Month: July 2022

Migrating from Docker to Podman

I’ve been eyeing Podman for a while. Being able to build and run containers without the Docker daemon hanging around in the background as an omnipotent demigod sounds nice. Less stuff running = good. Installing it is trivial on Ubuntu (unlike installing Docker): apt-get install podman and it’s done. As the documentation states, it can even be aliased to docker to make migration even smoother. It has tons of features that I don’t think are there in Docker: generating Kubernetes YAMLs from running containers, using container LABELs to make run commands simpler and so on.

Which is all nice and it’s to stay for sure, but I had to realize that at this point this won’t have much of an effect on me (anymore). I hardly ever build container images locally, instead using some CI (Github Actions) to do it automatically. Not having to remember all the build options (even if just in .bash_history) is nice, and if I have to put it into code anyway, I might as well automate it completely.

As for running containers, podman run will be now my go-to for experimentation (instead of docker run). However, most of the time I use docker-compose to orchestrate a local development environment. It’s possible to use Podman’s system service to use docker-compose, but then is there much of a difference from running that and having the Docker daemon running? It’s also possible to use (simple) Kubernetes Pod/Service YAMLs to achieve something similar, and I’ll be sure to consider that the next time. But I don’t think I’ll make the effort to change these stuff for existing projects on my current computer. Not unless Docker kills free users altogether…

Sweet base camp

After the unsuccessful summit push, it was time to return to camp 3. Needless to say without the darkness, distances became much smaller. At the same time, people (especially those of us climbing without oxygen) were getting very tired. For some of my teammates, climbing down the rocky “chimney” back on the snowy slopes took maybe even longer than climbing up did. While I’d have preferred to use a figure-8 to safely descend at that point, the ropes were such a tangled mess that it wasn’t an option. So I just rolled some of the ropes around my arm and climbed down as quickly as the altitude allowed. Personally I found that the easy part.

The long, snowy descent back to camp 3 was really hard for me. It was the soft, unstable snow. Basically every step meant sinking or sliding with the snow, making it a very frustrating challenge to go down in (needless to say it was an extremely exhausting challenge to climb too). Even with my trekking pole set up for snow, it was hard to take a step without losing my balance. This quickly drained the last of my energy reserves, so I was getting slower and slower, and I reached camp last of the team. The first thing I did (after reporting in) was collapsing in a tent to sleep for three hours or so (I think I didn’t even turn off my GPS tracker, making some people worry—sorry!).

To the summit of Broad Peak

Multiple people have commented on the trail how tight our summit push is. We didn’t have a third rotation, nor did we get a chance to sleep in camp 3. This is because we were there at the mountain early in the season, but had to sit on our hands for almost ten days due to snow (and then some more snow) and logistics issues. At the same time those of us who got up to camp 3 were fit and strong, ready and eager to go for the summit. Having some more leeway would’ve been nice, but with bad weather incoming, we had limited options. Being a commercial expedition means that we can’t just stay for three more weeks waiting for the next window.

Up, up, up!

After the second rotation on Broad Peak, I hoped we’d get a day rest. Resting a day in base camp after sleeping in the high camps helps a lot with acclimatization. Not to mention on the way up a rock rolled under me and I hurt my left knee. Luckily it got better after the rest in camp 1, but it’s still not good. Letting it recover would’ve been nice.

Except the weather forecast had other plans. There wasn’t enough time anymore to waste: snow was coming, and probably to stay too for weeks. Which meant we only had this one chance to move for the summit. Those who wanted a rest in camp 1 would start the next day, while those who were okay with the long way straight to camp 2 would get a day.

Second rotation

After taking a day of rest after the first rotation in base camp, we were heading back for the second rotation on Broad Peak. This time the goal was to sleep in camp 2, and preferably hike above camp 2, maybe even touch camp 3.

We started early in the morning so the glacier rivers would still be frozen and the snow firm. Once the rivers on the glacier thaw, traversing the moraine becomes much more difficult. The rivers’ flow is too strong and the terrain too slippery to cross just anywhere, and finding a safe crossing isn’t easy even with experience. (Hell, a local porter died trying to jump one of the rivers…) Higher up on the mountain having firm snow means much easier progress. Not sliding back a step for every two steps forward is a huge difference in exhaustion.