Got it. I know people have used the USB SSD for rpi4 too, which is probably fine, but if you want the new quicker route. pcie might be worth getting into… https://www.tomshardware.com/news/hatdrive-m2-hat-for-raspberry-pi-5
A lot of NAS that you buy now have “containers” where you can side load apps like plex or jellyfin for your you videos vs just installing on the main rpi OS.
Do you plan on doing anything special with the NAS (like containers) or just straight up NAS storage services?
If you are doing containers, I would probably say get the 8GB and get the 5 and use for NAS.
Actually, it looks like there is a setting to avoid the warning all together: https://www.raspberrypi.com/documentation/computers/raspberry-pi-5.html#usb-boot-and-power-supplies
Thanks for that feedback. I also understand and have read that you are basically taking a bunch of tech that is it’s own tech in it’s own and then openstack glued it together (like you are stating).
I also see that there are many different projects in how you can deploy this too. Some say “don’t use ____ if you want multi node” and some are like “this _____ is the easiest to run”. Did you just use whatever Openstack suggested to install or did you use one of these other wrapper projects that helped too? I understand that some of this may have changed from 6 years ago though.
Don’t “need” to, no. I can put it on top, so that is what I planned to do to POC, but if I like it, I would rather not have software/resources taken up on top of resources. But I am asking if anyone has already done this either direction and their experience.
Yeah, but also remember that if there is one thing history taught us, it is that a non popular voted president can win still. Many of us were in disbelief when it happened with this guy.
Also, don’t underestimate the fact that people may disagree with Trump, but they only have to disagree/hate Biden more in order to vote for this guy.
So yes, I do agree that his polls may be increasing or rather, I never underestimate that they could be.
Anyone have any theories on this? Is kbin operating with faster servers, bandwidth and etc? I am sure the code is growing/learning on how to optimize too, but I am curious. I do plan to review issues to watch/track/contribute here:
https://github.com/LemmyNet/lemmy/issues
https://codeberg.org/Kbin/kbin-core/issues
I don’t have experience with the NUC for my own homelab and kubernetes, but the main things I will say with any kubernetes is the RAM. I have 4 servers on VMware (just to get experience there) and then have Rancher (like Openshift) across them (both a management server and then a 3 VM cluster). Every resource is important, but the RAM is what was eaten up before anything else for me. I have Lenovo Tiny m910q (x 4) with 64GB and that allows me a full on cluster and then some with other VMs too.
Sounds like you made a good choice going with a NUC and 64GB ram. You just may find you want to add another or two depending how much you go beyond just experimenting with k8s and using it to host most of your homelab services.