| .. |
|
cluster_state
|
65c1cfcdc8
HA add-on + recipe repos (specs/tech-architecture/ha-addon-recipe-repos.md)
|
1 week ago |
|
config
|
8f78bcdc45
serve --cold: recipe-declared cold_args overlay for first-launch safety
|
6 days ago |
|
docker
|
bde9a8c427
follow_with_liveness: don't mislabel a slow boot as 'engine down'
|
6 days ago |
|
native
|
89013fbb75
native/dgpp: fix provisioning from the first live DGPP w2 run
|
1 week ago |
|
orchestrate
|
65c1cfcdc8
HA add-on + recipe repos (specs/tech-architecture/ha-addon-recipe-repos.md)
|
1 week ago |
|
power
|
c937008b42
power off: never cut a shared extra (fabric switch/fan) another cluster's live run needs
|
6 days ago |
|
report
|
503fe88e78
web/report: recipe logs open a dashboard WINDOW, not an inline stream
|
1 week ago |
|
staging
|
6ae3c5159c
readiness: watch() must not call an API that never opened 'dead'
|
3 weeks ago |
|
tmux
|
8f78bcdc45
serve --cold: recipe-declared cold_args overlay for first-launch safety
|
6 days ago |
|
web
|
8f78bcdc45
serve --cold: recipe-declared cold_args overlay for first-launch safety
|
6 days ago |
|
after_stop.rs
|
8f78bcdc45
serve --cold: recipe-declared cold_args overlay for first-launch safety
|
6 days ago |
|
ansi.rs
|
9bf6e76e6c
adopt r35 (latest lab rev): full 3-fork pin profile + r35 4x-dcp4 recipe; image built on dgx5 (71m, checksum 8bb87348), boot gates green
|
1 month ago |
|
before_serve.rs
|
8f78bcdc45
serve --cold: recipe-declared cold_args overlay for first-launch safety
|
6 days ago |
|
bench.rs
|
0e826d4b35
bench: add --no-think and --timeout passthrough to tool-eval-bench — unblocks valid quality benches for thinking-by-default models (Ling, GLM-4.5-style) whose default reasoning blows the 120s/turn timeout and excludes most scenarios (the 10-16 fp8 Ling run scored 67/100 at 3.1% completion for exactly this reason)
|
6 days ago |
|
bench_artifacts.rs
|
24d6737bc3
de-pin model revisions: kindling 2x/4x model -> floating org/model id (engine resolves via NAS refs/main -> latest snapshot, same as every other recipe; the pinned da920bb0 stays as an informational comment); 6x notes its tp6pad source rev. Bench manifest now records the model's resolved refs/main snapshot so a floating model stays reproducible (backward-compatible #[serde(default)]).
|
3 weeks ago |
|
bench_depth.rs
|
d5a2032562
bench: warm-prefix depth measurement (--benchy-args=--enable-prefix-caching) — the c2 long-context 'collapse' (13.7/5.6/2.1) was llama-benchy's default cold-monolith branch, proven artifact (flag flips c2@65K 13.7 -> 52.1, TTFT 28.3s -> 1.5s); drop DEBUG from sparkring recipe (diagnosis closed)
|
3 weeks ago |
|
bench_io.rs
|
b70649a547
feat(bench): --ssh-host — run tool-eval-bench on a remote host over SSH
|
2 weeks ago |
|
bench_remote.rs
|
425da04ac3
bench: --label optional (default 'perf' for perf-only), depth suite now opt-in (--depth), parse perf-box caveat asterisks, quiet json warning for perf-only
|
1 week ago |
|
bench_report.rs
|
45462379ae
bench: re-attach the perf table the --json-file path drops; fix aiden recipe toml
|
3 weeks ago |
|
builtin_scripts.rs
|
2dcf9a3e9e
exl3: temporary 2 GB display-reserve unlock (unlock/restore builtins) + per-recipe device passthrough
|
3 weeks ago |
|
cli.rs
|
8f78bcdc45
serve --cold: recipe-declared cold_args overlay for first-launch safety
|
6 days ago |
|
cli_sub.rs
|
5359f6f6b7
repos sync --watch: background re-sync loop (default 10 min) so the Mac recipe-repo cache never goes stale — mirrors the add-on entrypoint loop
|
1 week ago |
|
cli_tests.rs
|
9c39603968
feat: [registry] — push built images to a registry (ghcr.io) after build; docker push via tmux on the node, localized (6 locales), --no-push opt-out
|
1 month ago |
|
cluster_state.rs
|
e00a59ec49
Log probe SSH failures (previously swallowed as Err(_))
|
1 week ago |
|
cold.rs
|
8f78bcdc45
serve --cold: recipe-declared cold_args overlay for first-launch safety
|
6 days ago |
|
config.rs
|
e87b6706bf
THE leak: recipe watcher watched cwd ('.') — in the add-on container cwd=/ and notify followed /sys symlink loops, walking the container filesystem forever (~4.6 GB/h churn → allocator ratchet → daily Pi OOM). Watch only real recipe workspaces (watchable_recipe_dirs). Also: one shared HA client (a reqwest+rustls cert-store build per probe tick was another ~2 GB/h churn), shutdown now closes SSE connections on signal
|
1 week ago |
|
detect.rs
|
65c1cfcdc8
HA add-on + recipe repos (specs/tech-architecture/ha-addon-recipe-repos.md)
|
1 week ago |
|
doctor.rs
|
8f78bcdc45
serve --cold: recipe-declared cold_args overlay for first-launch safety
|
6 days ago |
|
ends.rs
|
69ba71ad59
feat: health probes [mikrotik] devices via RouterOS API login (ping = on, dead/rejected = off, blocking); power-on routes through on_entity relay
|
1 month ago |
|
engine.rs
|
8f78bcdc45
serve --cold: recipe-declared cold_args overlay for first-launch safety
|
6 days ago |
|
exec.rs
|
65c1cfcdc8
HA add-on + recipe repos (specs/tech-architecture/ha-addon-recipe-repos.md)
|
1 week ago |
|
gconfig.rs
|
ee2855da38
config errors name every searched path
|
1 month ago |
|
handlers.rs
|
3acc8fa855
power on: ensure-powered — nodes believed down with an already-on plug get a 30 s plug power cycle (2-5 s cuts left the PSU rail energized and the machine in soft-off); per-node outcome report becomes the job message; CLI uses a 1.5 s TCP reachability probe, web uses the last snapshot
|
2 weeks ago |
|
health.rs
|
0181645b9e
fix: make IB config optional for single-node clusters
|
2 months ago |
|
home_assistant.rs
|
e87b6706bf
THE leak: recipe watcher watched cwd ('.') — in the add-on container cwd=/ and notify followed /sys symlink loops, walking the container filesystem forever (~4.6 GB/h churn → allocator ratchet → daily Pi OOM). Watch only real recipe workspaces (watchable_recipe_dirs). Also: one shared HA client (a reqwest+rustls cert-store build per probe tick was another ~2 GB/h churn), shutdown now closes SSE connections on signal
|
1 week ago |
|
host_keys.rs
|
1ba1d99e05
fix: resolve review findings across modules
|
2 months ago |
|
locale.rs
|
519a63ca71
i18n: translate the recently added modules + drop dead FTL entries
|
1 week ago |
|
main.rs
|
8f78bcdc45
serve --cold: recipe-declared cold_args overlay for first-launch safety
|
6 days ago |
|
mikrotik.rs
|
69ba71ad59
feat: health probes [mikrotik] devices via RouterOS API login (ping = on, dead/rejected = off, blocking); power-on routes through on_entity relay
|
1 month ago |
|
power.rs
|
c937008b42
power off: never cut a shared extra (fabric switch/fan) another cluster's live run needs
|
6 days ago |
|
pre_serve.rs
|
afa6999213
feat(tmux): support inline venv_setup + venv_setup_script
|
2 months ago |
|
promote.rs
|
519a63ca71
i18n: translate the recently added modules + drop dead FTL entries
|
1 week ago |
|
promote_tests.rs
|
65c1cfcdc8
HA add-on + recipe repos (specs/tech-architecture/ha-addon-recipe-repos.md)
|
1 week ago |
|
publish.rs
|
8e20fd1dad
feat: versioned registry tags (build checksum) + latest; publish --push; duplicate guard via publication ledger with --force
|
1 month ago |
|
publish_render.rs
|
95e45ae40d
repo renamed on Gitea: sm121/sm121-launcher → sm121/launcher
|
1 week ago |
|
readiness.rs
|
a82e875c2c
karmic-kraken: image build + GLM53 B12X recipe + fail-fast readiness
|
1 week ago |
|
recipe_repos.rs
|
5359f6f6b7
repos sync --watch: background re-sync loop (default 10 min) so the Mac recipe-repo cache never goes stale — mirrors the add-on entrypoint loop
|
1 week ago |
|
recipe_repos_tests.rs
|
65c1cfcdc8
HA add-on + recipe repos (specs/tech-architecture/ha-addon-recipe-repos.md)
|
1 week ago |
|
registry.rs
|
51912523c5
feat(registry): publish --registry upload + sm121 registry list/show/install
|
1 month ago |
|
runlock.rs
|
2f717631ff
report: stale cross-check of run ledger vs live probe (orange + runlock hint); new 'sm121 runlock list|close' offline ledger command
|
1 month ago |
|
ssh.rs
|
5f1783f3e1
memory: mimalloc allocator feature (add-on builds), lazy report refresh (file-event-driven, fleet cross-check rides the dashboard bus), SSH key cache, graceful web shutdown (SIGTERM) + dhat profiler feature — the Pi OOM-kills (2.7 GiB RSS) traced to musl's never-trimming allocator ratcheting on per-tick transient spikes
|
1 week ago |
|
ssh_tests.rs
|
5a6ec550d7
fix(ssh): pass through already single-quoted tokens in shq
|
2 months ago |
|
staging.rs
|
8f78bcdc45
serve --cold: recipe-declared cold_args overlay for first-launch safety
|
6 days ago |
|
stamp.rs
|
1d73fd0945
serve/stop reliability + run ledger + archive curation
|
1 month ago |
|
ui_test_lock.rs
|
f4aee53e0c
single working set: recipes/ + docker-images/ move to the shared repo checkout (../recipes) — dev config points at it via recipe_sources/docker_image_sources; validate gates over every configured source; web hosts the recipe-repo resync loop (--repo-sync-interval, default 10, 0 disables); playwright UI suites run serially (browser lock + split test legs)
|
1 week ago |
|
validate.rs
|
5194805e77
native/dgpp: emit paths.resident_cache in cluster JSON; exempt Native deploy from multi-node-tp1
|
1 week ago |
Alexander Korolev
c937008b42
power off: never cut a shared extra (fabric switch/fan) another cluster's live run needs