OrneryTaurus

I think the server weight itself is about 50-75 lbs before adding anything extra into it. Adding a full system of hard drives, you'll probably increase it to 150lbs. I would be more worried about the rails and sliding the server out to service it with the existing rack you have.

I have concerns about this rack being able to support the weight of a 4U server. You'll want to double check the maximum loads the rack supports. It might require permanent installation to increase the maximum load it can handle.

I run my docker containers in br0 purely for my sanity. Ran into a networking problem long ago involving docker containers and just have done it that way since. I know the security implications, and that I can re-map a docker container's ports if needed; but I find it easier to just assign an IP so I don't have to worry about remapping.

Hey there. I also have a CSE847 server case. I would say you probably need something a bit stronger. An empty CSE847 weighs around 50lbs / 22kg. which is already a quarter of the weight for the maximum load when ground mounted.

Yep, tried using ich777's RCON Plugin and it won't connect. Also tried RCON-WebAdmin docker container and that wouldn't connect either. Found ARRCON and that allows for connectivity. Edit: Found the same symptoms and potential workaround here:

RCON doesn't seem to want to work on the PalWorld docker. RCON is enabled in the PalWorldSettings.ini, and I'm using the rcon plugin with -H IP -p 25575 -P AdminPassword. Am I missing something? Nothing else is using the port. EDIT: Looks like the rcon plugin and another docker, RCON WEB Admin do not work with the PalWorld RCON setup. I found a third party app that allows for connectivity. RCON itself is working.

Well shoot, there goes my theory. Someone else on the forum has the same board but their network drivers weren’t loading. Previously I had the on board NIC driver screw up on my X10 board which would make it appear as if it crashed.

Are you running your server headless or do you have a monitor/mouse/keyboard hooked up to it? Is it accessible physically when it crashes?

Depending on how the official docker is setup, you would need to change a variable named VERSION to one from here: https://hub.docker.com/r/plexinc/pms-docker/tags/ or you'd need to change the repository line to include :version (version being the tag you want from the hub.docker.com link) so if it has a VERSION variable, you'd change your entry in it to 1.32.6.7557-1cf77d501 If you needed to change at the repository line, and it says plex.hub.blah you'd put plex.hub.blah:1.32.6.7557-1cf77d501

^ - Some notes from your list.

Having a GPU for transcoding, you might want to downsize your CPU, unless you feel you need all those cores.

The only thing I foresee a problem with is heat dispersion. I also noticed you didn't list a cooler for your CPU, which one are you planning on going with?

Based on the pictures, I don't see a mounting point for a desktop PSU except for on the right side. The size of the hole when you remove the extra panel looks like it could. Looks big enough that even if it doesn't properly fit one, you could have one sitting there loose in it.

That chassis does not support desktop power supplies. Best to buy at least one PSU from the manufacturer so you can utilize all the available connections it's intended to have. Server grade power supply setups are intended to be hot swappable when you have redundancy to reduce downtime. One thing to watch out for is what kind of power is needed for these PSUs. For my server, I have https://store.supermicro.com/1280w-1u-pws-1k28p-sq.html which have a rating of: 1000W: 100-140V; 1280W: 180-240V. Meaning I can only pump out 1000W of power when hooked up to a standard 120V / 15A circuit and plug. In order to get 1280W, I would need to have a 240V / 20A circuit and plug. Thankfully it supports both but that may not be the case for the 2000W PSU, it may require the 240V / 20A circuit and plug, this of course is providing it's a US plug setup.

Bought a similar case from another seller. Love it. Works great with a Tesla P4 for transcoding too. I was shocked that it runs pretty quiet even with 7 fans inside.

looks like you named the key NVIDIA_VISIBLE_DEVICES GPU BIND for NVIDIA_VISIBLE_DEVICES instead of NVIDIA_VISIBLE_DEVICES for both name and key. Might be contributing to the problem. Your 750 Ti is pretty limited on what it can be used for when it comes to transcoding. Looks like it's limited to 5 streams and only h264 content.

Unfortunately you will have to find an alternative method of power monitoring.

Short answer: do not mix ECC with non-ECC ram. Long answer: If your motherboard supports both, there is the likelihood it would have to be unbuffered ECC ram and you would need the functionality in the BIOS to disable ECC functionality all together. You would not be able to use ECC functionality on some modules and not others.

That In-Win case is definitely half height slots as the design is exactly like a supermicro 36-bay. I would go with your original case.

It might be an issue with PCI Express lane bandwidth: I see this note in your motherboard manual: * When using PCIe x1 slot (PEX1_1), PCIe x16 slot (PEX16_SB_1) will be adjusted to x1 lanes. This to me seems like as soon as you use one of the x1 lanes, it will make the PCIe x16 a x1 lane as well - Really strange if that is the case.

That's a beefy chassis. Expander backplanes should allow you to use a single cable per backplane to see the drives IIRC. You will need at least 2, maybe 3 cables to connect to it (not sure if they split the front into two backplanes or not). I was looking at something similar from the same place for more storage with full height slots. However, I personally picked up a 36-bay supermicro superchassis and upgraded from a P2000 GPU to a Tesla P4 instead ($800 all in for the two). Only issues I can think of would be related to shipping and availability of replacement parts / racking the damn thing.

I used to have Dual Samsung EVO Plus 1TB drives, never had an issue. That being said, I did opt to switch to WD Red NVMe drives instead when I wanted 2TB instead of 1TB.

Hi @ich777 - switching the PCIe port and it's now working with the latest driver. Plex transcoding is also working. Thanks for the kick in the rear I've attached diagnostics in case you needed to review. smalls-diagnostics-20230510-1209.zip

Hey there. Here are the diagnostics for the system currently running in case it provides anything. The card was removed. I'll install the card again and run new diagnostics shortly. smalls-diagnostics-20230510-1146.zip