Hoopster

Here is an example of what @jonathanm said above. If you have Deluge configured to save to a /data location, which in your case appear to map to an unassigned device, you must have the deluge app configured to save there as well. The spelling must be exactly like it is in the setup including capitalization. /data and /Data are not the same location. Here is an example: The above saves downloads to incomplete and complete folders in the specified /data location. If something is mis-configured, it will be saving to the docker.img file causing it to grow a lot. Has it grown a lot recently from a smaller size to now 75% of specified size? 20GB is the recommended docker.img size, so you are at about 15GB now? You can also click on the CONTAINER SIZE button in the Docker tab to see if any of your containers appear to be unusually large. This could be an indication they are mis-configured and storing data in docker.img. Check all of your docker containers which download, transcode or otherwise store data to make sure some error is not causing their to do their thing in docker.img

^^^^This First of all, you want to avoid transcoding locally if possible and direct play your content. This means the content must be in a format the clients can handle without asking the Plex server to transcode. Playing remotely or to mobile devices often causes transcoding because of bandwidth and other constraints. When the Plex server must transcode, the recommendation is for 2000 passmarks per 1080p stream. unRAID needs about 2000 for overhead operation to keep the system running normally, so, in theory, your CPU should be capable of 4000 passmarks just to transcode one 1080p stream. You CPU has a passmark score of 3491. With a score that low, it will struggle to transcode a single stream and the CPU will be at 100%. Your CPU not only has no overhead, it does not technically even have enough power to do what it is being asked to do. Plex also recommends a single thread rating of at least 2000 for 1080p and your CPU has a 1608 single thread rating. I think you see the problem now. If you don't want to upgrade the CPU, you can get an Nvidia GPU (AMD is not officially supported by Plex) and let the GPU do the transcoding with the Nvidia Driver plugin.

Yes. 5.10.24 is the necessary interim step to upgrade from LTS.

Looks like they have increased it to 3 now; used to be 2. The Professional and Server GPUs are unrestricted. https://developer.nvidia.com/video-encode-and-decode-gpu-support-matrix-new

Noctua estimated that the i9-9900K/Xeon E-2288G could require up to 220W when under very heavy loads in Turbo Boost. They spec their CPU coolers a little differently than most manufacturers and don't base recommendation purely on how may watts of heat they can handle; nevertheless when I wrote them about my E-2288G and recommendations, they replied that they had seen it draw up to 220W in a very heavy AVX workload test. That just seems nuts to me. I think the larger PSU will help; especially since a "generic" 265W PSU is likely a bit under-powered with your new CPU.

I would suspect this may be the issue. I have the Xeon equivalent of the i9-9900K (E-2288G) in my server and when doing multiple stream transcoding work, it can require some juice. Depending on what other things are going on with your system, of course more wattage from the PSU is required. I seriously doubt your CPU is drawing anywhere near 244 watts unless your have some heavy AVX transcoding jobs and the CPU is going into Turbo Boost, nevertheless, I would probably not be comfortable with anything less than 450 watts in a typical server PSU and more if you have graphics cards for VMs and a lot of drives in the system. A PSU should also be over provisioned and not just capable of providing what the maximum draw may be. PSUs tend to lose efficiency over time. It is a good idea to have 20-25% more capacity than anticipated maximum demand. My 600W PSU is probably overkill for my system, but it does give me room to grow without wondering if I need to upgrade the PSU.

I get that it is a new feature of the forum software and the forum developers think it is an improvement, but I do not see it as such. The result is some active threads totally take over the first screen(s) and it takes a lot more paging through the "Unread" screens to see which forums have new posts. Same with "Content I Posted In" I don't really need to see every new post in a summary, just the fact there is a new post in a thread is enough to let me know I should check it as there is an indication in the thread where the new posts start.

UnRAID loads into and runs from RAM on every reboot. Any changes you made to the filesystem (outside of /mnt which is the array and /boot which is the flash disk) will be lost on reboot as those changes only existed im RAM. Installation of plugins, system configuration and modifications to syslinux write to the flash drive so those changes persist. If you want the creation of certain folders in the file system to be persistent, do so through a startup script or modifications to the "go" file on the flash drive.

I have the EnableInsecureGuestAuth key on 3 of the 4 and it is set to 1. The one without the key can still access the server by name from Network in the File Manager. The one way in which this PC is different is that I have drive mappings direct to several shares for my wife's convenience.

Or setup a VLAN for docker containers, or try a different NIC with its associated bridge network - br1 for example, or look at the last post in the linked thread and see in any of the suggestions there help you.

I’ll check them all when I can (not home now) but I know it exists on two machines as I added it about a year ago as that was the needed fix after some past Windows 10 update. I have also disabled SMB 1.0 on them all and run WSD on the unRAID server. Sent from my iPhone using Tapatalk

You have Plex Pass and have enabled hardware acceleration in Plex server, correct? I assume so since it is mentioned in the guide. but we just need to verify this. I use the Extra Parameters method, but, the device method should also work.

And I have nothing but Win 10 Pro PCs (4 of them) all on 20H2 and none of them have any problems at all accessing the unRAID server by name with nothing but guest access. I have no users other than root on the unRAID server. That's what makes this all so maddening. It seems to be a total crap shoot and there is no solution that seems to always work reliably.

Actually this is a Micro-ATX board with 3 PCIe slots and an M.2 slot. Perhaps the SuperMicro X11SCA-W would be more to your liking if you really need a full ATX socket 1151 board and do not need IPMI.

You cannot use a discrete Nvidia GPU simultaneously in multiple VMs, but you can use an iGPU simultaneously in multiple docker containers.

In theory, yes. I have not tried transcoding in multiple Media management systems simultaneously; however, I have successfully used Plex and HandBrake (added /dev/dri to both docker containers) at the same time doing media transcoding.

I had forgotten about that episode. Yes, I generally keep my shirt on so as not to scare any women, children or people with sensitive eyes who may be in the vicinity. You could have chosen a different thread for your example in order to preserve what little dignity I have left! 😁

unRAID 6.9.0 RC2 (latest available) supports Nvidia GPUs for use in dockers for transcoding. See below for details.

How many drives do you need it to hold? Some other options worth considering: U-NAS Has 2, 4 and 8-drive cases that are fairly compact and take Mini-ITX motherboards. My backup unRAID server is in the popular Fractal Node 304 case which can accommodate up to six drives and a Mini-ITX motherboard.

$1500 x 8 for the SSDs? Yeah, a mere $12,000 for 64TB storage + $1000 for the PCIe card. We all can afford $13K just for storage, right? I have only 32TB storage, so I could do it for the bargain price of $8.5K (got to add one more for parity). One day, perhaps a pure SSD array will be reality in unRAID.

Yeah, that should do it. I assume that is in your 'go' file and /tmp/PlexRamScratch is your transcode host path.

How are you doing this? Just setting a RAM limit on the container as illustrated above does not affect the RAM Plex transcoding can use. That only has an impact on the Plex application itself. If you are not limiting Plex RAM transcoding by setting up a RAM disk, Plex transcoding can use ALL the RAM on your server and cause it to crash; especially if more than one transcode is happening simultaneously or a long transcode is taking place.

It should be if things are properly configured, unless you have a really high number of docker containers or if you are running multiple VMs. When things start to slow down on your server or after it has been running for a while, check how unRAID memory is being used by looking at the Memory graph in Stats --> System Stats tab. Having a lot of cached RAM is not an issue; that's just how Linux works as cached RAM can be allocated as needed. If your Used RAM is consistently high, you need to investigate further why that is happening.

If it is caused by one or more docker containers you can check which ones are using the most RAM by turning on the advanced view in the upper right corner in the Dockers tab. You can also limit how much RAM a docker container uses by adding the --memory= parameter to the docker container Extra Parameters. Here is an example of one of my containers which I have limited to using a maximum of 4GB RAM:

I have updated firmware on my USG, two switches and three APs many times through the UniFi app with no problems; however, I don't usually do it immediately after a new firmware shows up. I usually wait a couple of weeks to see what feedback is in the Ubiquiti forums. I have three devices with pending updates right now that I have chosen not to update yet as current firmware is working well.