I bought a new Beelink mini-PC that has an 8-core AMD CPU, 32GB memory, 1TB m.2 drive for $429. Added two 2TB SSD drives (software based RAID1). Attached a USB powered 120mm fan on top.
Running a bunch of virtual machines. Plenty of power, stays cool and takes up very little room.
I live in NJ, so I usually preconfigured my equipment and ship it. I've also sent them bare servers and had them connect a KVM for me to do the setup remotely. They have common replacement parts on hand, I've bought some memory sticks from them and had them swap.
You can check back, they update the list at server availability changes. Other providers there are Dedispec and Joesdatacenter, may have something in stock you're looking for.
I'm currently doing individual COLO (I don't have a rack atm). But I haven't added any space in three years. I've replaced some older equipment, but that's about it. I've been talking with another customer about splitting a rack ($200 each) as I'm spending spending $270 for 6 COLO servers ($240 for the space and $5x6 for additional IP space per-server).
joesdatacenter.com (Kansas City) has single server COLO for $50 a month.
Another good low cost provider to consider is WholeSaleInternet.net. Their datacenter is located in Kansas City. I've been colocating and renting dedicated servers through them for since 2005.
Servers start at $9 per-month. A comparable example:
I'm colocating a couple of servers there for $40 per-month each, bandwith is 1Gbit unmetered and comes with 5ip's. A couple 1U's and towers. I recently bought a used 1U server off Amazon for $400. It has 48 cores, 96 GB memory and 4x1TB drives and came with a one year warranty on the components.
Hetzner was solid, but their network was sketchy at times.
They make these for hvac systems, mines installed in the air return, fully enclosed it works well. When the heat or AC isn't in use, I just run the hvac system fan.
Based on my own experiences, one of the most disappointing elements of devops is their lack of understanding how things work. They run to GitHub and pull other peoples recipes to build out certain environments. Even when they build automation from scratch, they miss tons of core OS level tunables and rather than adjusting those, they'll just add more servers to the mix.
I worked at a place that had an elastic search cluster with 30 nodes, because they kept hitting open file limits, when 8 servers could easily handle the traffic with basic tuning.
This is great news as well. @ Wildbit we have a dedicated logging server consisting of Rsyslog, ES, LogStash and Kibana3. It's been improving considerably each month.
Running a bunch of virtual machines. Plenty of power, stays cool and takes up very little room.