eblume/hephaestus
1
0
Fork 0
generated from eblume/project-template
Code Issues Pull requests Projects Releases 10 Packages Wiki Activity Actions

ci: cap CARGO_BUILD_JOBS at 1 to stop the DinD engine OOM-killing
All checks were successful
Build / validate (pull_request) Successful in 5m23s
Details

Browse source 
The recent CI failures ("Cannot connect to the Docker daemon") are the
DinD build engine being OOM-killed mid-compile, not flakiness. Even 4
parallel rustc invocations spike memory too high on the runner; serialize
to jobs=1. Slower but survives. Temporary mitigation pending more host RAM.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This commit is contained in:
Erich Blume 2026-06-03 22:46:09 -07:00
commit d93f84958c
1 changed files with 4 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

7
.dagger/src/hephaestus_ci/main.py
View file

@ -15,9 +15,10 @@ class HephaestusCi:
dag.container()
.from_("rust:1-bookworm")
.with_exec(["rustup", "component", "add", "clippy", "rustfmt"])
# Cap parallel rustc — unbounded (= ncpu) spikes memory on heavy
# crates and OOMs the build engine on a many-core runner.
.with_env_variable("CARGO_BUILD_JOBS", "4")
# Serialize compilation (jobs=1) — even 4 parallel rustc invocations
# spike memory on heavy crates and OOM-kill the DinD build engine.
# Slower, but the runner survives; revisit if the host gets more RAM.
.with_env_variable("CARGO_BUILD_JOBS", "1")
.with_mounted_cache(
"/usr/local/cargo/registry",
dag.cache_volume("heph-cargo-registry"),