I am trying to run mapr sandbox on a windows pc and with 8gb ram. But when I am trying to import the ovf its always saying ovf is corrupt while I have used multiple sources the ovf that is running on the other machine is not running in my one.I have tried to play with the configuration as well I also tried to extract and run the ovf as a vmdk but than there will be no config setup done for so that doesn't works as well. Now I have tried that on vmplayer it got install and said that the ovf format is unsupproted and when you try again it will not see the ovf file specification concern so it imported the file successfully but now its says that the vmx file is incompatible. I cannot find any way out?
I did the following for install it on Ubuntu 14.04 (being virtual machines the final destination, shouldn't be mayor problems):
On VirtualBox
Don't use the ovf file.
Create virtual machine (Machine -> New...)
On operating system, choose red hat 64 bits
On memory, you should asing 8 GB for the VM (or less, if you have an old computer like me :D)
Don't add virtual drives, you can't add both drives. Use the option "Do not add a Virtual Hard Drive"
After creation of the VM
Add both disks to the virtual machine, from settings
Configure the network of the machine as following
Attached to "Bridget Adapter"
Name: Eht0
Adapter Type: Intel PRO/1000 MT Desktop
Promiscuos mode: Deny
Cable Connected: yes
After this small steps, you should be capable of doing right click -> start, and start using MapR. Basically, we import the machine in a very complicated way, because the ovf file that is supposed to use for importing doesn't work!!
I was facing same issue on my Windows & machine. Here is what I did:
Again downloaded MapR sandbox for VMWare for windows.
Uninstalled previous version of VMWare which was giving this issue and downloaded VMWare Workstation Player for Windows 64 bit.
This time it worked.
As I had the chance to experiment with MapR recently-
MapR needs 6GB RAM
at least for the Virtual Box
(or the virtual machine you are using on windows)
if you don't grant the MapR these 6gb it is just not starting with some strange error saying nothing about that issue. You have 8gb ram on your windows machine so I recommend you to spend at least 6.2gb ram for the process.
p.s. Later I had other problems with the mapper as you can see with no support. (previous I found 1 more bug that they say will be fixed in MapR 6)
I am currently using MapR 5.2
Related
I want to run a test with KAA, so I was trying to install the sandbox in my laptop but it has only 4GB in RAM, so when I try to set up the Virtual Machine the system won't let me set up over 1,6GB and the VM won't start.
So I was trying to install in other old laptop so I installed Ubuntu 16,04 and I followed all the step by step instructions in Kaaproyect's WEB. I could do it, but when I try to start the server can't do it. I was checking the Log error and say me that the problem is in the Java's Virtual machine, can't start because only have 2GB in RAM. I need to test a Little application so is it possible change this requirement in the Java machine and start the system?
PS: I can't buy more Ram.
I recommend you to use amazon AWS. The basic installation where you can run Kaa is free for one year, and it runs quite well there.
Not sure why it was so challenging to bring up Cloudera CDH 5.5 on VirtualBox 5.0.14 on my Windows 7 workstation (64-bit). My desktop is a Lenovo 30AGS01Y00 w/ 1 Intel64 CPU, 16GB RAM, and 1TB of HDD. The detail OS version: 6.1.7601 Service Pack 1 Build 7601.
After installing VirtualBox and unzip Cloudera QuickStart VM 5.5, I created my VM with Red Hat (64-bit), Memory Size: 8,192MB and the "Use an existing virtual hard disk file" option to point to the vmdk file of the Cloudera quickstart for virtual box file. After the VM was created I adjusted its settings, as recommended, such as "Shared Clipboard", "DragnDrop", "Boot Order" (leave Hard Disk only).
The chipset setting was PIIX3 by default. It made no difference when I tried both PIIX3 and ICH9.
I left Processor as 1 CPU because my desktop has only 1 physical CPU even though its VT-x and physical HyperThreading were enabled. Nested Paging was also enabled.
The Storage of the Cloudera VM was created with SATA and Type: AHCI. The rest of settings remained no change (as default).
When I tried to boot the VM my VM screen looked exactly the same as depicted in this question:
Virtual machine "Cloudera quick start" not booting
I've been Googling this issues for about a week. The above question is the closest case I could find on the web. I tried a variety of VM settings but no luck. Not sure what the root cause is.
I tried to fall back to Cloudera QuickStart VM 5.4.2. No luck either.
Looks like zip file cannot be attached. Some key elements in VBox and VboxHardening logs were excerpted as follows:
Vbox.log
00:00:02.535619 VMSetError: F:\tinderbox\win-5.0\src\VBox\Storage\VD.cpp(6410) int __cdecl VDOpen(struct VBOXHDD *,const char *,const char *,unsigned int,struct VDINTERFACE *); rc=VERR_NOT_SUPPORTED
Note: I don't know what F: drive is. There is no F: drive on my desktop.
00:00:02.520998 AIOMgr: Endpoint for file 'C:\CDH_5.4.2\cloudera-quickstart-vm-5.4.2-0-virtualbox-disk1.vmdk' (flags 000c0781) created successfully
00:00:02.535626 VMSetError: VD: error VERR_NOT_SUPPORTED opening image file 'C:\CDH_5.4.2\cloudera-quickstart-vm-5.4.2-0-virtualbox-disk1.vmdk'
00:00:02.567924 AIOMgr: Preparing flush failed with VERR_NOT_SUPPORTED, disabling async flushes
VBoxHardening.log:
62a4.318: NtOpenDirectoryObject failed on \Driver: 0xc0000022
This may not make a difference, but have you tried importing the .ovf file (not the .vmdk one)?
Also, for future reference, there is much Cloudera VM knowledge here: https://community.cloudera.com/t5/Apache-Hadoop-Concepts-and/bd-p/ApacheHadoopConcepts
I still can't get the "Use an existing virtual hard disk file" part to work.
But, "Import Appliance" via .ovf does bring up the CDH VM. Here is what I do in case if anyone hit the same issue as me.
To import the QuickStart VM (.ovf or .ova):
Select File and then "Import Appliance (Ctrl I) " in VirtualBox.
Browse to find your .ovf (or .ova) and click Next
You may leave everything default and click Import.
Via this "Import Appliance" method, I was able to bring up QuickStart 5.4.2 and 5.5.0, on VirtualBox 5.0.14 on my Windows 7 desktop. It took about 4-5 minutes to finish the boot up process.
I have a windows 7 laptop and I need to setup hadoop (mutlinode) cluster on it.
I have the following things ready -
virtual softwares, i.e. virtualbox and vmware player.
Two virtual machines, i.e.
Ubuntu - for Hadoop master and
Ubuntu - for (1X) Hadoop slave
Has anyone done a setup of such a cluster using Virtual machines on
your laptop ?
If yes please help me to install it.
I've searched over google but I am not getting how to configure this multi-node cluster on hadoop using VMs?
How to run two Ubuntu OS on windows 7 using VMware or virtualbox?
Should we use same Ubuntu version VM image or
vm images with different versions of Ubuntu linux?
Yes you can use ubuntu two node. I am using five nodes(1 master, 4 datanodes).
If you want install multi node in vm ware.
Just download ubutnu from this link: http://www.ubuntu.com/download/desktop
And install two machine. And install java and openssh.
And download shell script for multinode from this link::
https://github.com/tonyreddy/Apache-MultiNode-Insatallation-Shellscript
And try it .....
All the best............
Since you're running Hadoop on your laptop, obviously you're doing it for learning purposes or building POC or functional debugging.
Instead of going through the hassles of installing and setting up Hadoop and related Big-Data softwares, you can simply install a pre-configured pseudo-distributed VM.
Some good options are:
Cloudera QuickStart VM
Hortonworks Sandbox
I've been using the Cloudera's VM on my laptop for quite sometime now and it's been working great.
Cloudera and Hortonworks are the fastest way to get it up and running.
Make sure you have enough RAM installed on your laptop for the Operating system already running, else your laptop will restart abruptly often while you use the Virtual machines.
Let me give you an example -
If you are using Windows 10, it needs 3-5GB RAM to be used to work smoothly,
This means if you load a Virtual Machine of 5GB size in your RAM, Windows may crash when it does not find enough RAM to operate.
You must upgrade the RAM from 8GB to 12GB or best 16GB for smooth operation of your laptop.
Hope it helps
I'm searching for a solution to boot a native OS on a hard disk as a virtual machine.
It's like what VMware Fusion did on a Mac which boots Windows in Boot Camp as a virtual machine.
In detail, I have Windows installed on /dev/sda2 and Ubuntu 11.10 on /dev/sda5.
Is there anyway to use a virtual machine software to boot the Windows on /dev/sda2 as a virtual machine while I'm using Ubuntu?
Yes, I did this long ago following this guide:
https://lists.ubuntu.com/archives/ubuntu-us-nm/2008-February/000521.html
of course, always backup and be careful!
Essentially:
Used a USB 3.5 HD enclosure and connect the XP drive to it.
If the drive was shutdown uncleanly you may need to manually
mount it with the following command.
sudo mount ntfs-3g /dev/whereyourdriveis /mount/somemountpoint -o
force
Once the drive is mounted under linux contiunue to step 2.
Launch VMWare.
Go to File -> New -> New Virtual Machine.
Select "Custom"
Select Next
Select your operating system (i.e. Win XP)
Select Next
Give it a name like "WindowsXP"
Select Next
Specify processor One or Two
Select Next
Choose public or private (on a single-user machine this doesn't
matter)
Select Next
Select the memory to devote to the virtual machine. 512 MB is a
pretty useful number.
Select your network connection
Select Next.
Leave SCSI set to BusLogic
Select Next
Select Use Physical Disk
Select Next
Select Use Entire Drive
Select Next
Specify the place to save the VM
At this point you're done Select Power On to boot the Physical drive
in VMWare!
More Info: I should add, I have successfully done this, but I also had success using this method years even years before. So there are at least two known and tested ways for accomplishing this that I can tell you.
You can do this via VirtualBox raw disk access.
(http://www.virtualbox.org/manual/ch09.html)
It basically creates a "virtual" disk file that points to the actual partition and loads it as a disk drive in the VM. I've installed Linux guest in VB on Windows host in such a way, and the installation can boot from the VM or by itself.
As answered, this also can be done in VirtualBox, this is the way that works for me
Always, make sure that you are running as Administrator(Windows) or Sudo(Linux), any changes that you do will write to the REAL disk, so be carefull
In Windows
C:\Program Files\Oracle\VirtualBox>VBoxManage.exe internalcommands createrawvmdk -filename "E:\virtualbox\linuxhd.vmdk" -rawdisk "\\.\PhysicalDrive1"
RAW host disk access VMDK file E:\virtualbox\linuxhd.vmdk created successfully.
In Linux
$ VBoxManage internalcommands createrawvmdk -filename "~/linuxhd.vmdk" -rawdisk "/dev/sda"
It will create a file with something around 1kb that is a link to the physical hard drive.
Then create a Virtual Machine as ever you do.
If you want to map only a partition
At Windows
\\.\Physicaldrive1 -partitions 1
(Disk start with 0, partitions
with 1)
At Linux (Much more intuitive)
/dev/sda1
/dev/sda2
etc.
Eventually you can get resolution issues
Eventually you can get resolution issues even after install vboxadditions, in my experience the problem is your /etc/X11/xorg.conf it is configured to your specific real hardware specs(I have a offboard GPU for example), least in my case I solve it simply removing this file (xorg auto configure at boot, only will not work if you set some specific setting), so run:
sudo cp /etc/X11/xorg.conf /etc/X11/xorg.conf.original && sudo rm /etc/X11/xorg.conf
Reference
http://www.virtualbox.org/manual/ch09.html#rawdisk
https://forums.virtualbox.org/viewtopic.php?f=6&t=36694
https://romaimperator.com/?p=29
I started on Ubuntu and have had the first considerable error. I'm looking for help.
I have an HP Pavilion dv6 i7. I had installed windows 7 and I decided to also install Ubuntu using a USB.
My first attempt was to install Ubuntu 11.10 following the instructions of the official Ubuntu website. When loading the pendrive, my PC stucks at the main menu of ubuntu, so after searching, I found could be due to a problem with my AMD Radeon graphic card (or not), but I decided to change.
Then I used Ubuntu 10.4. This could happen from the start menu i get into Ubuntu live. There I decided to install it because I liked it and I need to develope with Google TV (in windows is not posible).
And I fail in the partitions section. I tried to follow the instructions on this page:
http://hadesbego.blogspot.com/2010/08/instalando-linux-en-hp-pavilion-dv6.html
but there were things that changed a bit so I improvised. I took the windows partition of 700000MB and went to 600000Mb leaving 100GB free to install Linux there. The error was to set it to ext3 (it was ntfs). I thought the new 100gb partition will be set to ext3, and windows partition will stuck at ntfs system, but not.
Total I ran out to boot windows, and above I can not install ubuntu on the 100GB free.
Someone thinks I can help. Is there any easy way to convert back to ntfs windows and not lose data?
Thank you very much.
You should be able to hit F11 when the machine is booting up and go to the HP recovery application. This should let you reset to factory default.
You should definitely be able to install Ubuntu on the new 100GB partition as well. Just make sure you choose the right partition to install it on.
You will need to recover using recovery CD/DVD's. You must have been using the install gparted utility in Linux to "re-partition" your drive. You scrubbed some boot files.
If you successfully recover using the recovery media you can use Disk Management in Win 7 to shrink or extend your volume. In your case you would shrink it down 100Gb's and then when installing Linux gparted will see that available 100 GB and install there while Windows will still run.
Also, you should probably be running ext4 fs, not ext3. you would only want ext3 for compatibility reasons.