I got this for cheap in the UK, gen5 blades. I am reluctant to even plug the thing in! Apparently it works though. Heard its a huge energy drain. Worth the nominal fee it took to acquire it as a homelab in a separate room? Part out (one blade i checked had 16gb ram and a drive) or sell as whole system? Thanks.

Restrictions due to rental apartment. NAS is placed in the living room corner. What do you guys think? Debating of putting a slim nocturnal fan in there (there are cooling vents in the door).

The proxmox box n300 idels around 48C (Summer); with the door open it is 40. load is ca. 55 degrees.

I've designed the case for myself, to make a low power consumption server at home, as the electricity is not the cheap where I live, but if people are interested, I can make more of them (only in Europe).

The case is made out of galvanized steel and powder coated in black. You can fit inside: - Two mini-ITX motherboards (I have in mine i7 12700T 35W TDP and i7-1165G7 with TDP 28W) - Two SFX Power supplies - Four 80mm Fans - 4x SSD / 3x SSD + 1 HDD / 2 x HDD + 1 SSD can be installed

I improved the design a bit for the next case, but looks more or less the same.

A picture of our core router

Hello Home Labbers,

When I started homelab, a couple of years back, one of the toughest thing I found was the difficulty around self-hosting and getting my servers internet facing servers, as most of the residential ISP, dont offer static IPs OR they are priced high.

Since I moved my servers to a datacenter, I was able to use BGP to get my own prefixes online. As a result, I am here to offer everyone within this subreddit, free IP BGP transit over a tunnel, via AS21738 or AS25759.

If you have an ASN and want your prefixes up and running for your homelab, I will be able to downstream you for free. If you are interested, feel free signup here. Any question, feel free to comment below.

Disclosure: We use the signup form to verify you are an authorized to use your ASN

It’s interesting how it works out for me - my pet projects turn out by chance. There is no final goal, there is only an impulse: “Oh! This sounds interesting, how can this be done?” And all: “sleep is for weaklings”, “beer on Friday? Of course I won’t!” and stuff like that. As they say, there is only a way. And this story began in much the same way... It was getting dark. At work I had nothing to do, I needed to install a certain number of servers and services for monitoring, but due to the large bureaucracy in the company, this was not easy to do, and the monitoring system itself worked on SNMP database, but where can I get SNMP from a self-written service? And then the brilliant idea came to my mind to try it myself. Besides, it didn’t look complicated: monitoring ports, http and sending an alert somewhere. “Why not,” I thought, besides, I’m learning more about Python. And so he appeared...

Simple monitoring that somehow does something, shows something, and even has a console tool:

Old SMON

A couple of years later, I remembered that I had homemade monitoring and why not add it to my main pet project, Roxy-WI. No sooner said than done. After all, the more functions the better! And it so happened that over time, monitoring became “crowded” within the walls of Roxy-WI: on the one hand, it was necessary to develop a web interface, on the other, monitoring, so that there was no preponderance in one of the parties, I decided to move monitoring into a separate project. Greetings - RMON! Yes... my names are so-so.

RMON status page

Pfft... one more monitoring, how many?

100500? Yes, perhaps so, they probably also said about Prometheus at one time: “Why is there Zabbix?!”, and before that about Zabbix: “Why is there SNMP, MRTG and Nagios?!” Yes, there is, but why not? Maybe you'll be able to do something better. Of course, I don’t yet put RMON in the same category as these monitoring systems, not yet. What if we can do something better ;)?

What do I see as the “competitive advantage” of RMON over existing monitoring systems, primarily over Prometheus (as an industry standard) and Uptime Kuma (as closer in functionality)? There are, in my opinion, at least five main killer features:

1. Agents - you can install several pieces both inside and outside the perimeter and monitor availability from several points. Agents can be combined into “regions” to balance checks and move between groups.
2. API.
3. Role-based agent access model.
4. Easy to install and configure, Web interface and Status pages.
5. 7 HTTP connection metrics + SSL certificate attenuation monitoring.

There is also Ping monitoring, DNS records and TCP. In the future I plan to expand the capabilities of inspections.

New RMON

We've seen it all before

Yes, agents are essentially implemented in Prometheus and Blackbox exporter: Blackbox exporters can also be installed at different points and monitored from there, + - the same thing. Yes, Uptime Kuma is even easier to climb and also has a web interface. The API can be replaced with the same Ansible, for example. But there is one thing - it is not here and there. You can’t give a playbook to a person and say: “Don’t create anything on those exporters, you’re bad!”, you will have to raise several instances to share access, plus he needs to be trained to work with Ansible. It is also impossible to automate work with checks. More precisely, most likely it is possible, but these are crutches and a high level of entry.

As a result, for those who will write: “The Web sucks, the console is our everything!”

Yes, sometimes it is, and sometimes it is not. Sometimes even the most advanced and technologically correct solutions are not suitable. Somewhere it’s a pity to waste time and resources, somewhere you don’t want to dive, and somewhere you need to “get everything done in 2 minutes.” And sometimes advanced solutions are simply not needed and it is more convenient to work with simpler ones. We must proceed from a specific situation, and not force everyone into a framework: “%UserName%, use only %ProgrameName% in all cases of life!”

P.S. If you want to try, then write, I’ll be happy to show/explain :).

Hi fellow homelabbers,

I need your guidance. Recently did a lot back and forth on the German variant to Craigslist and ended up with two RX2530 M2s and one RX2540M1.

Found out that one Raidcontroller wasn’t working anymore from the M2s so put everything into one and migrated my vms onto it. Right now the proxmox install is on a single Samsung 850 SSD and inside are 2 old SSDs as one virtual device and 2 new ones ( Crucial 500GB) as the second virtual device. I’m running homeassistant, the arrs, gitlab, Mattermost, jellyfin etc.

From the 2540 there are 9x 800GB SAS Disk available. It came as RAID6 preconfigured and the speed was massive. At least that feels like it. However the noise of the 2540 keeps me away from wanting it running 24/7. you can easily hear it through the closed door and I’m getting a bit crazy about that. Sad part here: I have an ARC310 with passthrough working with jellyfin but I will not make use of it due to the noise.

There is my old TX120 S3 which also has 2x Crucial 500GB SSD and 2x 300 GB SAS Drives I could use.

The single RX2530 M2 has two v4 xeons with 10C/20T each. So it will satisfy whatever I need to run for eternity and together with parts from the 2540 it will have 192GB of Memory.

Which kind of storage layout would you shoot for in the M2. Given there are 10 slots available and my media and data is on the synology NAS?

Cheers

I started working on my mini homelab.

• 3 × Trigkey Green G4 Mini PC
  • CPU: Intel® Processor N100 @ 3.40GHz
  • RAM: 16GB
  • SSD: 512GB
• Netgear GS305v3 switch:
  • Ports: 4
  • Speed: 1000Mbps

https://github.com/ivanlemeshev/homelab

https://preview.redd.it/ncouid8mevad1.jpg?width=3072&format=pjpg&auto=webp&s=4be767d2c046f13bb945458cb651ea69d9fc7118

This is my sort of mobile server. It's a pi 5 with a hyperpixel 4. I am not using a router since I don't need too, this just provides some small services and connects directly to my hotspot/hotel wifi. I runtipi, code server, Adguard (fallback Adguard if my main server dies or is powered off), uptime kuma, portainer and olivetin. Of course everything happens through tailscale. Although I would love to add a small router and SSD in the future.

https://preview.redd.it/ypqebncecuad1.jpg?width=3000&format=pjpg&auto=webp&s=1690e148c05c8390d9db4e3db428931bffdb8927

https://preview.redd.it/onv8vmcecuad1.jpg?width=3000&format=pjpg&auto=webp&s=02f1e0fa5eff03a053e39979f5f9805081aca1f9

Yeah, this lab has grown from one R720 to 2x R720's, an HP EliteDesk 800 G6 and an OptiPlex 7080.

So the reason it grew is I finally dumped my Dedicated Server, got the R720 and the HP and Dell, upgraded to 1Gbit Business Internet, and moved all my production services to the house. Saves me some money!

Running Gitlab, some Webserver stuff, Build Servers, and TrueNAS on a VM with my HBA passthroughed for my NAS for the Prod stuff.

Plex is on the Dell for HW Transcoding.

Network wise I'm running all Cisco except for the firewall (that's a Zyxel Flex 200) and the Wireless AP (Ubiquity). Got it all VLANed and a DMZ network (that's what the little Netgear switch is for, keeping the DMZ network separate), The Synology is for some legacy stuff. The VG204 is for my house landline plugs for playing with old phone system stuff.

Only bad thing so far is Cali weather and the heat. Sounds like some 747's lifting off in there, least they ain't cooking to death!