ConnerVT

I have a keyring of flash drives. Each is loaded with bootable software needed to recover each of the major devices (two Unraid servers, my firewall, a few PCs) I have at home. Each has a tag on it, identifying to which it goes. I also have a ISO files for each saved on the external USB HDs I swap out to keep an off-site backup copy. (There's a few troubleshooting tools flash drives on the keyring as well). An important part of the process is to test and make sure that the restoration of your backup actually works (and you know how to do it). Too many folks just click on the "Make Recovery Key" button, burn a flash drive, then toss it in the desk drawer. You need to try it out, otherwise it is just security theater. Makes you feel good, but may not be helpful in a crisis.

4Kn and 512e does not have anything to do with the filesystem format. It is the hardware low level formatting for how sector size is mapped to the physical sectors of the drive. What is the exact model # of the drive and LSI card?

4Kn and 512e format? I'll let the LSI HBA experts comment. (Also, He8 doesn't use the Pin 3 Power Disable).

Set up syslog server to start with.

You're not wrong Jonathan. But time flies - The 1st gen Ryzen released April 2017, which soon will be 7 years ago. So Ryzen is no longer in the early adopter designation. The funny thing is, soon AMD will likely have the more stable platform, as Intel starts updating its lineup with the mix of high performance and efficiency cores, as well as the rebranding of the Core processor line and rolling out more mobile processors as mainstream. It will be fun to see how things look in a year or two.

NerdPack was deprecated quite some time ago. It has been replaced by NerdTools.

I would think that would be correct. Another piece of the puzzle is the AMD motherboard chip set, and the BIOS (and the AGESA it is built with). The 3xx series had a limited amount of NVRAM. With the AM4 socket having such a long life, the MB manufactures started to run out of space for their code, as new CPUs were released. The AGESA started to grow in size, and the pretty things (fancy screens, overclocking, fan control etc.) the MB makers wanted in the BIOS, sometimes settings and such would disappear from the latest version of their BIOS. I found using the last version of the BIOS released before the next generation of CPU released works best for 1st/2nd gen Ryzen. Such is to be expected with Linux and any brand new hardware model. Microsoft and the motherboard manufactures get pre-production information and CPU samples well in advance. They certainly sign non disclosure agreements, which is something difficult to de for Linux with it being open source. Early adopter Linux users should expect to have issues.

Looks as it was handling it. A snapshot in time of CPU utilization doesn't give much info to make that decision. Play around streaming different files, settings (resolutions/audio/subs) on different clients, and observe what the CPU/GPU usage looks like to see if all works. If you don't have stuttering and CPU threads aren't pinned to 100%, you're good.

Plex only hardware transcodes the video. The burned in subtitles and audio down to Opus is done by the CPU. The higher the bit rate of the source file, the more data the CPU needs to process.

There is really little difference between EPYC and Threadripper/Ryzen. They all use the same ZEN processor dies, with only the I/O chiplet that glues the processors together and handles connecting PCIe lanes and other connections to the physical package. First Gen Ryzen does have an acknowledged bug in Linux. My thoughts are that AMD was so focused on having Windows benchmarks that kicked Intel's butt that they didn't do enough to insure there weren't going to be problems in Linux. Understandable - AMD had nearly no data center business at the time, and Linux isn't as sexy as gaming and Cinebench benchmarks for Dr. Su to flaunt in product release Powerpoint presentations. Each generation has updated how the CPU power management is done on the die. By ZEN2 (your EPYC) they have addressed the Linux bug. If they hadn't, the AMD server market share would not have grown to its current 20%+ from a lowly single digit. I have been running a 1st Gen Ryzen 1500 with Unraid for several years. Was my first server, now is my backup doing NAS only service. I did see lockups until I set the Idle Current correctly in BIOS, then never a problem. My main server's 5700G (ZEN3) has been perfect since Day 1. The 3900X (ZEN2) the original poster is using should be sure to follow the info for Ryzen in the Unraid FAQ - Setting the Typical Idle Current in BIOS and not exceeding the RAM speeds from the charts.

The folks who own all of those datacenters running EPYC servers would disagree with you, Jorge. 😜

Hate to repeat myself, but it always depends on your needs. 1TB or 256GB as cache? All depends on your work loads. What is the largest amount of data you will copy to your cached enabled array share? I back up a few desktop systems, with full (compressed) images approaching 700GB, So 256GB wouldn't work for me at all. I also have Usenet downloads go to a cached share, but set my Media share non-cache enabled. Makes no sense to download a file to cache, then "move" it to cache for the mover to then move it to the array. Might as well have it moved to the array right after the download is complete. You need to understand your work flows, then decide what is best.

Perhaps posting in the NUT Plugin Support Thread will get you an answer?

Or motherboard power stages are marginal/failing, BIOS settings wrong. BIOS needs updating, just a few that come to mind.

https://forums.unraid.net/topic/119052-psa-on-sandisk-usbs/

Isn't this a bit like saying your car overheats when you drive on the highway, and you fixed your car by not going on the highway any more? Seems you need to find the root cause of the issue, Something isn't right with your system.

Looks as there isn't one. You can reach out to @welfare, but hasn't been on the forum for a month, and last post is Aug 2023. https://forums.unraid.net/profile/129748-welfare/

I had the same issue, going back to Halloween. I finally just turned off VPN, which wasn't my desired outcome, but it crippled my download speed otherwise.

It should be sufficient. It isn't always Can it do all these things, but more how hard you work things. The only limitations I see is I would add more memory (16GB is limiting when spinning up a few VMs) and note that you only have 4 SATA ports on the motherboard and 1x16 and 2x1 PCIe slots if you need add a HBA/SATA card for more drives.

Or you could be even worse, like me. 64TB in my array at ~55% utilization. +100TB total between main server, backup server and external backup drives on the server/network bench. 50TB of cold storage in the bench drawer. I need to update the thread in my signature. There is no doubt, I have issues. 😲

The first generation Ryzen CPUs have a bug which prevents them from going into the low power C-States in Linux, causing system freezes. So you won't find much power savings there. What are you using the RX 590 for? That is probably a good part of your power consumption.

1164 - 1064 = 100 (Sorry. I just couldn't help myself) 😝

The Devil is in the details. If you go with a 400W UPS, you will need to set the shut down to be basically immediately (say, 1 minute). It takes Unraid a few minutes to shut down - It spins up all drives, shuts down Docker, VM and other services, unmounts the file system before taking the array offline, then powers down. Always good to time how long a clean shut down takes for your system (there are Unraid settings you can tweak to allow extra time before Unraid kills processes if something needs extra time to shut down gracefully). Another thing to consider is how a UPS actual behaves vs. how it theoretically behaves. Even a momentary power sag, one which you may not even notice your lights flicker, will kick it to battery for a moment. And this alone can drop your battery by several percent. And take the UPS a couple of hours to top off. Most power events are not simple. A thunderstorm may give multiple power sags as above. Or maybe even several longer power outages, of a minute or so, over an hour. Each will delete battery reserve power. How much depends on how much/hard the attached load is. So each time the power comes back, you are not starting with a 100% charges 400W UPS, but of something much less. So my recommendation isn't really overkill. A UPS is insurance for not losing data by unclean shut down of the server. The only thing worse than not having insurance is to spend money, think you are covered, then find out you aren't. Auto start - Setting the BIOS to power on the server when power is restored won't work. The APCUPSD native to Unraid has a setting - Turn the UPS AC power output off after shutdown. If you turn off the AC power, it won't come on once main power is restored, so your server won't start. Or you leave the UPS power on after shutting down the server, so your server won't start. You will need to power the server back on by another method - via USB, Ethernet, etc. Most basic UPS do not have such a feature. APCUPSD goes back to the very early UNIX/Linux days, and was written for APC brand hardware. Most major manufacturer's of UPS support APC's command set and communication protocols, so work with APCUPSD. Once you find a unit you think to buy, you can do an Internet search to confirm. Another popular program is NUT which I believe can utilize other brand's formats/enhancements (disclaimer: I've never have used it).

SSD cannot be trimmed when in the Array. While they can be assigned to the array, they may degrade over time and give you errors. Not a good solution for a long term plan. What I have done is create separate pools for three of my drives: 2TB NVMe, 2TB SSD and an ancient 120GB SSD I've repurposed many times since 2012. 2TB NVMe- Used for system and appdata Shares, as well as cache for most of my other shares (except Media) 2TB SSD - Used to support VMs. It has my domain and iso Shares. Fast storage for VMs and I like having it on a dedicated drive. 120GB - Used for my transcoding working directories/shares (Plex/Handbrake/ffmpeg). I've had this drive forever. I lost count how many systems it has been used. Too small to use for much else, so I will keep trying to get it to error - something it has never done (according to SMART). What's great about Unraid is there is no "Right Way" to do things. Go with what works best for your needs.

Most major brand consumer UPS will do most of what you ask. For a system that is (actually measured to run) 300W max, you are going to want a good sized unit. 1000W at a minimum, 1500W would be better. A UPS is not meant to keep a server running indefinitely, but rather to ride out a few minutes of power gone time then gracefully shut down the server (which again will take a few minutes to accomplish. Ideally you don't want to run your battery down below 50% charge/run time. It is bad for battery longevity and it takes *much* longer to charge than discharge. The only thing that is difficult to implement is "Auto boot". After power goes out, you then shut down the server (power off state). Depending what else you may have connected to the UPS (router, modem, network gear, etc) you can then decide if you wish to keep the UPS power output on or off. (Leaving it on continues to discharge the battery, even if everything connected is powered off). Once the AC power returns, you want to power up the server. There are many things to consider: How charged is the battery? How likely will the power stay on (think power blips during a storm)? Surely there are other things to consider. So you need something smart enough to decide it is OK to restart the server, then turn back on the ac output of the UPS, then power on the server. You won't find that in any not "fancy" UPS. This is a snapshot of my APC RS-1350 MS looks like. It powers the two servers in my siggy (the 1500X is sleeping), a mini-PC firewall, a Raspberry Pi, external HD, and two small network switches. With the 1500X still asleep, max power at shutdown (which will spin up any Unraid drives that are sleeping) reports a UPS Load approaching 200W. (This is why I questioned your 300W max earlier. Don't go buy PSU calculator results, they are looking at momentary peak surges and to sell you a bigger, more expensive power supply).