jcoffey-dev is traveling from Thursday 1 October through Sunday 4 October. Issues and pull requests are welcome, and will get an answer after that. Thanks for your patience.
The v2026.9.30 tag build (GitHub run 36764295827) lost its arm64 publish job at the final link of inbuxa: rustc was SIGKILLed after 38 min with ResourceExhausted … cannot allocate memory. The release profile is fat LTO with one codegen unit, and the 16 GB ubuntu-24.04-arm runner ran out. amd64 got through on the same size of runner. Because of that, index, ghcr, release, binaries, github-release and the announcement were all skipped. Production is unaffected: the nodes run the amd64 image, which was published.
Each publish job now adds a 16 GB swap file before the build. buildx's container has no memory limit of its own, so the linker can use the host's swap.
The step logs free -g and df -h /, so the next tag build shows the headroom.
This can't be tested on a PR, because publish only runs on tags. It first runs on the next tag (2026.9.30.1, planned right after this lands). A re-run of the v2026.9.30 run would still use that tag's old workflow.
No new strings.
The v2026.9.30 tag build ([GitHub run 36764295827](https://github.com/inbuxa/inbuxa-server/actions/runs/36764295827)) lost its arm64 publish job at the final link of `inbuxa`: rustc was SIGKILLed after 38 min with `ResourceExhausted … cannot allocate memory`. The release profile is fat LTO with one codegen unit, and the 16 GB `ubuntu-24.04-arm` runner ran out. amd64 got through on the same size of runner. Because of that, `index`, `ghcr`, `release`, `binaries`, `github-release` and the announcement were all skipped. Production is unaffected: the nodes run the amd64 image, which was published.
- Each `publish` job now adds a 16 GB swap file before the build. buildx's container has no memory limit of its own, so the linker can use the host's swap.
- The step logs `free -g` and `df -h /`, so the next tag build shows the headroom.
This can't be tested on a PR, because publish only runs on tags. It first runs on the next tag (2026.9.30.1, planned right after this lands). A re-run of the v2026.9.30 run would still use that tag's old workflow.
No new strings.
The v2026.9.30 tag build's arm64 publish job was killed linking the
inbuxa binary (fat LTO, one codegen unit): cannot allocate memory on the
16 GB ubuntu-24.04-arm runner. index, ghcr, release, binaries and
announce were skipped. amd64 got through on the same size of runner.
Each publish job now adds a 16 GB swap file before the build; buildx's
container has no memory limit of its own, so the linker can use it.
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
The v2026.9.30 tag build (GitHub run 36764295827) lost its arm64 publish job at the final link of
inbuxa: rustc was SIGKILLed after 38 min withResourceExhausted … cannot allocate memory. The release profile is fat LTO with one codegen unit, and the 16 GBubuntu-24.04-armrunner ran out. amd64 got through on the same size of runner. Because of that,index,ghcr,release,binaries,github-releaseand the announcement were all skipped. Production is unaffected: the nodes run the amd64 image, which was published.publishjob now adds a 16 GB swap file before the build. buildx's container has no memory limit of its own, so the linker can use the host's swap.free -ganddf -h /, so the next tag build shows the headroom.This can't be tested on a PR, because publish only runs on tags. It first runs on the next tag (2026.9.30.1, planned right after this lands). A re-run of the v2026.9.30 run would still use that tag's old workflow.
No new strings.