I had a good year. I wouldn’t be writing this if I had a bad year. You shouldn’t judge your year against anyone who writes up a nice little year in review like this. Here are some reflections.

On being 30

I wasn’t happy to turn 30, but I am happy to be 30. I have enough time out of the way to feel comfortable for the first time since I was young. I still learn new things, but I can start to see more people who have always been in my life have either recently stopped that or stopped that some time ago. Some people haven’t learned anything new in decades. Others might have stopped and then started again.

I’m 30 and I have the confidence of someone who is comfortable with themself – so I can share something about the world. I might even be right! What if I had stopped learning new things though? There were some people in my life who had wisdom; they stopped. They should be sharing wisdom about the world with me. I see it now.

On work

I don’t think I’ve ever had such a consequential year at work before. I worked on three projects that have taken years to create. I don’t know if I’ve communicated this well at work, but now as a I sit reflect, I’m impressed. In my first couple of years on the job (2017 & early-2018) I looked around me and made some suggestions – here are the relevant three:

I wish I could share the “Goals & Accomplishments” notes to prove this, but I have been pushing for these things for five years.

Logging cluster

One of my first big undertakings at work was rehabilitating an old Graylog server. It was centralizing logs for the networking team. It was clearly overwhelmed and needed some new nourishment. I poked around on this issue for sometime – avoiding something I suspect I knew pretty soon. I acquired some out-of-warranty hardware and in the summer of 2017 built my first Elasticsearch cluster. This pet cluster was fine, but clearly not going to win any shows. I felt the scrutiny:

I was patient and continued to expand where I could. Collecting more logs and improving the reliability. My coworkers started to see the rationale. I didn’t have a lot of time to dedicate to the project, but over the years progress was made. In January of 2022, I finally was able to build a production grade OpenSearch cluster. It took five years to complete this project. Nearly immediately after finishing the set up and migration, a new coworker took on the task of bringing in all of the missing logs. He’s added monitors and we are addressing things I could have only dreamed of five years ago when I first drew a diagram of what a “log cluster” could be. Coming up on one year in, I’m working on passing the management to the new coworker. I’ve done the hard part of being patient and learning the tools.

I asked for a promotion after completing this project.

High performance compute

Sometime ago I inherited a compute cluster. It would have been high performance when it was bought, but alas it had become just compute. It was woefully out of date and infeasible to improve. I kept it running, but just at that. Some limited research was being done utilizing it, but many professors around campus were left to use their own specialized machines or rely on publicly accessible HPC resources.

A request came in searching for a solution to using a popular biology compute tool for a class project. We waffled on this, but explored the options anyway. Smelling doom on the horizon, the original cluster took the opportunity to die in 2020. This confluence of events forced me to implement a temporary setup. It was two old vSphere hosts, some scrap storage and yet – more power than the previous cluster. A couple of hooks and the biology tool could handle the classes simultaneous workload. I had to get cozy with Slurm on the fly, but after years of dabbling in HPC when called upon, I had just enough grasp to get it running.

In 2021, we purchased an actual cluster – and this summer it was installed. It’s small, but for the students who need to learn the basics – the size doesn’t matter. The new cluster has all the infrastructure the previous ones were missing. I took all of August and September this year and built modules and integrated tools. It’s surreal seeing this happen. Requests come in for things like Julia, R, GSL, and AlphaFold. I can finally meet those.

The phone system

In my interviews for my job (back in 2016), I insisted I didn’t want to work with phones. I had previously worked with Avaya, and hey, what do you know, my new job had (a MUCH larger) Avaya deployment. They ignore my protests and I’m too nice so there I was – working on phones. If you’ve never worked with black box enterprise systems like Avaya you might find some of this extreme, but it’s not.

The 24/7 support isn’t typically the safety net you’re looking for. Phone systems are typically extremely reliable. We’ve had scenarios where unplugging a USB from a VIRTUAL server causes all the phones to reboot, but regardless of the stupidity of that behavior, the phones come online again mostly without any hiccup. The issues you face are much more…edge-case. It’s finicky to work through this with support and immensely time consuming. This system is built on jargon, clever 1970s engineering, and coked up managers from the 1980s (the last time Avaya innovated). After chasing some of this I was looking for something to make the process of manually managing thousands of users and extensions more tolerable. I was a fallback on this stuff – so things sat on the back burner.

In 2019, we spent far too much money to upgrade some stuff and I began to revisit my previous suggestions that we make a five year plan to get off this system. I really thought about it. We couldn’t afford to own this anymore. We couldn’t afford a new hire (or even find one!), we couldn’t afford the contracts, we couldn’t afford to pay specialized services to preform the necessary updates. We don’t have enough staff in the department to have someone doing manual labor that Avaya requires. It was time to go. I made my pitch again.

2020 reared it’s ugly head. It was time to go. A year of reviews and we selected the replacement. I helped, but only that. It was a team and a very hardworking coworker that did the physical labor of swapping phones and met with every group to build their programming. We did a nearly daily move of around 20 people for months. That’s the tip of the iceberg though. If we wanted this phone system to pay for itself, my job starts now that most nearly all the phones have moved to the new system. It’s time to tether the floating satellite back to the mother ship. Every little thing that can interconnect will. Users will automatically be provisioned, extensions will be dynamically added. We will watch logs, we will monitor devices. I spent five years asking for more work to do. What a fool!

Other things

On LASIK

I’ve been a glasses-wearer for half my life, but during Covid enough was enough. I was tired of fumbling with contacts and my fogged up glasses while wearing a mask or ALL OF WINTER. I ponied up and got LASIK in March. I can’t recommend it enough. It’s a breath of fresh air. There’s a cult of LASIK and I’m in it.

On friends

My friends are changing. There was one marriage and one divorce in our friend group this year and we couldn’t be happier for both. Losing a friend to the alt-right and another who mistreated his wife (we kept her, now happily divorced). I didn’t expect that either of these events would affect me nearly as much as they did. I am filled with pride to be friends with people who know what they’re about.

If you know someone who is leaving their spouse after being manipulated for years, I hope you can see just how brave they are.

On grad school

I graduated college with a B.A. in History years ago. I was a social science education major for much of my college years, but switched to history towards the end. Focusing on history was one of my favorite times in college.

I’ve always wanted to get a computer science degree and after years of kicking it around I just did it. I enrolled in a distance grad program that will allow me to acquire first a graduate certificate in cyber security and then a full masters in it. I have to achieve a certificate first (it’s all part of the same program), as part of a probational process since I don’t have a CS undergrad. I tell people I’m converting my History B.A. to a CS masters degree.

School is fine, but it definitely hits different after being away for eight years. It’s common that people with CS degrees will suggest that most of your day-to-day is self taught. That’s likely true, but never fall into the trap of dismissing foundations.

On a new home

My wife and I lived in our apartment for seven years. We moved in shortly before getting married and started looking for houses in 2017. We got serious in 2019. We were picky and took the availability of our local market for granted. 2020/21 housing market changes alarmed us, but we stuck to our guns. We had a certain type of house in mind and we liked our apartment. Don’t rush a decision. We bought a house we intend to die in. It’s from 1929 and filled with original hardwoods and builtin unpainted details. There’s room for me to work from home too. We walk to downtown and the farmers market.

It was a good price and no bidding war occurred. Every star aligned and we painted the whole place, removed a ton of carpet, and moved in in October. We are happy and look forward to a lifetime of projects. This is the stuff that makes us happy: well thought out decisions and a task done well and right.

Conclusion

It was an unusually good year. I wonder, were my other years bad? Maybe I was coasting. I cannot sense these things. I don’t subscribe to some of the belief systems about accomplishing goals in certain time frames or anything about when life events should occur. My wife and I have always grown and changed together. I think we have become wiser and learned to be confident in ourselves. We are not only individually more competent, but together we are stronger than ever. Could I have had a good year without her help? No. Don’t get into the habit of dismissing the foundations.