dumps\SLTN-MCEXPORT-2010-07-22.dmp ';

    Then a table fails MC. CURVEPOINTS

    And finally we are disconnected from Oracle

    The determined type SCHEMA_EXPORT/TABLE/INDEX/INDEX processing
    UDI-00008 operation error 3113 and 3114 error.

    and I can't get back. I again install Oracle XE.

    Anyone know why this is happening? Is it a memory problem? Or is there a problem when the user, the tablespace is created next to me?
    Maybe the export and import between these two Oracle database versions are not compatible?

    Any help would be greatly appreciated.

    Many thanks in advance,
    Thomas

    Please find the journal that I have:

    D:\COSMOS_BATCH_RUNNER\Automation_oracle\Scripts\IMPORT DMP > impdp mc/mc DIRECTOR
    Y = DUMPFILE EXPORT_DIR = SLTN-MCEXPORT-2010-07 - 22.dmp PATTERNS = MC LOGFILE = SLTN-MCEXP
    ORT-2010-07 - 22.log

    Import: Release 10.2.0.1.0 - Production on Thursday, July 22, 2010 17:35:36

    Copyright (c) 2003, 2005, Oracle. All rights reserved.

    ConnectΘ α: Oracle Database 10g Express Edition Release 10.2.0.1.0 - Production

    Master table 'MC '. "' SYS_IMPORT_SCHEMA_01 ' chargΘe/dΘchargΘe with succΦs
    DΘmarrage of 'MC '. "" SYS_IMPORT_SCHEMA_01 ": mc / * DIRECTORY = DUMP EXPORT_DIR
    FILE = SLTN-MCEXPORT-2010-07 - 22.dmp PATTERNS = MC LOGFILE = SLTN-MCEXPORT-2010-07 - 22.lo
    g
    Treatment of the determined type SCHEMA_EXPORT/USER
    ORA-31684: the USER object type: 'MC' is dΘjα
    The determined type SCHEMA_EXPORT/SYSTEM_GRANT processing
    The determined type SCHEMA_EXPORT/ROLE_GRANT processing
    The determined type SCHEMA_EXPORT/DEFAULT_ROLE processing
    The determined type SCHEMA_EXPORT/TABLESPACE_QUOTA processing
    Treatment of the determined type SCHEMA_EXPORT/PRE_SCHEMA/PROCACT_SCHEMA
    Treatment of the determined type SCHEMA_EXPORT/SEQUENCE/SEQUENCE
    The determined type SCHEMA_EXPORT/TABLE/TABLE processing
    Treatment of the determined type SCHEMA_EXPORT/TABLE/TABLE_DATA
    ORA-31693: failed loading/dΘchargement of the donnΘes of table object 'MC '.
    'CURVE_POINTS '; the object is ignorΘ due to the error:
    ORA-02354: error during export/import of donnΘes
    ORA-02368: the file is not valid for this load operation
    . . import: "MC". "' BLOCKORDER_INTERVAL_RESULTS ' 46,54 MB 2280571 lines

    . . import: "MC". "' BLOCKORDER_INTERVALS ' 19.54 MB 1359850 lines

    . . import: "MC". "' BLOCKORDER_HEADERS ' 6,162 MB 149683 lines

    . . import: "MC". "' BLOCKORDER_RESULTS ' 246805 lines 6,205 MB

    . . import: "MC". "' PRELIMINARY_RESULTS ' lines 4,247 58248 MB

    . . import: "MC". "' ALGORITHM_EVENT_LOG ' lines 1,689 MB 20388

    . . import: "MC". "' CURVE_HEADERS ' 1,816 MB 73568 lines

    . . import: "MC". "' LINE_RESULTS ' lines 1,937 58248 MB

    . . import: "MC". "' OS_HISTORYSTEP ' lines 2,155 MB 21460

    . . import: "MC". "' SESSION_ALGORITHMPARAMETERS ' 1,741 MB 57656 lines

    . . import: "MC". "' LINE_CAPACITIES ' lines 1,311 46883 MB

    . . import: "MC". "' MSG_TRANSMISSION_LOGS ' 6464 lines 1,143 MB

    . . import: "MC". "' FINAL_PRICE_POSITIONS ' lines 396,4 16428 KB

    . . import: "MC". "' OS_CURRENTSTEP ' 7456 lines 348,4 KB

    . . import: "MC". "" REPORTATCVALUES "254,5 Ko 12000 lines

    . . import: "MC". "' BUSINESS_STEP_SESSION_REPORTS ' 220,7 lines 2852 KB

    . . import: "MC". "' PTDF_MATRIX_FACTORS ' lines of 239,6 12612 KB

    . . import: "MC". "' SESSION_BIDDINGAREAS ' 207 KB 2648 lines

    . . import: "MC". "' ALGORITHM_DETAILED_LOG ' 526 lines 52,85 KB

    . . import: "MC". "' ALGORITHM_SOLUTION_LOG ' 607 lines 55,23 KB

    . . import: "MC". "' AUDIT_LOGS ' 69,25 lines 1516 KB

    . . import: "MC". "' BUSINESS_STEP_REPORT_COMMENTS ' 62,46 lines 2821 KB

    . . import: "MC". "' OS_WFENTRY ' 45,38 lines 1020 KB

    . . import: "MC". "' PTDF_MATRIX_CONSTRAINTS ' 3153 lines 59.75 KB

    . . import: "MC". "' SESSIONS ' 679 KB 61,64 lines

    . . import: "MC". "' SESSION_LINES ' 77,58 lines 2648 KB

    . . import: "MC". "' ALGORITHM_PARAMETERS ' lines 7,937 84 KB

    . . import: "MC". "" ALGORITHM_SESSION_LOG "317 lines 19.20 KB

    . . import: "MC". "' BALANCING_AREAS ' 6,320 KB 1 lines

    . . import: "MC". "' BIDDING_AREAS ' 10.03 KB 4 lines

    . . import: "MC". "' The LINES ' 7,140 KB 4 lines

    . . import: "MC". "' LINE_CAPACITIES_STATIC ' lines 7,601 96 KB

    . . import: "MC". "" MAPPING_CONSTRAINTID_ROWID "171 lines 11,17 KB

    . . import: "MC". "' OUTPUTCHECKS ' 419 lines 37.96 KB

    . . import: "MC". "' PART ' 6,257 KB 9 lines

    . . import: "MC". "' PERIPHERAL_SYSTEMS ' 8,851 KB 9 lines

    . . import: "MC". "' PTDF_MATRIX_HEADERS ' lines 7,195 72 KB

    . . import: "MC". "' SESSION_BALANCINGAREAS ' 662 lines 32,71 KB

    . . import: "MC". "" SESSION_REPORTS "135 lines 29,89 KB

    . . import: "MC". "' SESSION_REPORT_COMMENTS ' 128 lines 7,007 KB

    . . import: "MC". "' SYSTEM_PARAMETERS ' 27 lines 6.5 KB

    . . import: "MC". "" USERS "33 lines 11,71 KB

    . . import: "MC". "" USER_SCRATCH_CARD_DETAILS "11 lines 11.51 KB

    . . import: "MC". "" BLOCKORDER_EXCLUSIVEGROUPS "0 KB 0 rows

    . . import: "MC". "" BLOCKORDER_LINKS "0 KB 0 rows

    . . import: "MC". "" MATRIXROWRESULTS "0 KB 0 rows

    . . import: "MC". "" MSG_TRANSMISSION_ERROR_LOGS "0 KB 0 rows

    The determined type SCHEMA_EXPORT/TABLE/INDEX/INDEX processing

    UDI-00008: the opΘration a gΘnΘrΘ the ORACLE 3113 error
    ORA-03113: end of file on channel communication


    UDI-00008: the opΘration a gΘnΘrΘ error ORACLE 3114
    ORA-03114: not connectΘ α ORACLE

    DMP D:\COSMOS_BATCH_RUNNER\Automation_oracle\Scripts\IMPORT >

    Hello
    I think you're confusing the creation of storage with the expdp dumpfile. You tried to add the dumpfile to the database as a data file, and then it was corrupted.

    What you need to do is add a new space of tables with a brand new file - nothing to do with the dumpfile.

    And then perform the import with the dumpfile - you will probably need to this regenerate your source system.

    See you soon,.
    Harry

  • Problem with NVL in Oracle 10 g version 10.2.0.1.0

    Hello

    I am facing a problem with NVL. I have two configurations one with version 10g 10.2.0.1.0, UAT and other with 10.2.0.3.0.
    I defined a function (SQL written below). Both configurations have the same defined function and I'm passing the same value/setting as well. But, I get different results on both configurations.
    Pls let me know the reasons... .to this strange behavior of function, I think it's because of the NVL

    Function
    CREATE OR REPLACE FUNCTION Fn_Dt_Get_amnt (NUMBER of N_EXP_COVERED_AMT,
    V_SRC_SYSTEM_ID varchar2,
    NUMBER OF N_UNDRAWN_AMT_LCY
    NUMBER OF N_EXP_LIMIT_NCY
    NUMBER OF N_CCF_PERCENT
    NUMBER OF N_EXP_AMT_LCY
    p_undrawn_amt varchar2)
    + RETURN VARCHAR2 AS +.

    + UNDRAWN_AMT NUMERIC (22.3). +

    BEGIN


    IF P_UNDRAWN_AMT IN ('EXP_COVERED')

    THEN
    IF (UPPER (V_SRC_SYSTEM_ID) IN ("UBS"))
    THEN
    IF (N_EXP_AMT_LCY (case when NVL(N_UNDRAWN_AMT_LCY,0) < 0 +))
    THEN 0
    of other NVL(N_UNDRAWN_AMT_LCY,0)
    (END) * NVL(N_CCF_PERCENT,0)) < N_EXP_COVERED_AMT
    THEN

    UNDRAWN_AMT: = (N_EXP_AMT_LCY (+)
    -case when NVL(N_UNDRAWN_AMT_LCY,0) < 0
    THEN 0
    OF OTHER NVL(N_UNDRAWN_AMT_LCY,0)
    (END) * NVL(N_CCF_PERCENT,0));
    ON THE OTHER
    UNDRAWN_AMT: = N_EXP_COVERED_AMT;
    END IF;


    ELSIF (UPPER (V_SRC_SYSTEM_ID) ("VP", "FW", "FN"))
    THEN

    IF NVL(N_EXP_AMT_LCY,0) (CASE WHEN (NVL(N_EXP_LIMIT_NCY,0)-NVL(N_EXP_AMT_LCY,0)) < 0 +)
    THEN 0
    ELSE (NVL(N_EXP_LIMIT_NCY,0)-NVL(N_EXP_AMT_LCY,0))
    END) * NVL(N_CCF_PERCENT,0) < N_EXP_COVERED_AMT


    THEN

    UNDRAWN_AMT: = (NVL(N_EXP_AMT_LCY,0)(+)
    CASE WHEN (NVL(N_EXP_LIMIT_NCY,0)-NVL(N_EXP_AMT_LCY,0)) < 0
    THEN 0
    ELSE (NVL(N_EXP_LIMIT_NCY,0)-NVL(N_EXP_AMT_LCY,0))
    (END) * NVL(N_CCF_PERCENT,0));

    ON THE OTHER

    UNDRAWN_AMT: = N_EXP_COVERED_AMT;
    END IF;

    END IF;


    ELSIF P_UNDRAWN_AMT IN ('NO')

    THEN
    IF UPPER (V_SRC_SYSTEM_ID) IN ("UBS")

    THEN

    UNDRAWN_AMT: = (CASE WHEN NVL(N_UNDRAWN_AMT_LCY,0) < 0
    THEN 0

    OF OTHER NVL(N_UNDRAWN_AMT_LCY,0)
    (END);

    ELSIF (V_SRC_SYSTEM_ID) UPPER IN ("VP", "FW", "FN")
    THEN
    UNDRAWN_AMT: = (CASE WHEN (NVL(N_EXP_LIMIT_NCY,0)-NVL(N_EXP_AMT_LCY,0)) < 0

    THEN 0
    ELSE (NVL(N_EXP_LIMIT_NCY,0)-NVL(N_EXP_AMT_LCY,0))
    (END);
    END IF;

    END IF;

    RETURN UNDRAWN_AMT;
    END Fn_Dt_Get_amnt;

    Setting run

    Oracle 10g 10.2.0.3.0
    SELECT FN_DT_GET_AMNT(1002872,'FW',699737.78,752154,0.5,52416.22,'EXP_COVERED')
    the double - power - 402285.11

    Oracle 10g 10.2.0.1.0
    SELECT FN_DT_GET_AMNT(1002872,'FW',699737.78,752154,0.5,52416.22,'EXP_COVERED')
    the double - power - 1002872

    I think the problem is with NVL, such that the code works fine with Coalesce on the two configurations. Can someone explain why?

    It is a bug, as evidenced by the test case I posted.
    Apparently when multiple NVLs are used with the CASE in the same expression.

    Works fine in the following versions:
    10.2.0.4
    10.2.0.3
    9.0.1.1.1

Maybe you are looking for