Physical Machines vs LPARs

Hello,
In order to install ECC6 on AIX machine (Production instance), is it better to have both ECC instance & DB instance on the same machine without LPARs configuration, or to have each instance on a separate LPAR?
Thanks for advice.
Waleed

HI,
The benefit, of having separate LPARs for your DB and CI, is that is is easier to troubleshoot hardware/software problems if any had to occur.
The benefit, of a single LPAR/physical, is that from an administration point of view, it is easier to maintain. E.G, if OS patches need to be applied then you need to patch 2 systems.
Kind Regards,
Chris
Edited by: Chris Geladaris on Mar 16, 2010 7:11 PM

Similar Messages

  • Need guidance in cloning my newly installed EBS 12.1.3 to a physical machine

    Hello Guys,
    We have just migrated EBS 11i to 12.1.3 with rdbms 11.2.3 on the same server. Our Test instance is on a physical machine and DEV is on a Virtual Machine. We need to refresh or Clone the Test and Dev systems with or from Prod. Which is the best approach for me to follow in cloning Prod? I am looking at Doc ID 406982.1 but it's not very clear to me. Has someone get a step by step documentation for such cloning?
    Thanks
    Mathias

    Mathias,
    The document you have is all what you need.
    You may also refer to (Troubleshooting RapidClone issues with Oracle Applications R12 (Doc ID 603104.1)).
    And, if you Google "Rapid Clone" R12 you will find many hits.
    Thanks,
    Hussein

  • How to convert a physical machine to virtual machine ?

    Hi,
    I'm confortable with Xen that I use for years, but with Oracle VM, there are some points that would need more explanation.
    Here is what I already did :
    - Create a disk file image, and make a filesystem on it. (with dd and mkfs.ext3)
    - Copy all the content of my physical machine to the virtual disk image. (rsync -a)
    - Create a vm.cfg file.
    My problems :
    - OVS use to create virtual disk image instead of virtual partition image. I don't know how to manage this. I'm using the image OVM_EL5U1_X86_64_PVM_10GB to understand how oracle encapsulated the /boot partition, the / and the swap within the same file that contains a partition table, 2 partitions : the first is the /boot, the second is a lvm pv. With lvm, they created 2 volumes : the / and the swap.
    Where can I get a clear howto to use losetup and set the correct offset in order to be able to mount the partition or to activate the lvm ?
    - I'm not familiar with pygrub. This looks very nice script, but I don't know why it doesn't work with my basic images. Is it necessary to have a /boot partition within the image file ?
    Regards,

    Hi,
    I'm confortable with Xen that I use for years, but
    with Oracle VM, there are some points that would need
    more explanation.
    Here is what I already did :
    - Create a disk file image, and make a filesystem on
    it. (with dd and mkfs.ext3)
    - Copy all the content of my physical machine to the
    virtual disk image. (rsync -a)
    - Create a vm.cfg file.
    My problems :
    - OVS use to create virtual disk image instead of
    virtual partition image. I don't know how to manage
    this. I'm using the image OVM_EL5U1_X86_64_PVM_10GB
    to understand how oracle encapsulated the /boot
    partition, the / and the swap within the same file
    that contains a partition table, 2 partitions : the
    first is the /boot, the second is a lvm pv. With lvm,
    they created 2 volumes : the / and the swap.
    Where can I get a clear howto to use losetup and set
    the correct offset in order to be able to mount the
    partition or to activate the lvm ?fdisk -ul OS.imghttp://wiki.edseek.com/guide:mount_loopback
    above link may give you some idea about how to calculate the offset.
    sample command,
    fdisk -ul OS.img
    losetup -f #get the next available loop device
    losetup -o $((512*208845)) /dev/loop2 OS.img
    pvscan
    lvscan
    vgchange -a y VolGroup00
    mount /dev/VolGroup00/LogVol00 root
    >
    - I'm not familiar with pygrub. This looks very nice
    script, but I don't know why it doesn't work with my
    basic images. Is it necessary to have a /boot
    partition within the image file ?pygrub used the kernel inside the vm image. so it is necessary to have the kernel available inside the vm. but the good old way still works, if you replace the "bootloader" line with "kernel"/"ramdisk" lines.
    >
    Regards,

  • Oracle EBS is patehtically working on VM Ware although it has same specs as of physical machine.

    Hi Everyone,
    We are running Oracle EBS on Physical machine having specs as follows:
    DB
    X2 Quad Core Xeon 2.4 GHz
    32 GB RAM
    x2 146GB SAS Drive with RAID-1 for OS
    SAN HDD x10 450GB SAS 15K i.e. 4.0TB but RAID-5 implemented (2.3TB allotted to DB)
    x4 NIC but 1 NIC used 10/100/1000
    Apps
    X2 Quad Core Xeon 2.4 GHz
    32 GB RAM
    x2 146GB SAS Drive with RAID-1 for OS
    SAN HDD x10 450GB SAS 15K i.e. 4.0TB but RAID-5 implemented (500GB allotted to Apps)
    x4 NIC but 1 NIC used 10/100/1000
    We are facing serious issues when migrated Oracle EBS on vmware infrastructure.
    Issue are related to high I/Os and high CPU usage hence performance of the application gets down badly whereas same application on physical machine working fine. We would also like to share physical machine is with lesser specification.
    Following are specs of Oracle & VM Ware machine;
    Oracle EBS is with 11g v1.1.07 Database and R12 v12.1.3 Apps
    OS for above applications is RED HAT Enterprise Linux 5.5 64bit
    ESXi version 5.5.0 build 1331820
    Hope to have a positive reply.

    Hi Linjo,
    Thank you for your response.
    VM MAchine specs are as follows:
    (There are two servers, everything is same except local storage.)
    Both servers are HP DL 585 G7 server with;
    Quad AMD Opteron Processor 2.5GHz
    12GB RAM
    x8 NIC
    Local-1 HDD x8 600GB SAS 10K i.e. 5.4TB, RAID-5 implemented
    Local-2 HDD x10 900GB SAS 10K i.e. 7.2TB, RAID-5 implemented
    SAN
    SAN-1 HDD x6 600GB SAS 15K i.e. 2.4TB but RAID-5 implemented
    SAN-2 HDD x6 2TB SAS 7.2K i.e. 8TB but RAID-5 implemented
    Yes, it is a shared environment, like we have other machines are running on VM but we have dedicated cores & memory for EBS Apps & DB servers.

  • Oracle XE on multiple virtual machines on the same physical machine?

    hi,
    does oracle's licensing for 11g express edition prohibit running XE databases on multiple virtual machines on the same physical machine? if not, does the machine have to be "hard partitioned" per oracle specifications (link below) to make it legal?
    http://www.oracle.com/us/corporate/pricing/partitioning-070609.pdf
    i found a similar question regarding 10g express edition in the forum (link below), but it doesn't seem like the question was ever definitively answered.
    Oracle XE on multiple virtual machines on the same physical machine?
    thanks

    I thought until I just checked that this was a clear cut no in that it specifies 'a single server'. Now I'm not sure as 'server' is an amibuous term. I at least know that as soon as you start giving developers access to theor own XE instance for dev/testing that there are many VM environments where you can't really know what physical macine your VM is running on to actually be able to tell if you had multiple XE instances running on a single 'Physical' server. I'm not sure the question has been answered by Oracle though.

  • WSUS, DC and CA on same physical machine?

    Hi!
    Have been playing around with WSUS as we are about to deploy that here and I had time over so I thought that I should learn a bit on how it works, how it's setup and all that.
    I have three physical machines to play with, all running WS2012R2. One is master, one runs a hyper-v and the last one I deployed as a DC, ADCS (Enterprise CA) and WSUS:
    Februaryserver: Master DC
    Marchserver: Hyper-V
    Aprilserver: DC, ADCS and WSUS
    At my last attempt I almost succeeded, but ended up with an error on the server that tried to use the WSUS Server and that was that a certificate chain was processed, but terminated in a root certificate which is not trusted by the trust provider (can post
    logfile if needed).
    I have now installed and uninstalled everything 3 times I think because I cant get this working and I started wondering about that maybe, just maybe, it's a bad idea to promote a single server to be a DC, ADCS and WSUS...?
    My questions are these: is there any complete guide to how to setup WSUS using SSL  - OR is there a guide to setup WSUS and NOT using SSL? Can't seem to find any appropriate information.
    This is my first attempt at trying to understand the CS role, creating certificates and what a CA/CN/CS etc is. Its also my first attempt at trying to setup a WSUS, but I actually might give up and let some third party do this for us instead, time is running
    out :(
    I have read and followed the following guides:
    http://technet.microsoft.com/en-us/library/hh852346.aspx
    http://technet.microsoft.com/en-us/library/hh831574.aspx
    http://social.technet.microsoft.com/wiki/contents/articles/12485.configure-ssltls-on-a-web-site-in-the-domain-with-an-enterprise-ca.aspx
    At my first attempt I didnt pay attention to the guide which covers setting up the ADCS and the result was that the CN was incorrectly named. The second time I thought I got it right but apparently I hadn't. At my third attempt I was just playing around and
    trying to setup the whole WSUS without using SSL as ALL our clients are located at the same physical location as our servers.
    So... that's about it. Story of my week so far :)
    Would really appreciate any help/guidance/support/tips/whatever that might help me out here.
    Thanks in advance and have nice holiday everyone!

    Ok, thanks for clearing that out. So, WSUS should not be installed on a DC. I suppose I should move both the DC and ADCS to one of the other servers and test again.
    The problem with the certificate will most likely remain though. If I setup ADCS on, for example, my machine called februaryserver, the CN till be februaryserver and FQDN will be februaryserver.xxx.xxx.xx, right? Other than installing the certificate as
    a trusted root certificate provider, is there anything else I need to do to get this to become a trusted certificate?
    Thanks!

  • Number managed servers per physical machine

    Hi
    I seem to remember reading somewhere that oracle recommend that you have 2 managed servers per physical machine. The reason I think was something to do with heap size or garbage collection.
    Does anyone know for sure if oracle/bea did recommend 2 managed servers? And if so what is the reason for this.
    many thanks in advance.

    There is no such restriction. The recommendation of the number of managed servers per machine would depend on each machines available resources.

  • Migrate ALUI portal from physical machine to virtual machine

    Currently the application is on physical machine. We are planning to migrate it to virtual machine.
    Is there any risk factor to migrate the application?
    Do we need to take any precautions while or before doing migration of the data?
    Thanks

    Hi,
    We've never had problems with multi-server VMWare deployments but just recently hit problems running WCI 10.3.0.1 onto Virtual Server 2005 in 2-virtual-server deployment.
    Whenever we started the Analytics, Collab and Notification services, all connectivity (inc ping) between the virtual servers dropped out after a few mins. Connectivity to the external network was maintained and the virtual network was fine without the portal services running. CPU and memory usage was ok on both servers. Stopping these services re-established connectivity.
    We've raised a support issue with Oracle who told us that 10.3.0.1 was not proven to work on Virtual Server 2005. Unfortunately we could not pin down the cause and as it was only for Dev we're running all portal services on a single Virtual Server as a workaround.
    Jonathan - what portal version are you running on Virtual Server please?
    Cheers,
    Adam

  • Different physical machines want to join a cluster

    Dear Expert:
    I have two physical machines which is located in different network,which means between them maybe have firewall or switch,
    Does Coherence support they join one cluster,If it can,how can I config?
    Thanks for your reply!

    Hi Ericnie,
    Yes it is possible!
    It is recommended to use Coherence*Extend though otherwise also possible using unicast/WKA configuration. But remember, if you don't use Coherence*Extend then your link between subnets becomes SPOF and will also limit your cluster scalabilty.
    Cheers,
    NJ

  • Need share folder for more presentation service on one physical machine?

    Dear gurus,
    If I want to scale more presentation service instances on one physical machine, do I still need set the network share folder for the catalogs? Can I just specify a local disk folder for them ?
    thanks!
    michael

    thanks, that really helps!
    A further question is do you have best practice recommendations about the ratio between presentation service instances number with the java host instance number ? For e.g, 1:1?
    I'm trying to set more instances of bi server/presentation services/java host to get better scale performance, while i can't determine the right ratio of them. One idea I can think out is watching the processes' resource costing, is that the only way? Have some experiences values? thanks!

  • Connecting my vm AD to another physical machine

    I have two physical systems(running windows 8), one have vmware with 2 virtual machines (windows server(AD) and sql) on it. Now i want to join my "other physical machine" to the AD on the virtual machine.how can i do this please?

    I have two physical systems(running windows 8), one have vmware with 2 virtual machines (windows server(AD) and sql) on it. Now i want to join my "other physical machine" to the AD on the virtual machine.how can i do this please?

  • Securely identifying a remote physical machine

    Hi,
    Is there a way of securely identifying a remote physical machine? I don't mean by using a public/private key mechanism, because I want to identify the physical machine, NOT the user, for program piracy issues.
    I posted this to the general security forum, but I think I have a better chance using some low-level network mechanism?
    TIA,
    Reggie

    I thought commercial programmers always had access to
    some unique network machine ID, like the one ActiveX
    uses to create a UID to create a unique ID for their
    ActiveX control?It's easy to generate a unique ID. But such an ID isn't secure. The user can change it, force it to be regenerated, move it to a different system, or otherwise compromise some aspect of its uniqueness/security.
    If you're willing to rely on users not being malicious, there are a lot of ways you can generate unique IDs which would identify the machine in question. If you were to combine the current time, the machine's IP address and some random value, you would have a nearly-guaranteed unique ID, but one that is only sufficient in a friendly environment.
    The "standard" UUID generator is based on the current time and the machine's MAC address. Here is some information on it:
    http://www.opengroup.org/onlinepubs/9629399/apdxa.htm

  • Run hyper-v on physical machine with server 2012

    I currently am running windows 8.1 with several 8.1 hyper-v clients. I am about to have to do a new install of the host os because i am getting a new faster drive. The convenience and speed of backing up and moving hyper-v clients to different systems is
    so easy i was hoping i could take advantage of it for the future. i would also like to use some of the functions of windows only available in the server edition of windows. as a dreamspark student you can get a free copy of windows server 2012. I have not
    set up a windows server, but i have run a Linux and bsd one. I only have 1 physical machine. Does it make sense at all to install windows server 2012 and then have all of my working systems be hyper-v clients? i play games and have a nice graphics card. the
    clients i don't think use them directly, but would use it via the host i think. I also use visual studio for app dev and i beliebe VS installs a Virtual like machine for windows phone. Can i do this? what would it prevent me from doing that i currently
    do? is there a better way to go about this? maybe do the reverse and run the server in hyper-v? thanks

    Personally I'd say it depends mainly on what you want to do with the desktop side of your machine. You certainly could install Server 2012 onto your desktop, and from reviews I've seen it's far more suited to being used on a desktop setup than previous server
    versions. You still have to do some additional work to enable some of the functionality you'd normally have on a desktop, but which would never be found on a server, but it is more plausible now. There's a bit of a discussion here of the possible issues and
    gotchas
    http://www.eightforums.com/general-support/7520-windows-r8-windows-2012-server-makes-great-laptop-os.html which cover some of the main things you'd need to do.
    If you're just looking to play with Windows Server and the functionality you're looking to use isn't Hyper-V related then I'd probably go with simply making the Windows Server 2012 another instance on your existing setup, eg within Hyper-V on your 8.1 desktop.
    There's no real difference between hyper-v on windows 8 and server 2012 so that side of things doesn't make a difference, and windows 8 is quite capable of handling server vm's (I've got 12 VM's running on my windows 8 desktop, and none of them are client
    OS's).
    The main thing which would dissuade me from installing Server 2012 on a desktop is support for other desktop applications. If you're planning to continue using the machine as a desktop, and especially if you play games on it, you're bound to hit issues
    with compatibility at some point. Not legitimate compatibility, more situations where some apps will try to detect the OS you're running, and either not recognise the server OS and refuse to install, or recognise it and refuse to install as they're only designed
    / licenced to be install on a client OS. Some apps which are free to use on a client installation for instance require a licence for use on a server (since it's assumed to be commercial installation).

  • Veeam and physical machines

    yes: http://www.veeam.com/endpoint-backup-free.html
    well, there are some restrictions with the O.S. supported.

    Can you back up a physical machine with Veeam ?
    This topic first appeared in the Spiceworks Community

  • Move Lync from a VM to a Physical Machine

    Currently we have Lync 2013 set up on a VM on a ESXi server, and I was wondering what would be the most efficient way to move it onto a physical machine? The way i see it there's 2 options:
    1. Install Lync 2013 from scratch on the new physical machine, and delete the VM.
    2. Grab the vmdk file and use vmplayer or some-other vitalization programme to create a Lync box, this way i dont lose the settings.
    Before you ask "why would you want to move the VM from one machine to another, it's because we want to free up resources on the ESXi server and want to use our spare physical machine to its full extent. It wont only have Lync 2013 on it, it will have
    vCenter, Veeam, and Windows 2012.
    Thanks in advance!

    Out of interest, Is it not possible to visualize your physical box as well, you could then Copy the VMDK to the second box and turn it on.
    No need to setup a new lync server and migrate users.
    If the physical box has Veeam, Cventer and Windows 2012 , why not do a  P2V ( physical to Virtual ) conversion. That way you would maximise the use of the physical box and improve/reduce server sprawl.
    Not sure if there is a SAN involved here else you could also setup HA , depending on VMware/Hyper-V licensing.
    You might be able to over commit the Host for a few hours depending on how much resources are available.
    but that is a risk you take, in my experience this would be far easier, and worth buying a few extra RAM chips for the sake of not having to re-configure lync.

Maybe you are looking for