< up >
2024-08-23

Home office via hot spot

tl;dr Working from home via hot spot isn’t as unreliable and traffic expensive as I thought. Pre-schedule suitable work and delay data intensive ones like meetings help along the way.

I recently moved into a new flat and the technician appointment to enable my internet connection was still in the future. At the same time I awaited some furniture deliveries and you know how they’re scheduled in Germany¹: expected delivery time between 7 and 13 o’clock.

To save my vacation for actual holiday, I got the idea of working from home and connecting my notebook with the hot spot of my mobile.

Limits

My mobile data plan is limited to 10GB where 5GB are already spent. Clearly if I pull some containers or install some updates, the remaining traffic won’t last long. Even building go binaries may come with a surprise as missing dependencies get downloaded automatically.

So this is a matter of preparing the right work and estimate the data needs properly. Audio- and video chat is hard to estimate for me, it’s intuitively data hungry but compression algorithms got better in the past years.

Preparation

There were some TODO’s where I could work on locally and watch ci pipelines via some neat git-platform-integration tools (e.g. github, gitlab,…). Meetings were only scheduled in the afternoon with the hope that the delivery already arrived and I had the time to drive to the office.

Home office

I started at 6 o’clock and monitored my consumption using the android lock screen widget for hotspots (see featured image above). I worked on some pipelines, monitored its jobs using proper integration which worked pretty well without consuming that much data. Even opening the git-platform e.g. view Pull Request worked well. Some web-based tools don’t need much, other consumed 10 megs on loading. Such learnings are helpful to categorize workloads into data-friendly and data-hungry ones.

When working on servers via ssh, I tried to use pagers like less² to control the amount of output sent to my screen, as much as possible.

The bell rang at 8:30 with the delivery of my bed, so I worked til 9 and drove to the office. I consumed about 77.34MB in those three hours of work. I had expected at least ten times more.

Summary

Those three hours of working via mobile data went very well. Git and git-platform-integrations work well without needing much data. Some web-based tools are hungry which need to be kept in mind.

Definitely something I want to do again to gain more experience.