deltaexray

Hello, Yes exactly: My array is purely mechanical, therefore when all drives are running it needs quite the cooling to keep all of that cool - to be precise: 6 Fans and 2 CPU Cooling Fans. When all of them run, the whole server is cool - literally. My idea is now, when the array is spundown, to allocate the non-mechanical to a specific fan Group i.e make the then „second“ Drive Group available as an input for the fans to cool them. It‘s oddly specific I know that but basically I have to different sets of drives, mechanical and non-mechanical and both of them need cooling but realising that with only one „Drive Temperature“ Group is not that elegant or practical. Cheers Alex

Hello again, It‘s me again. Since u can select the drives which temperatures are to be pulled for the „HDD Temperature“ Sensor/Value to then set the corresponding percentages, Would it be possible to create two groups so to speak? I know the HDD Temperature is meant to be used for all the drives and what the IPMI/UnRaid gives is limited but wouldn‘t it be possible?

Hey Simon, Thanks for the fast and more importantly helpful answer. I somehow never thought of them as zones but as of individual ports, therefore the question of why only 2 show up in the IIPMI but this makes way more sense haha Actually this will be kinda helpful tbh, given I need to cool chassi and disks separately anyway Thanks again & happy holidays ☺️ Alex

Hello y‘all, After reading almost the entire Thread/Post by now, I am kinda curious on what the „solution“ to hook up fans to a Supermicro H12SSL-NT mainboard is? I know the IPMI Plugin / the implementation of how the ipmi on the board works, only exposes only „2“ Fan Ports or from 1 to 4 so how do I hookup fans to that, so it can be controlled via the plugin? It would be lovely to control more than 2 ports for more groups so to speak of fans but I guess that is more a wish than a possibility Cheers and thanks in advance Alex

He y'all, Thank you for the reply's. As a quick update: Double checked that it isn't Parsec what it makes to "laggy" and it's the same with RealVNC / VNC Connection through UnRaid. No, there is no monitor connected - There is, however, a HDMI Dummy plug plugged into the GPU. Already checked if resolution is the problem, problems present themselves regardless of 2560x1440 or 1920x1080. That is actually a very good suggestion: I'll check right away and see if the 50/50 Split, so 4 cores for VM and 4 cores for UnRaid isn't a great split. EDIT #1: Nope, it's even worse with less than 4 Cores for the VM and more for UnRaid. Stil, it wouldn't surprise me if it the CPU itself, since it is one of the lowest EPYC's you can get. I might try to add a monitor directly to see if it is the video all the together that causes all of this. EDIT #2: With a monitor directly connected to the GPU, it's sort of the same experience, just a bit faster given it is not a networked connection. However, it has the same "running away" issue while editing. So I'm moving the VM on a passthrough SSD to let it have its own device. Cheers, Alex'

I don't know if it's the right category, if not please feel free to change ISSUES: While Videoediting, footage is running away after playing start and pause. With TS Doctor, to Edit recorded TV Shows of my Set Top Box, playing Pause/Start, footage runs away and cannot make a precise cute = skipping frames ends in standing image with 500ms of not moving. = Latency Issue? General not responsive, even after config changes. (Increasing Bandwidth) Gaming is laggy, even with low demanding Games at HD Settings with GTX 960. Parsec : Lags, drifting mouse and it feels like it is running at 10 FPS. CPU at MAX 75% GPU at MAX 70% EDIT #1: GPU Encode is between 40-70% usage. Nothing seems to throttle, there’s no indication that CPU or GPU are a Factor here. Bandwidth edited to 100 instead of 50, encoder set to 64 which is Highest Quality. Desktop / Display still look a little „blurry“. VDISK issue maybe? In general, through Parsec, it feels a little less responsive then through VNC Hardware: UnRaid VM: 8 7252 EPYC Cores, Host Passthrough MB: Supermicro H12SSL-NT 16GB (16384MB) RAM Hyper-V OFF I440fx-7.2 OVMF-TPM 3.0 qemu Primary VDISK Location: Manual, VirtIO, 200G, (Location is a single NVME 2TB) Said NVME: KIOXIA 2TB Gen 4 on an ASUS Hyper M.2 Card (All GEN 4 and Bifurcation enabled) GTX 960 Passthrough with VFIO Netwotk: br1 with virtio Software: WIN11 Images made with NTLite, removed a lot of bloat and did some tuning, however, I’ve been running the same image on two different machines with NO issues. Mind you, without parsec. IDEAS: VDISK Issue? Passingthrough a single NVME for direct and only storage for WIN11? Configuration Issue? Wrong Settings, both on WIN11 and UnRAID? Parsec maybe too? Parsec just being shit? In general, I don’t think that the Hardware is the issue here. Except for one small thing: The CPU, EPYC 7252, is a „special one“ since it is only 8 cores and effectively only 4 Memory Channels. Could it be the case that there’s simply not enough bandwidth in terms of System Memory available to have sufficient performance? ------ So this it the write up of the current state of these issues, with said VM and Parsec. Before the current WIN11 Image, I tested like 6 different ones before this one was stable. From my Point of view, I think Parsec might be the issue but also why is it so? Diagnostics are attached. Thanks in Advance, Cheers - Alex'

ISSUES: While Videoediting, footage is running away after playing start and pause. With TS Doctor, to Edit recorded TV Shows of my Set Top Box, playing Pause/Start, footage runs away and cannot make a precise cute = skipping frames ends in standing image with 500ms of not moving. = Latency Issue? General not responsive, even after config changes. (Increasing Bandwidth) Gaming is laggy, even with low demanding Games at HD Settings with GTX 960. Parsec : Lags, drifting mouse and it feels like it is running at 10 FPS. CPU at MAX 75% GPU at MAX 70% Nothing seems to throttle, there’s no indication that CPU or GPU are a Factor here. Bandwidth edited to 100 instead of 50, encoder set to 64 which is Highest Quality. Desktop / Display still look a little „blurry“. VDISK issue maybe? In general, through Parsec, it feels a little less responsive then through VNC Hardware: UnRaid VM: 8 7252 EPYC Cores, Host Passthrough MB: Supermicro H12SSL-NT 16GB (16384MB) RAM Hyper-V OFF I440fx-7.2 OVMF-TPM 3.0 qemu Primary VDISK Location: Manual, VirtIO, 200G, (Location is a single NVME 2TB) Said NVME: KIOXIA 2TB Gen 4 on an ASUS Hyper M.2 Card (All GEN 4 and Bifurcation enabled) GTX 960 Passthrough with VFIO Netwotk: br1 with virtio Software: WIN11 Images made with NTLite, removed a lot of bloat and did some tuning, however, I’ve been running the same image on two different machines with NO issues. Mind you, without parsec. IDEAS: VDISK Issue? Passingthrough a single NVME for direct and only storage for WIN11? Configuration Issue? Wrong Settings, both on WIN11 and UnRAID? Parsec maybe too? Parsec just being shit? In general, I don’t think that the Hardware is the issue here. Except for one small thing: The CPU, EPYC 7252, is a „special one“ since it is only 8 cores and effectively only 4 Memory Channels. Could it be the case that there’s simply not enough bandwidth in terms of System Memory available to have sufficient performance? ------ So this it the write up of the current state of these issues, with said VM and Parsec. Before the current WIN11 Image, I tested like 6 different ones before this one was stable. From my Point of view, I think Parsec might be the issue but also why is it so? Diagnostics are attached. Thanks in Advance, Cheers - Alex' server-diagnostics-20240722-1725.zip

From the back of my mind: I can remember something like that too. Specially with VC1 encoded Disks that get transcoded, I got a Notficiation the the image was full. But it was random as hell and only came back when trying to download movies to the iPhone. So you are not alone

As an update: I stil have some "old" VC1 encoded Blu Rays in Plex and played them on a TV where they needed transcoding. The result was, that nothing went wrong or hasn't worked. So the Initial problem stil exists

Thanks for the reply, I was busy during the day. I have the feeling I'm using a different Plex Docker build 'cause mine doesn't look remotely like yours. These are all the mounts I've been running But just as a note, that I remembered now: Whenever I do transcoding, in this case it runs purely on the cpu side of things, I do not get the Issue with the filling up of the image. Funny enough

I can relate to the Notifications/Mails thing, I get them too. and when it fills up

Thanks for the reply: It is set to /tmp. And as far as I know, it does transcode in RAM. What makes me question all of that is just the fact that when I try to download a movie onto my phone in ORIGINAL quality I get warnings from unraid that my vdisk fills up but only with large movies, with smaller ones it does work. but even if's using RAM I've only got 32GB of Ram, so why does it work sometimes and sometimes not? If you catch my drift, so to speak.

So, I don't know If this is the right Thread for it but it was quite similar to the issue, I am having currently. Whenever I try to download a movie to one of my devices, i.e iPhone or something similiar, the VDisk Image on which the docker for plex runs, fills up. Completely. paralell to that, the download fails too. The VDisk Image Size is 60GB, so when I try to download a movie with 45Gb in Size, it works. If I try to download a movie with 80GB, for example, it fills up the image, spits out warnings without an end in sight and then, when the vdisk has returned to normal size, the download has crashed. Has anyone encountered something similar like that?

It's kinda funny that I get this answer in a random conversation in a thread about MSSQL, than in a official conversation about bitwarden but I guess, sometimes u do not even see your hand right in front you - Thank you:) And yes, UnRaid is pretty much open as it gets. Which is a blessing and a course.

Reading this, I get the feeling I need to do everything in the Terminal of Unraid, right? Lile I get the whole picture but I somewhat do not understand where to put all of this? *InsertHeadscratching* I always thought Unraid is a purly U can only install what there is OS, not something I can go deeper with if you get what I mean.