Jcloud

I honestly have no idea, but I would hazard my guess on the answer being, it is the guest OS that will do the TRIM. Such that guestOS sends TIRM, passed to SATA (guestOS), which is passed through to host sata layer, which gives it to block device and ultimately the drive's own internal logic does trim. Only the guestOS is going to be aware of the unused blocks and issue the addresses to clear.

I've been happy with my Threadripper 1950X setup, for build details check out my forum signature. If you have any specific questions about my build (scope I feel confident enough to answer about), feel free to PM, or ask in that thread - I'll answer as best I can.

Same error or different kinds of errors each time, if it's different I'd call that progresss/learning; if it's the same error then something is up. That makes a lot of sense, actually, the WindowsPE environment (automatic repair feature) I've seen come up missing with drivers. Do you know if that card has its own bios/firmware? If it does, do you have video-bios dumped, and the file added to your VM's configuration?

If this is a brand new VM, with nothing (not much) invested, I'm tempted to say, "just start over," as there is little diagnostic information for pointers on what to try. Since you are getting to the automatic recovery part Windows, I would try a System Restore (just to see if it has a restore point, if so - try it). You mentioned the drive doesn't show up in diskpart; are you passing a HD through to your VM, or is it an image file?

Your video card looks fine to me, a cheapo for video - should work just fine. For the two SSDs general consensus seems to be to just have those on the SATA controller on your motherboard, and just keep the HBA card managing the hard drives.

Perhaps the file hasn't been made executable? (hazard guess) chmod +x /mnt/user/appdata/netdata.sh Then try to add it to extra parameter as: ;/mnt/user/appdata/netdata.sh I know the above looks odd, but the thought of leading ";' is to terminate /usr/local/emhttp/plugins/dynamix.docker.manager/scripts/docker run..... Also, I'm a newb - so I'll ask for clarification, are is that shell script inside or outside the container you are running?

Perhaps encapsulate it inside a file, for a shell script, and then call the script file? That's what I would try. myscript.sh #!/bin/bash -e /usr/sbin/netdata -D -u root -s /host -p 19999 sleep 10 chown -R root:root /usr/share/netdata/web/ change file mode for executable, and user permissions. Then call myscript.sh in the template. It wouldn't surprise me if this way is the ghetto way, and there are better ways. Just my two cents, for what it's worth.

I tried doing that with the checkmarks, but choose poorly in colors and perhaps didn't fill them in as well as I could have.

So I was going to as for this feature, but I found the request is already here. I made some mock-ups. I'm not a graphics artist, so I simply copy/pasted existing elements. In the first picture for /VMs all of the running cpu-pins are listed, and colorized - the user has a better visual of assignments. Second picture, is same as before; this way the user has the information for new provisioning. Devils advocate (the non-programmer's edition, me) counter points: Does this create too much over-head this element in a pure NAS setup? Will this create a color-chaos rainbow effect with too much VMs - looks like clowns. Visual style - would care if there is overlap of VMs what then? Is this worth programmers time for this visuals? If can colorize and not make a mess, don't forget about color-blindness. I know this is an old post too, so I would like to throw in, that I have appreciated the changes made in VMs - so that you can look at individual vm assignments.

+1 I have had this thought. Stop VM_X then start VM_Y

If you can (or want to) hold out for a few months, you'll probably start seeing some deals on the TR 1950x, since TR2 is due to come out at the end of the year. Very happy TR owner here, but I should be upgraded from a 3930K.

DDR4 2666 -- I'm 85% certain the answer is yes, they will work.

I'll have to go looking for that, I didn't know it was there, hence my suggestion. Your other post where guy was looking for WinDirstat makes a whole lot more sense to me. I guess we suggest the tools we know and use. Some people just have a better tool-kit/knowledge base (suggesting yours is >> mine).

If you're looking for visualization of biggest files, take a look at QDirStat a WinDirStat/GrandPerspective clone - if you think that will help you. Glad to read the solution set is working for you. Have a good one.

Google Chrome Version 67.0.3396.87 (Win) extensions: Disconnect 5.18.23 uBlock Origin 1.16.10 Edit: Ah yes our own Acid3 test. Looks the same in incognito-mode (no extensions).

Recently tried @jbartlett DiskSpeed Docker, nifty little tool and graph. I think it's fair to say Mr. Bartlett put in a fair amount of work into this project. "Let's give Dr. Klahn a great big hand!" Measured in 10% chunks.

It's possible to do the same thing with the other image, it's just a lot more clunky. StorjMonitor container was in response to the community wanting less clunky and a few features; Storjstat.com monitor script (which the authors of the last repo added shortly after I did), and log deletion to be specific. I'm glad my instructions were comprehensible!

I'd just keep your Samsung 860 as your cache, and when you make a VM, in the webui, it should default to the path: /mnt/cache/domains/NameOfYourVM/vdisk1.img In general you don't want a virtual machine's disk image on the array, because it's too slow. The reason the array is too slow is because of the Parity over-head. And, afaik you can't allocate part of a cache drive to be part of the disk array (directly answering question of subject-line). The two useful configurations seems to be: image file on your cache drive (mover isn't going to touch it), or pass-through an entire drive which the VM installs onto. Either way, you're right you're not on the disk array, but you could make a snapshot-copy of the vdisk1.img file. Also, see a lot of people map a user share inside the VM, and that gets used for user data so that gets array benefits. My two cents, I hope that helps.

Another thing to consider is to lurk on the "Buy, Sell, Trade" and this forum thread (Threadripper 2! - Lounge) maybe you could find a community member willing to sell their TR for a deal? -- just a thought.

Go to Community Apps plug-in do a search for, "StorjMonitor" and grab that; the template is going to look similar but have more options. Before this will really work well, you'll want to open each ../storj/Node_*/config.json file with nano, vi (or notepad++ something which handles UNIX carriage-returns): change: "doNotTranverseNat": false, ----> "doNotTraverseNat": true, For each Nod_#/confi.json file I recommend changing "rpcPort": 4005, to 4001 rpcPort for Node_1/config.json becomes 4001 rpcPort for Node_2/config.json becomes 4002 rpcProt for Node_3/config.json becomes 4003 . . . way less network ports needed that way. Hope that clarifies things. If something is still unclear just ask.

Couple of quick searches, on web, suggests Plex isn't that ram intensive, so 8GB probably ok (hazard guess). I'm also of the opinion that RAM is cheap so go for 16GB! Easy to say when it's not my money. RAM is also easy to install, so you could just start with 8GB, run it for a while, and drop more ram in later if you need it. Hello from Idaho.

1KW is over-kill in my opinion, you probably only need a 650W-750W for what is listed. I would search through the forums, if you haven't already, for that board - just to see if anyone has any issues with it. I don't know of any off hand, for your chosen board, but the motherboards I feel have the most snags for running VM's.

One thing I've ran into, related to this topic, is that some games don't like the network share system we do with unRAID - especially if it has cheater-protection app on it. For example ARK runs fine, but if you try and use their cheater-protection ARK will go APE-S*&^ bonkers trying to load. The solution here, is the same as for Battle.net, make a vDisk for it, or passthrough an entire drive. Just FYI.

If your VM is in an HD file I would shut down the VM and make a backup of the VM. Then I would restart the VM and try and do an in-place repair install of Windows (of the same build number) and then see if the Windows VM is behaving itself afterwards. It might also be worth it to build a new a new VM, just to see if it behaves the same as your current VMs; this might give a clue as to whether the issue was the unraid/KVM or an internal issue with the VMs. That's where/what I would start off with. EDIT: if you want to start off with some less drastic measures to try here are some command line things to try (assuming Windows here, you did say blue screen): These will require an Administrator power-shell or command prompt. DISM /Online /Cleanup-Image /RestoreHealth /source:WIM:X:\Sources\Install.wim:1 /LimitAccess Where "X" is the drive letter where the ISO is located. Simply change the "X" to the correct drive letter. That command goes through the Windows msi and driver store and will replace any corrupted files - can help, but no guarantees here. netsh winsock reset

Maybe the unBALANCE plug-in might be a useful tool here?