Installing OnCommand Unified Manager on CentOS

CentOS is my favorite Linux distribution because it is the closest to RedHat you can find (for a reason). And I like RedHat for the simple reason that it is the most supported Linux distribution for any enterprise applications. I know usually both RedHat and SuSE are supported, but SuSE and I was never a love story and I will save you the history of this little drama for this time.

So, the idea is not to run CentOS in a production environment, unless you really have the heart of a warrior and do not care about being supported by a rock solid editor, but instead to run your OnCommand Core lab on a free OS.

I used a CentOS 6 minimal install for this tutorial.

The problem

Here is what happens if you try to install the package right away after downloading it from NetApp download site :

# chmod 755 
# ./ 
Preparing to install OnCommand Core Package
Unpacking files needed for the installation ...
error: This is not a supported platform.

See ? We will have to trick the installer to force the installation

The solution

[EDIT] Thanks to Hadrian in the comments, there is actually a very simple way :

# ./ -o

That’s it !

I leave the rest of the original article for the archives !

The trick is just to interrupt the installer before it gets a chance to remove the uncompressed installation folder, the easiest way I found to do that is to use gdb (the linux debugger) and break on « unlink() » which is the C function that removes files on disk. It sounds complicated like this, but it is really trivial, as you can see : (you might have to install gdb with « yum install gdb » if it is not present on your system)

[root@localhost ~]# gdb ./ 
GNU gdb (GDB) Red Hat Enterprise Linux (7.2-60.el6_4.1)
Copyright (C) 2010 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law. Type "show copying"
and "show warranty" for details.
This GDB was configured as "x86_64-redhat-linux-gnu".
For bug reporting instructions, please see:
Reading symbols from /root/
(gdb) break unlink
Breakpoint 1 at 0x400d28
(gdb) r

At this point, the installer will run and just stop when it tries to remove temporary files

Starting program: /root/ 
Preparing to install OnCommand Core Package
Unpacking files needed for the installation ...
Detaching after fork from child process 2020.
error: This is not a supported platform.

Breakpoint 1, 0x00007ffff7b343b0 in unlink () from /lib64/
Missing separate debuginfos, use: debuginfo-install glibc-2.12-1.7.el6.x86_64
(gdb) q
A debugging session is active.
Inferior 1 [process 2017] will be killed.
Quit anyway? (y or n) y

Here we just quit with « q » and confirm to do so with « y ».

Ok, now let’s move the installer to a safer place :

# mv /tmp/dfmsetup-*/ ~/dfmsetup
# cd ~/dfmsetup

Now we just have to edit the installation script and disable the system check.

Just edit and change the line :




Now you can run the installer :

Note that « NTAPdfm » here is whatever the RPM package is named before the version string. For example, mine is called : NTAPdfm-5.2-16233.x86_64.rpm.

If you are instaling an OEM release like IBM’s, it might be different.

#  ./ -p NTAPdfm

WARNING: A minimum of 3072 MB of RAM is required for proper functioning of OnCommand Core Package.
You have only 989 MB RAM on this host.

By default, the DataFabric Manager server sends AutoSupport event messages
and weekly reports to NetApp Technical Support. Messages are sent
over secure HTTPS, HTTP or SMTP(HTTPS by default). This information is used to
assist you with troubleshooting and problem resolution. AutoSupport will
significantly speed problem determination and resolution should a problem occur
on the system.
Status AutoSupport event messages will be sent to NetApp once per
week and will start 24 hours after installation. If you do not want
AutoSupport event messages sent to NetApp, you can disable
AutoSupport by setting the AutoSupport Enabled global option to "No".
This option will initially be set to "Unknown" unless manually set, and
will automatically become "Yes" after 24 hours if not set manually to "No".
If you do not want to include private data such as IP addresses, host names,
and user names, set the AutoSupport Content global option to "minimal".
Please respond with "Yes" if you have read and agree with the use of the
AutoSupport feature on this DataFabric Manager server or "No" to abort the

Storage systems running Data ONTAP 8 operate either in 7-Mode or
in Cluster-Mode. Each installation of OnCommand Core can manage storage
systems operating in one of the two modes:

[1] Cluster-Mode
 Data ONTAP 8 operating in Cluster-Mode.

[2] 7-Mode
 Data ONTAP 7G and Data ONTAP 8 operating in 7-Mode.

Which kind of storage will you manage with this installation
of OnCommand Core Package?.[1,2](default 1):

Enter your NetApp DataFabric Manager server license key [?,q]: **************
Beginning the OnCommand Core Package installation ...
Verifying checksum of OnCommand Core Package package.
Préparation...              ########################################### [100%]
   1:NTAPdfm                ########################################### [100%]
Installing scripts in /etc/init.d directory.
Creating the DataFabric Manager server keystore.


        The Standard edition of the OnCommand Core Package, including the DataFabric Manager server, has
        been installed on your computer.

        You can now access OnCommand console at


        to configure and monitor hosts and storage with DataFabric Manager server.

        For help on the dfm command, enter

            dfm help

        at the command prompt.

DFM should be up and running, you can point your browser to the address mentioned at the end of the installation.

2 réflexions sur « Installing OnCommand Unified Manager on CentOS »

  1. fourmi

    Il y a un moyen beaucoup, beaucoup + simple ( there is an easier way ) :
    # vi /etc/redhat-release
    Red Hat Enterprise Linux Server release 6.5

    That’s all, c’est fini

    1. ybizeul Auteur de l’article

      Même encore plus simple comme mentionné dans le [EDIT]
      # ./ -o

      J’ai quand même laissé l’article original pour référence !


Laisser un commentaire

Votre adresse e-mail ne sera pas publiée. Les champs obligatoires sont indiqués avec *