In Europe, we can handle 32A on a single plug, so I just use a powerstrip with the C14 connector, to extend my UPS.
I’m nowhere near pulling that, so I’m safe. I’m at ~3A continiously, so I have litterally no reason to worry. Plus all the powerstrips are in spec for the country I’m in, so that’s another ‘yay I’m good’ thing.
The R740 should be able to boot from an NVMe drive (whether this is M.2 or U.2).
RAM speed support depends on the CPU. From the top of my head, the R630 doesn’t go further than 2400MT/s, neither do more than threequarter of the CPUs that can work in that machine.
You can put in DDR4-4000 if you want, but the machine will downclock it to 2400, because it can’t go much faster than that.
And why wouldn’t you want to go with ECC REG? It’s cheap, there is plenty of it and you can stuff in 3TB of it. Non-ECC can only go to 64GB I believe.
Why do you want to import VMs from a different hypervisor? Why not build an ‘golden image’ on your Proxmox machine, and create a template out of it.
This is basicly how I create my new VMs. I keep a relatively updated ‘golden image’ in my library and clone that to a VM as many times as I need. To be fair, I’m running ESXi. I have no idea if Proxmox even has this as a feature.
What are the best features of this board
The fact that there is an EPYC CPU on there.
if it was yours, what would you use it for?
ESXi server, combined with at least 256GB RAM.
there is a very annoying squealing sound coming from near the processor.
Welcome to the world of capacitors and chokes.
Basicly: The power delivery to your CPU is the squeely bit.
M.2 to SATA ports are available. I’ve seen more people build with those. But I have no idea about compatibility though.
Never used such a thing myself either. I have no use for such an adapter.
Those modules use some weird memory though: so-dimm ECC
SO-DIMM ECC isn’t weird though. It’s been out there for a longer time than you might think. Some professional laptops (ZBooks, Precisions, ThinkPads) have support for these.
I completely read over the part where you mentioned ‘ZFS’. Yes, IT mode flashing is better for ZFS.
I try to flash firmware from iDRAC web interface I repeatedly get “upload failed”
Then try to do smaller steps in firmware versions. iDRAC does firmware version checking to make sure you don’t upload the wrong one.
so can I just buy another two identical sticks?
Yes.
Do they have to be exactly the same
At least the same ranking (2Rx4), the same size and the same speed or faster.
And is 4x 16GB optimal configuration?
For one CPU, sure. For two CPUs, sure. It all depends in which slots you are sticking them. The proper RAM layout is on the inside of the lid. Check that for the proper installation of RAM.
for some reason ESXi just can’t see it.
I Googled with Bing
You what now?!
Try to use the APC Firmware updater. Your firmware might be stuck and the way to go is to update the firmware. This might be a little hard to find, because APC recently did a dick-move and threw all their NMC firmware utilities behind a MASSIVE paywall. If you really can’t find the tool, I can link you one. Send a DM if you’re interesed. I think I have a tool for the AP/9630/AP9631, but I’ll have to check to be sure.
is there a good solution for this?
Get a ODD caddy to HDD adapter, put in a proper quality SATA SSD and use the cableing of the ODD to connect it all. Install the OS and boot from it. I’ve been doing this to many servers over the years, also in my current R730. Works splendidly.
I was thinking of using a PCIE m.2 adatpter
The R720 and newer R730 do not support boot from NVMe. So without some Clover bootloader magic, you can’t boot from this.
I am also running into problems with the lifecycle controller
You can update firmware via iDRAC itself. If you can access iDRAC with a web browser, just upload the Windows .exe file in the ‘file upload’ section in iDRAC.
Oh, and if your firmware of your iDRAC is still on 1.x.x.x, DO NOT UPGRADE STRAIGHT TO THE NEWEST VERSION OF 2.x.x.x. This will brick the controller. Just take your time with updates and just go slowly through every other firmware update available, until you’re on the latest update. I really can’t emphasize this enough.
but it could be a fun proof of concept?
Having gigantic clusters of small ARM computers, has been around almost as long as ARM is on consumer hardware. Definitly since the very first Pi 1. Loads of people have constructed many different configurations with all kind of different hardware. Loads of people here even run K8s and other stuff on their ARM cluster.
Tell me why I shouldn’t do this!
If you have no use for it, why bother? Only if you want to learn from it, it can be useful. But still, you really have to think about what you want to do with it.
Oh, and power reasons. If you’re in the USA, and your power is cheap, then why not. If you’re not so lucky with the powerbill, then just get one big system (x64 or ARM) and just run stuff on there.
Nope. You need to buy a seperate case for this and throw everything over.
ever since I put it in the server gives me a fatal error on bay 1 drive 0
Some SSDs cause this. Don’t know for sure why, but all the Samsung consumer SSDs I’ve put in servers over the years, never triggered those warnings. Maybe for your next round of SSDs, stick with Samsung?
Mostly my own eyes… /s
I run Dozzle as a container on my host and I use the command ‘docker stats’ on the CLI on the dockerhost for in-depth stuff.
And to answer your own points you made:
I have nothing against Windows, so don’t get me wrong. But there are so many better ways to do stuff, also where you don’t have to pay for licensing.
I did a BIOS power tune recently, and that shaved off 14 watts on idle.
So that’s a thing you can also do.