Linux versus windows

I am a new DBA and caught in between the choice of OS(Linux/Windows) to use. I have just concluded my university education and will be in the job market soon. I need to know which OS I should learn to better suite oracle database.

I absolutely agree with Jaffar, especially from Oracle point of view. Just remember that Linux is a bit more difficult to learn than Windows, but once you know Linux it's easier to learn Windows than the opposite.
If you start learning Linux, I'd just advice you one thing : use graphical interface as less as possible, at least at the beginning. Graphical interface simplify your life, but you don't know what you're doing...
Anyway, in my opinion you should learn both : Linux is the best solution as database server, while Windows could be advisable as Client. Windows is surely the most popular Client in the world.

Similar Messages

  • Linux 10g versus Windows XP 10g Issue - ORA-07445: exception encountered

    I did a full database export using exp from a 10g Linux database:
    Oracle Database 10g Enterprise Edition Release 10.2.0.2.0 - Prod
    PL/SQL Release 10.2.0.2.0 - Production
    "CORE     10.2.0.2.0     Production"
    TNS for Linux: Version 10.2.0.2.0 - Production
    NLSRTL Version 10.2.0.2.0 - ProductionLinux version:
    Red Hat Enterprise Linux ES release 4 (Nahant Update 5)I installed Oracle 10g on Windows XP:
    Oracle Database 10g Enterprise Edition Release 10.2.0.1.0 - Prod
    PL/SQL Release 10.2.0.1.0 - Production
    "CORE     10.2.0.1.0     Production"
    TNS for 32-bit Windows: Version 10.2.0.1.0 - Production
    NLSRTL Version 10.2.0.1.0 - ProductionWindows XP version is:
    System:
        Microsoft Windows XP
        Professional
        Version 2002
        Service Pack 3
    Intel(R) Core(TM)2 CPU
    6300 @ 1.86GHz
    1.86 GHz, 3.25 GM or RAMI performed a full database import after creating the required tablespaces from the 10g database on Linux to Windows.
    On the Windows 10g database, I set Maximum SGA size to 1300 MB, and Total SGA size to 1000 MB. I set Aggregate PGA Target to 500 MB.
    This Windows box is going to be used by at most 2 developers for a change to our application that is not backward compatible (with respect to the database) with earlier versions of the application that uses the database. I was asked to create this so we could develop this branch without affecting the rest of the development team that are working on other enhancements on the Linux box.
    I then ran some packages that refresh the data in a set of tables. This runs through some code that exercises a lot of the database so I could uncover any issues.
    At a specific point in the execution of these packages I get the following error:
    ORA-07445: exception encountered: core dump [ACCESS_VIOLATION] [_qkkIsOJKey+299] [PC:0x1E58F73] [ADDR:0x80] [UNABLE_TO_READ] []I read the trace file and isolated it to a single, somewhat complex query. I tore the query apart into smaller queries, trying to isolate what was causing the issue. Whenever the core dump occurred, I was able to recover by simply closing the connection and opening a new connection. In my testing I found the specific query that raises the exception, and a less performing version of that query that does not cause the exception. Please note that this exception has never occurred on the Linux box (nor has it occurred on the beta or production database), and the code has been in place for close to 2 years.
    The code that causes the issue:
    SELECT UNIQUE
           oc.ogc_case_id AS oc_ogc_case_id,
           ogc.ogc_number AS ogc_number
      FROM (SELECT SUBSTR (REPLACE (afd.field_text, '-', '' ), 1, 6 )    AS ogc_number
              FROM ewoc_hw_snap.activity_field_data afd
             WHERE afd.afdcd_activity_field_data_key = (SELECT afdc.activity_field_data_key
                                                          FROM ewoc_hw_snap.activity_field_data_codes afdc
                                                         WHERE afdc.name = 'Case Number'
           ) ogc
      LEFT OUTER JOIN lct_snap.ogc_cases oc
            ON oc.ogc_number  = ogc.ogc_numberThe code that returns the same data, only slower, and does not raise the exception:
    SELECT UNIQUE
           (SELECT oc.ogc_case_id FROM lct_snap.ogc_cases oc WHERE oc.ogc_number = ogc.ogc_number ) AS oc_ogc_case_id,
           ogc.ogc_number AS ogc_number
      FROM (SELECT SUBSTR (REPLACE (afd.field_text, '-', '' ), 1, 6 )    AS ogc_number
              FROM ewoc_hw_snap.activity_field_data afd
             WHERE afd.afdcd_activity_field_data_key = (SELECT afdc.activity_field_data_key
                                                          FROM ewoc_hw_snap.activity_field_data_codes afdc
                                                         WHERE afdc.name = 'Case Number'
           ) ogcI executed both queries in SQL Developer version 3.0.04. SQL Developer gives the error message: No more data to read from socket
    Any ideas?
    The query shown above is as simple as I could pare it down to and still raise and then fix the issue. The fix to the package is not as simple as the second query shown above since there is more than one column of data that I need to return from the lct_snap.ogc_cases table. In any case, I shouldn't have to forgo using the LEFT OUTER JOIN. There must be something wrong with the Windows database. I didn't want to patch the Windows database without first checking on this forum to see if anyone had any less intensive potential solutions.
    I searched and found so many posts with the ORA-07445: exception encountered: core dump [ACCESS_VIOLATION] error, that I realize it's a somewhat generic error message.
    Edited by: Starlight Rider on Jun 2, 2011 10:01 AM

    Check this metalink note : 452951.1
    Regards
    Raj

  • Deploy on Linux or Windows?

    Hi,
    Is there any difference in performance, stability and memory-consumption (leaks?) between Forms Server on Linux versus Forms Server on Windows?
    Thanks.
    Regards,
    Stephan

    I had to do this to get RAP (standalone with Jetty) to run in cygwin. I
    hope & assume this means it will run on Linux (and Mac?)
    (1) I copied bundle org.eclipse.osgi_3.3.2.R33x_v20080105.jar to my root
    (2) To the root directory, I added a Linux-specific config folder, which I
    call config-linux. I added a config.ini file inside this folder, with
    this text:
    osgi.bundles=plugins/org.eclipse.equinox.common@2:start,plugins/org.eclipse.update.configurator@3:start,plugins/org.eclipse.rap.ui@4:start,plugins/org.eclipse.equinox.http.jetty@4:start
    # do not run an eclipse application
    eclipse.ignoreApp=true
    # don't shutdown osgi after exiting (or not running) the application
    osgi.noShutdown=true
    # incoming port for the server
    org.osgi.service.http.port=7070
    (3) In the root folder, I created run.sh, with this text:
    java -Dorg.osgi.framework.bootdelegation=* -jar
    org.eclipse.osgi_3.3.2.R33x_v20080105.jar -configuration config-linux
    -consoleLog -debug -clean -dev -console
    (4) So to run the app on cygwin or Linux, you go to the root folder and
    type
    sh run.sh
    and this runs the application without any SAXException errors
    David Donohue

  • [solved] Video conversion problem: Mencoder on Arch versus Windows XP

    I have a cheap portable video player which (according to a supplied Windows software configuration file) requires any video to be converted (say from a YouTube .flv) using a command line like:
    mencoder myfile.flv -ofps 15 -vf-add scale=160:128 -vf-add expand=160:128:-1:-1:1 -srate 44100 -ovc xvid -stereo 0 -xvidencopts bitrate=300:max_bframes=0:quant_type=h263:me_quality=0 -oac lavc -lavcopts acodec=mp2:abitrate=128 -o myfile.avi
    If I do the conversion on Linux - my fully-upgraded Arch - the resulting .avi file has the audio and video out of sync when played with Mplayer (on Linux and Windows), and no audio at all when played with Vlc (on Linux and Windows). But if I do (supposedly) exactly the same conversion using Mencoder on Windows, the resulting .avi file plays perfectly on everything on both Linux and Windows (and also the portable video player).
    Here's the version information for the Mencoders being used:
    Windows: MEncoder Sherpya-MinGW-20060312-4.1.0 (C) 2000-2006 MPlayer Team
    Arch: MEncoder SVN-r32492-4.5.1 (C) 2000-2010 MPlayer Team
    I'd be grateful for any ideas as to how to make my Mencoder conversion work as well on Linux as it does on Windows XP! (Or is there something wrong with the way I'm playing the file with Mplayer and Vlc on Arch?!)
    Last edited by ninian (2010-11-16 23:28:02)

    Aha! I think I've cracked it, but using ffmpeg instead, with options prompted by this blog post:
    ffmpeg -i infile.flv -r 15 -vcodec mpeg4 -vtag XVID -s 160x128 -aspect 4:3 -qmin 3 -qmax 5 -ar 44100 -ab 128k -ac 2 -acodec mp2 -g 1 outfile.avi
    This seems to work everywhere, and the writer specially notes that:
    The most important option is the -g 1 to force every frame to be a keyframe
    which may well be where my conversions were going wrong. So I'll mark this as 'solved' even though mencoder is still a mystery, but I have learned to keep Vlc up my sleeve, so thanks for that tip again.

  • [SOLVED]Arch Linux, and Windows 8.1 Dual Boot issue

    Hi guys. I recently bought a new laptop, and decided to run Arch Linux and Windows 8.1. I installed Windows 8.1 first as recommended by the beginners guide, and then installed Arch Linux. I made sure UEFI was enabled in my BIOS, and made sure everything was on a GPT partition. The install itself went fine. My laptop loads grub, and Arch Linux shows up and boots, but there is no option to boot into Windows 8.1. The only way I can boot into Windows 8.1 is by accessing my motherboard, and choosing to launch the windows boot manager instead of the grub launcher. How can I get Windows 8.1 to show up in grub? Thanks in advance guys.
    Last edited by Painguy (2014-08-20 20:19:37)

    Painguy wrote:
    -----------Edit------------------
    Yeah just ran this after installing os prober
    grub-mkconfig -o /boot/grub/grub.cfg
    and it works now. Sorry for the crazy posts guys. Thanks for the suggestions and help.
    Yeah I'm sure I'm doing something wrong lol. I did not install OS-Prober. I think that's where the issue is. I'm missing the menu entry in grub.cfg right? Do I just run os prober or do I have to manually edit grub.cfg.
    If that's not the case then here is what I did up until now. What I did was start with a blank SSD drive. I used cgdisk to make the disk use a GPT partitioning scheme. I installed windows 8.1 and checked in windows to make sure it boots into UEFI-GPT mode and it does.  I then installed arch linux using the beginners guide. I made sure to use cgdisk to create any extra partitions, and installed GRUB to the efi partition that windows had created ( I did not format this partition ). I then finished the installation and rebooted, and Grub only shows arch linux. I don't think I'm dealing with anything MBR related. I made sure I used the GPT partitioning scheme....i think lol.
    All you need to do is run os-prober then run grub-mkconfig. It should pick up Windows at that point.

  • How to set owner-only access file permissions both on Linux and Windows

    Hi everybody.
    I have the following problem. I need to store some private user information in file system. So I need to set owner-only access permissions for some directory in user home. I did not find API for doing this. As I understand this is platform specific thing. Could anybody tell me how can I do this both on Linux and Windows?
    Thank you in advance.

    More ideas just came back to me (of something I did before)
    Each OS requires a different control.
    For Windows you will need to set the .policy file for each user (a pain if you have more than a dozen of clients) granting access (R or RW) to files/directories.
    Then you have to write some native code (access via JNI) to access/modify your authorization.
    A more professional solution that I used was J-Integra (other similar tool exist).
    Basically it is a bridge between Java and COM. It magically give you access to the COM API from Java (and vice versa).
    In my case it was to read/write/modify Winword documents transfered between PC and server.
    From memory you have to run the com2java.exe to build the bridge (classes in Java language) between the 2 families. Compile those classes and you have a "driver" for the COM (Microsoft Objects).
    Here I must confess my ignorance on Microsoft.
    Those COM objets must give you some easy/integrated access to the API for LDAP or Microsoft Active Directory Service (for centralize control with more complex setup).

  • Dual boot kali linux with windows 8

    i have ideapad y510p and i want to dual boot kali linux with windows 8 
    i want a full guide to do it please !!help!!!

    Hello,
    if you got problems with the kernel, you can try backbox-linux. The kernel there is newer. It is based on Ubuntu 12.04 LTS.
    Have a nice day
    thoschi
    LENOVO IDEAPAD Z710 59403383,i7-4700MQ,8 GB DDR3-RAM,1.920 x 1.080 Pixel,NVIDIA® GeForce® GT 740M,Hybrid (1 TB S-ATA + 8GB SSD)

  • Oracle DB sessions on Linux and Windows

    Actually I am confused the way session memory is allocated on 32 bit windows and linux OS.
    Windows is a thread based architecture where all user sessions are treated as threads of oracle.exe process. The addressable space is 4GB and oracle.exe can use up to 3GB in case you use /3GB switch in boot.ini file. It means that combined size of SGA ,PGA and other memory structures can not be more than 3GB. When a session gets connected with a database a PGA is allocated. It means that memory required by database sessions also comes from the 3GB limit. If your memory usage goes beyond 3GB you wont be able to connect new sessions with database or you would require to reduce the size of one of your SGA components.
    Linux is a process based architecture where each user session is treated as a separate process and has its own addressable space. Even on 32bit linux the addressable space is 4GB i.e. 2GB for process and 2GB for kernel. It means we need to accommodate our SGA and PGA in the 2GB limit (there are ways to increase the size of SGA but let assume we are doing nothing to raise the 2GB limit).
    As PGA is allocated for every session when it connects with database.
    My question is how PGA memory is allocated against a database session on linux.
    To further clarify my question I am stating the results of an experiment that I did on linux and windows.
    On windows I set the total size of SGA to 1730M and PGA aggregate target to 1024M. I was able to able to startup the database instance without any issue. Then I started connecting sessions with the database and even ran a query doing a heavy sort in one of the sessions but after connecting 10 sessions with DB I started getting ORA-12500 TNS Listener failed to start a dedicated server process error on windows.
    Then I did the same experiment on Red Hat Linux. SGA size was 1633MB and where as PGA aggregate target was only 10MB. I was unable to start instance when I set the SGA size more than 1633MB.
    With that setting I was able to connect more than 200 sessions with database inspite of running query doing heavy sort in 4 sessions without getting ORA-12500 error. I was even able to connect more than 400 sessions with database after setting PGA aggregate target to 1GB on Linux.
    Now my confusion is why linux was able to handle more DB sessions. I know it’s a process base model and each session is a separate process and has its own addressable space but as PGA is allocated against every session and PGA memory comes from 2GB limit then why PGA memory did not exhaust on linux. Is PGA allocated differently on linux as compared to windows?

    Perhaps the problem is easier to understand if you leave PGA_AGGREGATE_TARGET out of this issue completely.
    The out of memory problem you had, has nothing to do with PGA_AGGREGATE_TARGET setting directly. The difference in number of processes comes from architectural differences in how Oracle uses memory in windows vs unix.
    On windows the SGA + ALL PGAs must fit into one 4GB address space (in practice even smaller as youve noticed). This is because there's only one process, oracle.exe. One process = one address space.
    On Unix, you have many processes and every process can use portion of their own address space for PGA.
    So, on windows (32bit) you run out of address space for the single oracle.exe process. On unix you dont hit this limit just because the number of processes grows (but if some process wants to use a lot of PGA memory, then they eventually run out of address space (talking about 32bit again))
    On unix you probably ran into another limit - amount of total usable virtual memory (RAM+swap) available in your system
    Tanel Poder
    http://blog.tanelpoder.com

  • Which operating system: Linux or Windows Server 2003?

    Hi,
    it doesn't matter which operating system we install an our server.
    What are your experience with Linux and Windows?
    Which System do you recommend regarding performance issues?
    best regards
    Felix

    Linux has a higher per process memory limit. The fmscore process can use up to 2 GB on Windows and up to 3 GB on Linux.
    The performance increases in Flash Media Server 3.5 are roughly equivalent on Linux and Windows.
    HTH,
    Jody

  • Error in Transport Tablespace from linux to windows

    I am testing the cross-Platform Transport Tablespace. As per the oracle, we can transport tablespace from linux to windows without conversion because both are using same endian (Little).
    But i am fail to do Transport Tablespace from Linux to Windows.
    I am performing Transport Tablespace process as following:
    from Source Oracle Database server(red had linux as 4 32-bit oracle version:10.2)
    Sql> alter tablespace TEST read only;
    $ expdp system/pass dumpfile=test.dmp directory=export_dir transport_tablespaces=test transport_full_check=y
    after this i am coping test.dmp and data file (test.dbf) to the target machine (ms windows xp 32-bit with oracle 10.1) .
    on Target Machine (with Ms windows xp os) here i am giving the following command:
    impdp system dumpfile=test.dmp directory=exp_dir transport_datafiles=/exp_dir/test.dbf
    but it is giving following error:
    ora-39001: invalid argument value
    ora-39000: bad dump file specification
    ora-31619: invalid dump file "c:\pks\1103.dmp"
    what may by ...
    Prabhaker

    now for version compatibility i am inclusding version option with expdp
    edpdp scott dumpfile=1103.dmp directory=pks transport_tablespaces=prabhu version=10.1.0.2.0
    but now it is giving following error:
    Import: Release 10.1.0.2.0 - Production on Saturday, 11 March, 2006 19:07
    Copyright (c) 2003, Oracle. All rights reserved.
    Connected to: Oracle Database 10g Enterprise Edition Release 10.1.0.2.0 - Production
    With the Partitioning, OLAP and Data Mining options
    Master table "SCOTT"."SYS_IMPORT_TRANSPORTABLE_01" successfully loaded/unloaded
    Starting "SCOTT"."SYS_IMPORT_TRANSPORTABLE_01": scott/******** DUMPFILE=1103.DMP DIRECTORY=PKS TRANSPORT_DATAFILES=C:\PKS\PRABHU version=10.1.0
    Processing object type TRANSPORTABLE_EXPORT/PLUGTS_BLK
    ORA-39123: Data Pump transportable tablespace job aborted
    ORA-06550: line 2, column 2:
    PLS-00306: wrong number or types of arguments in call to 'BEGINIMPORT'
    ORA-06550: line 2, column 2:
    PL/SQL: Statement ignored
    Job "SCOTT"."SYS_IMPORT_TRANSPORTABLE_01" stopped due to fatal error at 19:07
    regards
    Prabhu

  • Linux or windows hosting?

    I am publishing my site on iweb via godaddy- i have a choice to use linux or windows for hosting- which one?

    Rhonda Ferguson wrote:
    I am publishing my site on iweb via godaddy- i have a choice to use linux or windows for hosting- which one?
    Linux

  • Linux or windows?   godaddy

    I'm looking at the godaddy hosting plans, and it asks Linux or Windows-based, not giving a choice for Mac OS. I presume that it still works with Mac. Which one should I select? I intend to make pages with iWeb on a mac, and then host them outside of .mac.

    Linux. Windows is just potential for trouble.
    You can find some tips here:
    http://homepage.mac.com/thgewecke/iwebserver.html

  • Linux or Windows

    I have the option to use either multiple Linux or Windows
    servers.
    These servers will be dedicated purely to the media server.
    Any opinions on which would be better for live streaming in
    the most part.
    Alan

    I wouldn't consider running a production database in a non-supported configuration. That's a bit like walking a tightrope without a net. If you can't purchase a supported Linux configuration, I'd strongly suggest going with Windows.
    If you can purchase a supported Linux configuration, there would be two criteria I'd use to determine which O/S to use. First is the ease of scripting. For 99% of the people out there, Linux's scripting facilities are orders of magnitude better than those in Windows. Given how many DBA type tasks ought to be automated, this gives a strong nod to Linux. The other criteria would be the organization's familiarity with each O/S. If you have a bunch of Windows wizards, go with Windows. If you have a bunch of Linux/ Unix gurus, go with Linux.
    Justin

  • Is possible restore backup full from linux to windows?

    Is possible restore backup full from linux to windows, I need know if is possible.
    There are any document about it?

    Hi
    Yes because They are in little endian.
    http://apunhiran.blogspot.com.tr/2009/03/how-to-moveconvert-database-from-linux.html
    Regards

  • Unable to set focus for Linux / X-Windows ($$$ for fix...)

    It appears whenever a modal window is opened on Linux X-Windows the focus is stolen from the window and you CANNOT use the keyboard to return to the form - let alone set the focus on the textbox.
    This is not just for my modal forms but for something as fundamental as JOptionPane or simple code like this:
    JFileChooser chooser = new JFileChooser();
    if (chooser.showOpenDialog(parentComponent) == JFileChooser.APPROVE_OPTION) {
    From trolling through posts I can see no answers which is really sad for Linux good for Microsoft I guess. How can we deploy on Linux if you are forced to get carpal tunnel clicking on every form with the mouse ?
    I would be willing to pay some consulting dollars to get this resolved.
    Greg

    Hm, I've developed quite a bit for X11 and linux, but I never noticed that effect. Can you post a SSCCE that reproduces the issue, just to be sure it is not a side-effect from your code?
    It this is a persistent problem, you might want to make your dialog re-request the focus programmatically:
    myDialog.setVisible(true);
    SwingUtilities.invokeLater(new Runnable() {
      public void run(){
        myDialog.requestFocus();
    }See if this does anything. If not, the problem is most likely not coming from swing.

Maybe you are looking for