VMDK Puns

Hi folks,

We have a customer that has a 25 tb vm/ vmdk with Linux on it. The vmdk is thick. (Should be thin but due to translation error in VMware esxi thick and thin were switched - another story)

The Datastore is only 28 TB so a clone copy to thin with vmdkfstools is not an option...

Actual data on the VM is like 9tb or so.

So I was wondering:

Could I force VMware esxi (6.7) to see the disk as thin? Then if it sees it as thin use DD to fill it completely with zeros and then use hole punching to make it real thin.

I know the later steps do work if I do a "thin" clone. I can use the later steps to make the clone thin over time with space reclaim.

However I'm not sure where the vmdk type is recorded. My guess would be in the vmdk itself?

Any easy way to mark the disk as thin?

My approach would be to recreate a thin 2.5tb HDD and then DD the first few bytes with the partition table into the thick disk. Making it think it is thin. Then do the steps above to make it real thin.

However I do not know if vmdks even have partition table s of some sort and how many bytes to record the thin /thick part.

Any ideas for this approach?

We do have backups from the data but we would like to avoid restoration of 9tb if possible...

I know this is a head scratcher.

Thanks for your help!

I've become fairly versed in exporting servers from my ESXi as OVF/VMDK, and importing them into AWS and then creating an AMI from them.

However I recently exported a server, which makes use of dual disks, and as a result is exported as a single OVF file with dual VMDKs.

But for the life of me, Im unable to figure out how to import dual VMDKs and combine them into a single AMI.

Does anyone have any experience with this, and possibly have a guide?

Hi All, how do you all convert vmdk files into vhdx? I’m trying to upload a vm to azure from a VMware environment

hi guys,

disconnected 2 vmdks from 1 file server and mounted on new server. normally vmotion storage over to DS i want so the path reflects the new server... i moved 1 of the 1tb vmdks okay but 1 of them keeps failing with -

A fatal internal error occurred. See the virtual machine's log for more details. 2021-11-25T11:54:20.256539Z Failed to copy source (/vmfs/volumes/*******_4.vmdk): No space left on device. Failed to copy one or more disks.

Any ideas - DS im moving it to has plenty of space .. think i need to power off VM and try.

vSphere Client version 7.0.3.00100

thanks-travis

FYI, I maintain a guide with VMDK's for download. It is geared towards DIY diabetes management software that requires a Mac to build. https://macosvmware.tech.blog/

I'll cut to the chase.

Windows 10 virtual machine installed on ESX 6.7

1 datastore that is on a single Intel 400gig SSD.

Both C: and D: drives of the Win10 vm are on the Datastore1 (SSD). Both thick provisioned, both SCSI

If I copy a file from my NAS on to the C: drive of the Win10 machine, starts off fast (400mb/s) then drops to around 210mb/s (expected behavior)

If I copy a file from the NAS to the D: drive of the Win10 vm, starts off fast (400mb/s) then drops to 60mb/s

I have NO idea why! Both C: and D: drive are VMDK's on the same SSD device.

Why is D: so much slower than C: ?

Hello!

Can I backup my VMware Workstation VMs and snapshots in Backblaze? Did I notice that Backblaze exclusives the vmdk and other VMware file types? If I remove it, will it back it up? Or does Backblaze does this for a reason?

I am using a personal Backblaze account. Just wondering if there is a reason why VMware files are exclusive.

Thanks!

So I had vbox working perfectly the way I wanted it to. I had to reinstall my host os. Reinstalled vbox w/ expansion. Used the command that i had used in the past to get it to boot a live usb OS:

sudo vboxmanage internalcommands createrawvmdk -filename ~/usb.vmdk -rawdisk /dev/sdb

the file usb.vmdk is created, no problem except its got a red ❌ over it and some other forbidden style icon. Didn’t happen to me last time I did it. Vbox won’t accept it as an “existing virtual hard disk file”. I tried changing the permissions the way someone suggested online with:

sudo chmod 666 usb.vmdk which removed the forbidden icons

But both ways return this in vbox:

VERR_ACCESS_DENIED

VBOX_E_FILE_ERROR (0x80BB0004)

VBOX_E_OBJECT_NOT_FOUND (0x80BB0001)

Could someone help me figure out what the hell is going on here please?

Using: vbox 6.1.3

Host OS: Ubuntu

Guest OS: Kali Live

Expansion pack installed but not Guest Additions (I could be wrong but I don’t remember needing to use this last time when it was working)

ps. I booted up my live usb not in vbox and it worked so usb is still working fine.

Hi Everyone, Very new to proxmox been using VMware workstation and now migrating my VMs to proxmox.

I have tried t way acronis images and restoring them to a proxmox drive but drive won’t boot up.

Tried importing the vmdk but I think I’m doing it wrong. Most guides online reference esxi.

Hello,

I have problems to convert vmdk to hyper-v with Microsoft virtual machine converter. Does anybody knows a better (free) tool to use? Or a combination?

Many thanks!

I'm looking for solutions through which we can convert vmdk into ovf/ova format. Ovftool is there but it is for Windows and I want to convert it in Ubuntu so is there any alternative tools out there? (Wine could be the solution but any other?)

I searched on Google but I didn't find anything related to it.

Thanks in advance.

It's fine that the VM file expands when needed but after a while ( installing and uninstalling apps ) it's size gets ridiculously large.

Total Space taken = +38GB
Total Space Used = 6 ∼ 7 Gb

If anyone knows a way to reduce it ( Without loosing data ), would be appreciated if he/she could write it down.

Also, NoxPlayer Developers: Please make that function... allow users to shrink and/or compact their VM.

P.S.: I tried transforming it into VDI, without success. At least loading Gparted on VirtualBox and shrinking the partition works ( But there's still space left and can't delete file 3, inside the .vmdk file - "0.img, 1.img, 2.img, 3" are inside it. 0,1 and 2 are partitions, 3 is the unallocated space. If we where able to delete it... ).

Other Reddit post were not that helpful.

Hi and ty for any input.. have a windows 10 vm that was running on windows 10 using workstation pro 16.. created a vmdk from that virtual machine.. transferred it to USB now whats easiest method for transferring to and running in azure VM?

Ty

I have a ton of VMs on my tinkering/testing computer, and one of them is running Kubuntu on ZFS. Periodically I run a 'dd if=/dev/zero of=/zerofile' operation inside each of my VMs to fill the unused disk space with zeroes, so I can trim down the VMDK files -- otherwise I would need terabytes and terabytes to store all of the VMs I have. Only problem is, 'dd if=/dev/zero of=/zerofile' doesn't work on ZFS, because even with compression turned off, ZFS still detects incoming blocks filled with zeroes and throws them away before they reach the disk. I learned this the hard way when I started the operation on my Kubuntu on ZFS VM and walked away, coming back a couple hours later to find about 10TB of imaginary zeroes written to a 127GB VMDK.

Is there a correct way to zero-fill the unused disk space? I have to do something. My Kubuntu on ZFS VM is starting to turn into a space hog, and it doesn't even contain that much actual data -- it's all just random bits left over from previously-existing files, which I haven't been able to overwrite with zeroes.

EDIT: The problem appears to be the setting 'dedup=on'. Apparently ZFS was realtime-deduplicating all the zero-filled blocks into a single block with a zillion reference links. Not sure how dedup got turned on, but turning it off finally allowed 'dd' to zero-fill the unused disk space.

I have had an issue with a hard drive failing, however, I think I was able to save/recover the vmname-flat.vmdk file.

Is there any way to recreate the virtual machine with just this file? Whenever I try to add it as a hard drive (as shown in VMWare KB 1002294) the vmdk file doesn't show up in the datastore". But when I go to the datastore I don't have any problem seeing it.

I have tried the following articles/videos:

https://kb.vmware.com/s/article/1002294

https://www.youtube.com/watch?v=FDdi5fvdeHw&t=76s

However, now I get the error "The file specified is not a virtual disk" when trying to start the VM.

Thanks for your time!

Additional pictures for reference to comments:

Files currently in my Zabbix \"datastore folder\"

When I follow steps in VMWare KB 1002294, I can see other vmdk's but not the Zabbix one.

The Zabbix folder just returns a blank page.

What the Zabbix server .vmdk file looks like

How other VM .vmdk files look.

Is there any publication for the different deployment sizes of VCSA somewhere for 7.x?

I have referenced VMware KB 78515. It only lists the default Tiny deployment and I cannot find the level of detail I need to expand the disks.

Restoring using Exporting Disks to with vmdk does not maintain file level permissions. i backup my file share drives ( D, E and F ) using Agent, when i try to restore using the Export disk with vmdk to a vmware datastore the file level permissions are not there, but when i explore the files in Veeam B&R i see the permission ?

Hi,

We will setup 2-node windows server file cluster in the VMWare ESX Cluster. I want to do VMDK clustering inside VMWare ESX 7.0. Anyone here who has any experience with File Server Clustetering? Any thoughts? We have around 1000 users.

thanks,

Hello,

I have a 10TB XFS filesystem hosted on several 2 and 3TB RDM disks aggregated with LVM. I wish to use VMDK instead.

I already know how to make the RDM → VMDK migration (pvmove) but I don't know what is the best practice for the size of the VMDK. I was thinking to create a new datastore dedicated to this VM with two 4TB VMDKs and one VMDK of 2TB.

I'm a little reluctant to make one big 10TB VMDK because if one day I have to move it with vStorage Motion Storage vMotion it will be more complicated than moving smaller 4TB VMDK.

Any thoughts?

I converted a Windows XP VM from Vmware with qemu-img, from vmdk to raw, but when I boot the image, I am stuck in a login screen, because I am unable to use keyboard or mouse. I've tried many, many things including passing "-usb -device usb-kbd -device usb-tablet -device usb-mouse" and nothing. The system appears to be running fine, I just can't... interact.

I am using both bluetooth mouse and keyboards, a Satechi X1 kbd and a Apple Magic Mouse, host is a Late 2013 rMBP 13-inch i7.

Anyone know how can I fix it? My VMware VM is very updated and customized, I didn't want to go through another two days creating a QEMU VM from scratch.

Hi,

I've just expanded the data disk of one of my file servers and it is now 2 TB. For various reasons (new engineering software and a lot more use of photos to document works) data are growing faster in the last couple of years than in the previous 10.

I'm wondering how big you let grow your .vmdk and what solutions you put in place after you reach that limit.

It is ok to have a 10 TB .vmdk or do you use something different ?

Sorry for the stupid question but really I didn't have need of a lot of storage in the past years... things are changing quite fast ...

ok, so i have a win10 system i want to use as a VM instead of a physical system. it hasn't been booted into or configured, i want to just move the disk into vmware and do it all there.

i put the disk in my desktop and made an image of the physical disk w/ `dd`. then, i converted to vmdk with `qemu-img`.

i also followed the instructions in this post to ensure the correct permissions on the vmdk, convert to thin provisioned, and to update the controller for the disk to lsi.

i'm still having trouble booting the VM.

anyone have any thoughts? i've gotten the following in the `vmware.log`:

    2021-09-02T09:20:53.334Z| vmx| I125: DISKLIB-SPARSECHK: [/vmfs/volumes/60d44015-6d58a252-8c6a-001e6787a5d2/old/desktop03.vmdk] Invalid GD or RGD [7629]: 15662336,7831808 vs. 61185,30597
    2021-09-02T09:20:53.334Z| vmx| I125: DISKLIB-SPARSECHK: [/vmfs/volumes/60d44015-6d58a252-8c6a-001e6787a5d2/old/desktop03.vmdk] Invalid GD or RGD [7630]: 15663360,7832832 vs. 61189,30601
    2021-09-02T09:20:53.334Z| vmx| I125: DISKLIB-SPARSECHK: [/vmfs/volumes/60d44015-6d58a252-8c6a-001e6787a5d2/old/desktop03.vmdk] Invalid GD or RGD [7631]: 15664384,7833856 vs. 61193,30605
    2021-09-02T09:20:53.334Z| vmx| I125: DISKLIB-SPARSE: "/vmfs/volumes/60d44015-6d58a252-8c6a-001e6787a5d2/old/desktop03.vmdk" : failed to open (14): Disk needs repair.
    2021-09-02T09:20:53.335Z| vmx| I125: DISKLIB-LINK  : "/vmfs/volumes/60d44015-6d58a252-8c6a-001e6787a5d2/old/desktop03.vmdk" : failed to open (The specified virtual disk needs repair).  
    2021-09-02T09:20:53.335Z| vmx| I125: DISKLIB-CHAIN : "/vmfs/volumes/60d44015-6d58a252-8c6a-001e6787a5d2/old/desktop03.vmdk" : failed to open (The specified virtual disk needs repair).
    2021-09-02T09:20:53.335Z| vmx| I125: DISKLIB-LIB   : Failed to open '/vmfs/volumes/60d44015-6d58a252-8c6a-001e6787a5d2/old/desktop03.vmdk' with flags 0x10a Object type requires hosted I/O (76).
    2021-09-02T09:20:53.335Z| vmx| I125: DISK: Cannot open disk '/vmfs/volumes/60d44015-6d58a252-8c6a-001e6787a5d2/old/desktop03.vmdk': Object type requires hosted I/O (76).
    2021-09-02T09:20:53.335Z| vmx| I125: DISK: Opening disks took 45 ms.
    2021-09-02T09:20:53.335Z| vmx| I125: Module 'Disk' power on failed.
    2021-09-02T09:20:53.335Z| vmx| I125: VMX_PowerOn: ModuleTable_PowerOn = 0
    2021-09-02T09:20:53.335Z| vmx| I1

1 of my client has the following set-up.4 servers installed with ESXi with multiple VMs (For naming purposes, A, B, C, D)
1 SAN Storage

Storage is split as such:
1 LUN for each VM to be installed with OS (C Drive), totaling around 30
1 LUN for each server as a datastore, totaling 4.

The client is requesting to share certain drives between VMs in A and B, C and D.

Tried googling and I can't seem to find any article of sharing vmdk between VMs of different ESXi hosts, only between VMs of similar ESXi host.

May I know if this is even possible and how should I go about doing it?

As stated above, I have VM that is receiving the above message when trying to power it on. Is there an 'easy' way to fix this? This is a server I inherited and I don't know much about VMWare. Any help would be appreciated.

Hi,

A strange thing just happened to me:

• On vSphere 7, I've created a new 16 TB datastore (VMFS 6.82),
• on this datastore, as a test, I've allocated a new 4 TB VMDK for a VM with no issue,
• then I've deleted the VMDK and since then the max size was limited to 2 TB.

Here is a screenshot of the datastore properties showing the 2TB limit:

https://i.redd.it/6l56tjytfyf71.png

It's like there's now a VMFS 3 limitation on a VMFS 6 datastore. I've tried unmounting / remounting without success, the only solution was to unmount, delete and re-create a datastore.

Hi there.

In my Organization we are looking for a solution to upload VMware VM's to Amazon S3 from a Synology NAS. We already have a backup tool called Active Backup For Business running daily backups and it really works fine. The problem comes when we try using Synology's solutions for uploading those backups to S3. For example, on the Internet I found Hyper Backup and also Cloud Sync. Both of them have troubles with Active Backup format, the first one isn't able to select ActiveBackup directory's 'as source and CloudSync somehow detects that the directory is managed by ActiveBackup.

Out final goal is: Make backups onsite on our Synology NAS and then upload those backups to Amazon S3.

I found something called ALTARO that's just fine but the pricing...

Thank you before hand.

Thank you all guys. At the end we are planning to use a deployment with Duplicati and also AWS-CLI.

Duplicati for incremental backups of the data uploaded to S3.

Hi all

Need to migrate some RDMs to VMDKs (trying to get rid of all RDMs) but have limited timeframes in which I can do it on production boxes.

I know there's no issue doing it while powered off, but can it be done with the VM live and the disks still accessible?

Thanks!

Today we'll be showing you how to install Home Assistant (aka. Hassio) on your Unraid system using a VMDK. This also shows you how to use any other VMDK on Unraid.

Home Assistant is free and open-source home automation software designed to be the central home automation control system for smart home technology. Both the Home Assistant "core" application itself and its software extensions are written in Python. Its main focus is on local control and privacy.

https://youtu.be/w28_cPoFn8k