Tip for using that last bit of space on your drives

No matter the size of drive, you probably have anywhere from 1-99GB of free space on any drive after filling it with plots. Ever wish you could fill that extra space with more plots? Now you can! (on Windows 10 at least)

The trick is to fill the last bit of space on each drive with a Virtual Hard Disk and then merge all of these little VHDs into one big Windows Storage Space. Virtual Hard Disks are normally used for Virtual Machines, but you may not know that you can use Disk Manager in Windows 10 to create new VHD files anywhere you like!

Go into Disk Manager and click Action then Create VHD:

Choose the path to your drive, the size of the free space (I like to leave maybe 1GB free just to be safe) and choose the newer VHDX format and Dynamically expanding size:

Make sure to “Initialize” the new disk in Disk Manager.

Now open Windows Storage Spaces by typing “Manage Storage Spaces” in the Start bar. Click “Create a new pool and storage space.” Select the “drives” you just created and choose “Simple” resiliency. Use all the space available. BE SURE TO ONLY CHOOSE THE NEW VHD DRIVES YOU CREATED, NOT YOUR EXISTING DRIVES.

This should create a new mapped virtual drive that is all of those little virtual drives smashed together! You can even use Disk Management to map that new drive to a new folder, just like your other drives. I name my folders after the drive’s serial number, so in this case I name this special folder “00000000” so that it floats to the top of the folder list.


You can also add more VHD’s and then add them to this storage space and then expand the storage space, all from this interface.

Using this method I was able to reclaim about 1.4TB of drive space that was just sitting there. That’s 14 plots! :sunglasses:

Bonus: this virtual-virtual drive works great with my PowerAutomate setup: Using Windows Power Automate Desktop to move plots - Chia Plotting - Chia Forum


Can you do something like that on Linux?

1 Like

I don’t know, but I have to imagine so. The concept of virtual disk image files and software RAID 0 is nothing new, but I was surprised to see them work together so nicely in Windows!

Here’s that same directory structure after several hours of adding plots to this new drive. Notice that “00000000” is filling up but also notice how all the other drives have also decreased in free space as well. All the little VHD’s on those drives are slowly filling up and using that space!


Wow niiice! I wonder what happens if I use on external drives and one get off-line, the VHD fail and recover after it goes on-line? The files`ll be intact?

Theoretically yes, but if one of the VHD is missing then the whole merged drive will not be accessible until you attach that missing VHD again. This is normally really bad practice cause there are lots of points of failure - but in this case, we weren’t using that space anyway and if it does fail, it’s no worse than losing a single disk anyway. Just replot! :slight_smile:

Also check all your use cases before maxing out the disk completely. That bit of space is not always there just for the lulz


1 Like

You are making a huge mistake. Filesystems don’t like to be filled to the brink. Going for zero-bytes-free (or close to it) is a bad idea from a NTFS or ReFS perspective as they need some free space to work. No filesystem should be filled up till the neck.
Sorry but this is really bad tip. Read up on how filesystems work before posting stuff like this.



Read up on how filesystems work before posting stuff like this.

Perhaps you should take your own advice.

NTFS or ReFS perspective as they need some free space to work.

This space is already “lost” when you format the drive (the filesystem will reserve the space it needs to address the whole partition). Because of the low number of files that a drive devoted to plots will have, you will be unlikely to run out of the “default” space provided.

There will likely be performance degradation but as long as your plot times are still <5s in your harvester, you are good. Now the VDDs are RAID 0 (striped), so the moment you lose one drive, you lose the RAID0 array. But you where not really using the space anyway.

1 Like

Unfortunately not true in the sense that some disk space can be needed for things not directly under the control of the file system. Read the link I gave above for examples. That is why I said check your use cases.


Lol. Take your ten-year-old different-use-case sky-is-falling advice to someone who is worried about losing real data.

1 Like

Thank you for offering this advice in the form of advice rather than absolute prophecy lol. I did read the link you posted but I honestly couldn’t find the issues you described there. Would you mind summarizing exactly what issues can be expected by filling a drive to within 1 gigabyte of its capacity as I said in the OP (“I like to leave maybe 1GB free just to be safe”)?

There is a few in the document, but as said , they might not apply to your use cases

on an MBR disk, GRUB 2 hides data in the otherwise unused and unallocated space between the MBR and the start of the first primary partition. This unallocated and unused space may not exist on a GPT disk, so another place is needed to store the data. If you’re in doubt, go ahead and create a small BIOS Boot Partition of about 100 KiB to 1 MiB. At worst, it’ll consume a small amount of disk space.

Assignment of 128 MiB of unallocated space after most partitions on disks larger than 1 GiB. (The ESP is an exception to this rule.) The intent is that this space may be used by disk utilities to help them do their jobs.Mac OS X 10.6 refuses to install if its target installation partition doesn’t have sufficient surrounding free space, so creating such gaps is particularly important on Macs.

Oh ok, so I did read the issues you were describing.

Do I read this right that these issues together would add up to…128mb + 1mb = 129mb?? Are we really having a worried discussion about 129mb? I did mention that I leave at least 1gb of free space but in actuality that is really like 4-5gb because I’m not able to complete fill up the virtual disks either.

So just to make sure I understand, as long as I leave 129mb free, there are no other concerns that you have?

If none of those use cases apply to you, then well, you do what you do. I have no dog in this race.

As for myself, I’m not to worried about a tiny bit of unused disk-space. If the system engineers used this approach as precaution, I’m fine with that.

1 Like

Fair enough! I think maybe if next time you replaced the nebulous “use-cases” with “exactly 129mb” we probably wouldn’t have had an extended discussion cause I’d be like “uh yeah I don’t care about 129mb especially since I already said I’m reserving 1gb” :slight_smile:

Sorry, it’s just that sometimes you have to really dig a little deeper with some of the advice in this forum! Look at poor lonely Steve up there - he’s apparently having a conniption fit about 129mb :laughing:

Oh look who is being over-sensitive :rofl:
Is your Caps-Lock broken?

But seriously, you might not care about losing plots or other data (which is fine, we don’t care) but others might want to hold on to their data and taking your advice is just bad advice in that context.
Someone needs to tell others to ignore your “advice”, might just as well be me (it’s a dirty job but somebody has to do it).

I’m not sure why you seem to take this as some sort of adversarial thing tbh. People are just providing information as they see it and together we learn.

I like your thinking about novel approaches.

Whether that one plot extra backing off all the other drives will be able to respond in time given that it will be spread out over 10+ or 20+ drives, now there’s some interesting data as well.

1 Like

I certainly don’t mean to be adversarial! Some might take Steve’s posts as adversarial for sure, but I can’t seem to find any language of the adversary in my posts to you. Apologies if my challenges come off that way! I’m here to learn too, so when I offer a tip and somebody tells me it is “bad advice” I want to know why and if that is true!

I will definitely keep an eye on the timing. We are only talking about 14-15 plots for now and actually I noticed that the file transfer on those plots was SUPER fast when writing them, so I’m taking advantage of lots of spindles. I assume the same would be true for reading, and this makes sense for RAID 0.

“Look at poor lonely Steve up there - he’s apparently having a conniption fit about 129mb”
My son, I get all worked up and sweaty about 1mb.
I remember when I was young and having a 10MB Harddisk in my system. We learned to savour every bit and byte back in them days.

Now go back to your basement and polish your Chia coins.