ConnerVT

No matter which version of Unraid you are running, you need to select the correct version of Nvidia driver. This is due to how Nvidia fractured their drivers, in support of new Products and making others End Of Life. Fortunately, the Nvidia Plugin allows you to select the proper one (at least for every version of 7.x.x). At the time of this post: v580.x - For Pascal (such as P400) and older, though some Docker containers no longer support cards older than Pascal. These are all EOL. v595.x thru v610.x - Turing through RTX (except RTX 50 Series) Open Source Branch - RTX 50 Series

New trial needs new flash drive.

You're a good man, Charlie Brown.

Also, I would get rid of the 3 3TB drives in the array, and instead of dual parity just use one of the 14TB drives and change the second to be a data drive. Dual Parity is a bit overkill for 7 total drives (Parity is not a backup). So one parity drive for 4 drives is good, you will have an additional 5TB of storage, and will reduce power consumption, heat and points of possible failure.

Two thoughts came to mind. -- Check that there is no other drive activity taking place on the array, except for the parity check. It doesn't even need to be something the server originates, another system using the array for a network drive could be reading/writing. -- I see you have a mix of 14TB and 3TB drives. How far into the parity check was it? if it is near the end of checking 3TB, this is likely the slowest speed the parity check sees. 2 things to remember - Drives are their fastest at the first sector, slowest at their last; and (as long as the drives are syncing up nice) the parity check will only run as fast as the slowest drive in the array. Once the 3TB are done being read, I suspect speeds will jump back up significantly.

Sorry, but I disagree with this. Unraid is not a turnkey, take it out of the box and use it toaster. It is a highly configurable OS, an ecosystem which can be used as a basic NAS to a full blown home lab setup. Things which are highly configurable require an understanding of how they work before using. Not doing so is certainly a recipe for painful lessons. There are quite a number of tutorials, both written and video, readily available on the Internet which explain things from the basic configuration of arrays and pools, setting up Dockers and VMs, to much more complex things people do with Unraid. This support forum is frequented by people who use Unraid and share their knowledge without accolade or compensation. I find these folks to be some of the most helpful, especially compared to other support forums I have visited over many years. Other than your posts in this thread, I only see one other post you have made on this forum (with this Username). The author of the plugin replied to you the same day.

I have seen a number of cases where wonky transcoding or playback issues are due to a corrupted codec file. You can safely delete the Codecs folder from your Plex appdata. Stop Plex, delete the folder, and restart Plex. It will create and repopulate the folder with fresh codec files. I have had several folks post feedback they have done this and it resolved their issues. Also be sure to test with a dedicated PLEX client, and not the web browser. The web browser is well known to have issues, especially when switching resolutions or transcoding bandwidth. Switching off subtitles is also a must, as subtitles will most times switch to CPU transcoding.

I have been following several threads here for weeks about this. Answers have been very ambiguous about it, leaning toward it does count as additional drive towards the license limit.

Both of those are issues with these applications, not with Unraid. Best to work with the folks supporting the applications, either in the docker's support thread or with the application support itself.

I am pretty confident that currently this counts as an additional drive towards the drive limit for Basic, Pro and Starter licenses.

You forgot one important fact - It counts a a mounted drive. Fine if you are running Pro/Unleashed/Unlimited license. I have 2 servers. One Plus and one Basic. The Plus has 11 of 12 available drives populated, the Basic maxed out at 6. So no NVMe boot for me without spending some money. I have enough early tech flash drives to last me as long as I live (I'm old). But then, If I lived as long as my stash, I would be immortal as I have yet to have one fail in the 5+ years I've been running Unraid.

Not certain, but would guess this is where these three would run: Basically, the first is run after each container backup is done. The second is after all containers are backed up (and maybe the flash?). The third is the last thing run before the plugin exits. You could test this by creating some (almost) do nothing scripts, configure them in the backup plugin, and running manually. Then check the logs.

Perhaps update your script to only run when the passed container name is for your desired script, and just exit the script for the other containers?

I will echo what @Hoopster said about using a PiKVM or some other remote management solution, especially if you don't have someone back at the server when issues arise. While my server runs many months at a time without issues, Murphy will always show up at the most inconvenient time. I have a PiKVM that allows you to power on/off the server. Boot your server in GUI mode, and you can access your server from a browser. You won't be wishing to do much work this way, but can restart the server, go into BIOS, do whatever you would as if you are sitting in front of the server. I secure this with a Cloudflare Tunnel and their security/authentication tools.

smh. I paid $75 for Plex Pass in 2018, when they ran a sale at the original launch pricing. Doubt I would pay 10x that today. Especially as they are far behind updating several of their client apps and ignore issues they've had for years.

Wow. Looking back through this thread, it is shocking to see street retail prices today vs 3 years ago. The 2TB Gen3 NVMe cache drive in my main server I bought for $70. The cheapest equivalent drive (similar brand/specs) now is 4 times that price.

@TheExoticBeard --- Both cards are now End-of-Life by Nvidia, so no more driver support after v580. Your Pascal based P400 is on a limited usable life, but eventually either the available Nvidia drivers or Docker container Nvidia runtime code will no longer support it. Clock is ticking on that one. The GTX 970 is Maxwell based, and having been released in 2014 (with a very short, 1 year run before Pascal was released) really has little use with Docker containers. What I would suggest doing is: In the Nvidia Driver plugin, select the v580 driver. This will prevent you from possibly updating to a later version driver. This will keep your P400 happy for use in Docker containers. In Tools > System Devices, pin the GTX 970 for use in VM. This will then keep the Nvidia driver from even seeing the card and complaining. You can use the GTX 970 with VMs. Start shopping around for a replacement for your P400. The time is coming when it will no longer be supported at all. I recently replaced mine with a RTX A400. They can be had on eBay for around $200.

PCIe slots really don't use power themself, only the card plugged into them does. The M.2 slot is the same, only use power when a card is plugged in. Generally, most all motherboards containing the same MB chipset basically use the same amount of power (unless they have add-ons, like 100W of LED bling). Compared with the power used by the CPU, drives, fans, etc the MB is a very small percentage of power used. It is better to select a board based on the form factor you need, number of on-board SATA connections and PCIe slots required, and a brand you trust.

Post Diagnostics

Not for me (sorry for the funky screenshot - only way to get everything to fit.)

@ich777 - Something to note, based on the posts above: The text in the Nvidia Driver plugin (at least in my 7.2.4) still reference the Driver Update Notification, but this is no longer available as an option to select. I only see DISABLE GPU-BASED FILTERING and UPDATE & DOWNLOAD (as well as radio buttons for selecting driver version).

By traveling back in time, to January 2026. It was an option in the Nvidia Driver plugin in Settings then. ich777 has made a number of changes since , as the proper version of the driver has been complicated between all of the recent Unraid releases and requirements for the latest RTX cards and the deprecation of the Pascal chipset GPUs. The setting likely has caused more problems than good. In the vast majority of cases, the "latest" driver won't make a difference for people. It usually addresses niche things that may affect the latest AAA video game or high end video processing. If you are having this kind of issue, the automatic notification isn't much help, for you would likely already be looking for it and usually isn't an application running in Unraid. Removing the option from Settings keeps those who update everything immediately, needed or not, from inadvertently updating then finding something broke in their server.

^^^ This Also be sure to do your testing on a real streaming device (such as AppleTV 4K, Roku, etc) and not with the web browser, and turn off sub titles. Web browser is known to cause issues when switching transcoding values, and sub titles will (in many cases) result in CPU transcoding.

ich777 trying to keep up with Unraid releases be like:

I had that also, for several weeks. Then early this past week, it started working correctly. Believe it was happening on both servers (unsure of the backup server, but believe it also did this). Main server has custom image, backup has stock PC-Q25 image. Main runs 7.2.4, backup 7.2.5-rc1 that I updated (may correspond to when things worked again?)