maxstevens2

Sorry for a late reply, you can check the frequencies using: watch -n 0.3 "grep Hz /proc/cpuinfo"

I did not have to do anything, strange

I have been using ISCI, and it shows the exact same! No issues speed related or anything so I think it's all meant to be!

I would say to use FileIO when using a disk that contains data, and you don't have any other disk that you can spare for the function of ISCSI. I would go for block when you are able to have a complete drive available for this function. My case: I have both, 1 4TB HDD completely available to Windows, and a 300GB image on my SSD for game storage (for VM and my gaming rig).

Bwoahh, using a block (or just a whole disk) is so insane! Getting these numbers in combination with a WD RED PRO!:

It actually worked flawlessly! The image is now resized, but it also found out the actual image size itself! I typed in 450GB but it probably saw the actual size was 483, so it changed. No data overwritten and all good! The function is based on this: Do note to disconnect the .img first (remove it from the initiators list) (after shrinking the volume in Windows!!!!!!!!!!!!!!!!!!! Hope this also helps others in the future.

I will try re-adding it Otherwise It doesn't matter if I break it, already made a copy of the data so will see what will happen lol.

So I tried to resize a volume, but it still reports at 2TB disk, any idea how to decrease the size? Size 'before' was 2TB But it still reports it as 2TB drive: Also; I am not decreasing the size for further usage, I am only decreasing the size to then move the files over to the same disk, but then added as a block. So basically from disk 1 -> Cache -> disk 1 as a block again (manually transfering data)

Unassign, so that will mean creating new config in my case. Because I am trying to switch over to mounting a complete disk. Thanks for the help @ich777!

Just a cute little hint: if your image is over the disk size you are trying to use mover for, then it will break /mnt/user and you need to restart your server. And yes I learned this the hard way by moving a 2TB image over to a 1TB SSD (from disk to cache).

I am planning to do this today, due to my 4TB WD Red Pro performing very good in reads, but writes seems to cut off goard some programs in Windows. Do I need to disconnect the disk from Unraid (via tools -> New config)?

I think it depends on the usage, if you are gonna do a lot of writing, then you can mount the drive, however you can use a fileIO image on the btrfs-RAID setup. if you are gonna be using 1gigabit connection, I would recommend sticking with FileIO, 2.5gbs and above use the disk, which I think cannot be used in a raid-config. Hope this helps!

Well I don't think there actually is. Just use this to guide a bit:

I've not seen plex do it. What are your other dockers? Or do you have a VM stored onto the drive? Whats te filesystem type?

And based on these variables, I have to make my decision 😜 . But since I am pretty new to ISCSI and the disk is also pretty new (the server disk is a 4TB WD RED PRO), I have to see what I am gonna do with storing data an such (I am not new to unraid btw, got already 4TB+ of data). I will try and see how it will all end up. It does. Thanks so much for all the help!

Just tested this myself, it does not. Maximum read performance comming from the disk!

Well the weird thing is (that I saw with iotop on unraid), see screenshot after this quote: "Actual Disk Write 0 B/s", while on the picture before the quote it was running at 165MB/s (while probably exhausting buffer). Thanks so much for the replies and help! Does this not use RAM as cache/buffer?

Yes, 2 Tb FileIO image. I've noticed.. I had 15MB/s write speeds while testing I need to think how I am gonna realize this. The main focus of the project is to force out my PC's HDD, so this might be the best option then. I guess performance will increase when using the whole disk (block) too. Will this be close to 1:1 performance (although I am already pretty happy with the performance)? This is not the same for READ performance right? Because I don't think I would mind it then. It's mainly made for gaming (and other things like game recordings). Okay good, was already scared I would lose data. Are there maybe any tweaks? I could limit my network card to 2.5G or 5G (but 2.5G means ~300MB/s which still is above the 200BM/s limit of the disk).

It seems to write data onto the image, write onto the actual disk, then gets hung up due to the upload speed being too much, then de disk is ready (goes to 0MB/s) and then the upload goas upto 400MB/s again.

Thanks for your reply, At that point I was installing a game, so I guess on heavy writes yeah. Is this a very bad thing? And if so, where is this issue located? Disk, Network, CPU? Edit: In addition to that I also noticed this happening while copying ~8GB files: Notice the 'hills' created, swerving from 15MB/s upto 375, while the disk can do around 160-200MB/s It seems to be a bit 'stuttery' pushing all the data trough:

So yesterday I got it all working, and seems to work quite good, except that I get these messages in my log: Any idea what this is?

Little update: I just hung my notebook with an MTU of 1500 on the 10GbE adapter (but my notebook only supports 1 GbE so it went to 1GbE) Which worked flawlessly. Will try to see if changing MTU on my pc makes it work! Edit: Yep, it was the MTU. I set it to 1500 and instantly worked! Performance is perfect!

Yes I do Well I guess but also not. If I disable the bridge, this means VM's can't connect to it right? Which whould mean I can't have internet over this due to my firewall being a VM. Honestly I don't think this would be a big of a deal though as I don't mind using the 10 gig link only for Unraid -> PC usage. I will do this later, because then I have to stop all the VM's and such which I can't do right now.

Btw. The MTU of the 10 gigabit adapter is set to 9000. Is this a bad thing maybe?

Yes 192.168.179.XXX is the main network (connected via 1 Gigabit), this works flawlessly and reaches up-to 1gigabit without a problem (read and write). 192.168.181.XXX is the 10Gig lan network, which is a different network adapter than the other. More here: Later today I can make a picture to show a bit how the network is set-up (the 10-gig side is pretty simple due to it being peer to peer). Yes I did, it connects but then the whole ISCSI service just freezes in Windows when trying to disconnect or look for the disks. (I connected to 192.168.181.253 in my case.) This is a picture with it now connected to 1GbE: