In the last post I showed how my cluster is maintained day to day and using various industry standard tools. This is a fairly robust process as I build out the cluster, but just having a cluster isn’t particularly useful. To build upon this we need to put up a means for outside network traffic to come into the cluster and become inside network traffic. This traffic might be HTTP destined for a service, TCP as a lower level application protocol, or some kind of UDP ingress to a specialized application.

In the last post, I talked about how I setup my Nomad cluster running on bare metal using Void Linux and Ansible. I talked a lot about the basic design, but not how I actually executed on that design. That’s what we’ll get into in this post. There are 3 main elements to configure in the cluster. First, the network, routing, and data-plane needs to be configured as a foundation that can be used to PXE boot the rest of the fleet, and then managed on an ongoing basis.

I recently decided to rebuild my development cluster so that I can finally nail down a few services that I’ve wanted to understand better for some years now. To get to the point that I have a working cluster meant setting out some requirements, rebuilding the entire software stack, and then finally making physical changes to the cluster itself. This will be a series of posts detailing the adventure to build a private cloud with actual cloud-like functionality, namely multi-chassis scale-out and the ability to treat arbitrarily many computers as one large resource pool.

Its late at night, you’re on a walk, and you see something wrong. What do you do, who do you call? I walk several miles daily to try to keep my health up given that my career has me sitting at a desk all day. As a result of this, I somewhat regularly encounter situations that need to get called in, and over the years, I’ve become increasingly weary to try.

People who’ve known me for a while know that I’m an alumni of a program called BEST Robotics, and that I still actively volunteer with the program. This is an experiential learning program with a strong focus on industry exposure. Students compete in many different aspects in Spring and Fall, but I am most heavily involved with the flagship BEST Robotics Competition in the fall. This 8 week program sees students presented with a challenge and turned loose for 8 weeks to come up with an answer.

One of the topics of discussion in my circle of friends we keep circling back to is security of personal, work, corporate, and production systems. Its no surprise since many of my circle have backgrounds in highly academic fields and have between us bordering on a century of experience in operating systems at all points on the scale spectrum. From active red-team work to defense of mission critical systems, our discussions tend to span it all, and usually with a fairly blunt and honest look at reality, something I don’t often see in the industry.

My home network is complicated. A good friend recently described the network as having the complexity of a small to medium enterprise. I’m inclined to agree, since my network participates in multiple internal and external Border Gateway Protocol (BGP) arrangements, has close to a dozen internal security zones each with unique set of firewall rules and access control rules. The network includes split horizon DNS, both dynamic and static DHCP configurations, and a number of servers and distributed switches.

TL;DR - If you’re using USB-C displays with Intel 12th or 13th gen hardeware, its probably a DRI firmware issue in the i915 kernel module. Roll back your DMC blob to adlp_dmc_ver2_16.bin. I’ve recently been building a machine for doing video production work at the robotics competitions I support as well as other small events where a video mux is required. In my pursuit of ever smaller footprints to load in and load out, I switched late this year to using a pair of portable USB-C monitors that accept displayport over USB-C.

I was recently asked at my day-job to think about what the global fleet would look like if I had the powers of Star Trek’s Q and could snap my fingers with a flourish and re-form prod into what I think would be better. Without the restrictions of design documents, dozens of project management strategies and more meetings than I can shake a stick at, I’ve put a lot of idle time into thinking about this.

Recent changes in high profile open source projects has got people talking about sustainability in companies that build open source again. Usually I don’t really care about this as it comes across as just poor business management, but Hashicorp taking the entire core portfolio out of the Open Source ecosystem has lensed this through a new perspective for me. That perspective is when business people try to solve for market conditions, and seem to forget about established solutions.