The OCaml Planet RSS

On the base image builder, we have recently seen a significant increase in the number of times we hit the Docker rate limits on manifest pushes. Most of these work on the second attempt, but this week, a new class of permanent manifest push failures has emerged.

After deploying the QEMU RISC-V machines, the base image builder ran into issues on the first Docker build.

mtelvers/day10 can now export build results as multi-layer OCI images, where each opam package becomes its own layer.

Over the past few days, I migrated several OCaml CI services from Scaleway to Cambridge, consolidating them onto fewer machines with fewer services.

We are increasingly hitting the Docker Hub rate limits when pushing the Docker base images. This issue was previously identified in issue #267. However, this is now becoming critical as many more jobs are failing.

Jon asked me to make a Docker image that contains OxCaml ready to run without the need to build it from scratch.

The moves of registry.ci.dev, opam-repo-ci, and get.dune.build have followed the template of OCaml-CI. Notable differences have been that I have hosted get.dune.build in a VM, as the services required very little disk space or CPU/RAM. For opam-repo-ci, the rsync was pretty slow, so I tried running multiple instances using GNU parallel with marginal gains.

As noted on Thursday, the various OCaml services will need to be moved away from Equinix. Below are my notes on moving OCaml-CI.

For some time, we have had issues on Ubuntu Noble when extracting tar files within Docker containers. See ocaml/infrastructure#121. This is only an issue on exotic architectures like RISCV and PPC64LE.