kevinveenbirkenbachandClaude Opus 5 980a47e916 feat(swarm): dump ganesha thread states on rescue ...

The artifact of the wedged run carried journal, unit status and mounts but no process data, so the hang could be localized to the export config commit and not pinned to a syscall. wchan and syscall per thread close that gap.

MainPID 0 is its own likely state given Restart=on-failure with the start-rate limit disabled, so that branch reports the unit counters instead. Every read keeps 2>&1 because /proc stack and syscall are EPERM on hardened kernels.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

2026-07-29 11:55:26 +02:00

..

clean

fix(swarm): release the host state mount between drill cycles

2026-07-29 02:05:16 +02:00

collect

feat(swarm): dump ganesha thread states on rescue

2026-07-29 11:55:26 +02:00

topology

fix(swarm): carry the drill handover in-process, not via $GITHUB_ENV

2026-07-28 23:53:43 +02:00

unmount

fix(swarm): release the host state mount between drill cycles

2026-07-29 02:05:16 +02:00

_context.sh

fix(swarm): carry the drill handover in-process, not via $GITHUB_ENV

2026-07-28 23:53:43 +02:00

diagnostic.sh

…

trigger_units.sh

…