admin
e70b5feebe
Completes the spike once the venue came back. Q4's hang case and teardown are now measurements, not plans. Against a dmsetup-suspended device (I/O queues instead of returning EIO): - P1 (devno compare) and P2 (ext4 abort flags) completed in 364us / 206us. They read /proc, so no block device is involved. - statfs and getdents completed and reported HEALTHY — on a wedged device they do not even hang. R-117b confirmed in a second failure mode. - EVERY probe that touches the device blocked, including a buffered write with no fsync: the O_CREAT metadata path needs journal access (wchan=do_get_write_access). There is no cheap-and-safe write probe. - The blocked process survived SIGTERM AND SIGKILL (stat=D, wchan=folio_wait_bit_common, still alive 3m50s after kill -9) and died only when the device was resumed. So `systemctl restart felhom-agent` would hang, leaving the agent unrecoverable until the device returns or the host reboots. The thread count does not reveal the leak (5->5, 5->6). Filed as R-117f. A timeout protects the caller's control flow and nothing else, so "the fix must issue no block I/O" is now a fence rather than a preference — the thread-leak hypothesis the probes were built to test turned out to be the weaker half of the result. Teardown done, all three layers: guest 9301 destroyed, r117scratch removed, both dm and both loop devices gone, scsi_debug unloaded, local back to 37.02% against a 37.00% session start. Fences re-verified AFTER teardown: 9201 running, drill-r50 stopped, local-lvm 38.84% byte-identical, felhom-backup content unchanged, live /mnt/felhom-drives intact with both submounts, agent active. Layer 3 genuinely empty — 9301 had no NIC and ran no controller. Trap recorded: a suspended dm device must be resumed BEFORE any umount, or the teardown blocks on the same uninterruptible sleep.
Felhom — Documentation
Felhom is a managed home-server service for Hungarian households, built on a three-component model over Proxmox:
- Hub — operator backend on k3s (
hub.felhom.eu). Repo:felhom.eu/hub/. - Host agent — one per Proxmox host; operator-tier; owns all Proxmox interaction. Repo:
felhom-agent/. - In-guest controller — one per customer LXC; Docker-only; manages the customer's apps. Repo:
felhom-controller/.
This directory is the central, code-verified documentation home for all three components plus the platform and the security-audit record.
Sections
Controller (in-guest) — controller/
The Docker-only app-domain controller. Full per-area docs grounded in current source (v0.59.0).
→ controller/README.md: module map, deploy & stack lifecycle, backup
architecture, storage/monitoring/metrics, auth/hub/sync/integrations.
Host agent & platform — architecture/, proxmox-platform.md
The operator-tier agent and the Proxmox platform.
architecture/01-topology-and-trust.md— topology & trust modelarchitecture/03-host-agent.md— the host agent (Go; v0.29.1)architecture/04-control-plane-authorization.md— signing, escrow, authzarchitecture/02-controller-module-map.md— historical v0.33 planning map; the live map iscontroller/module-map.mdproxmox-platform.md— Proxmox platform reference
Hub (operator backend) — architecture/05
architecture/05-hub-architecture.md— hub architecture (v0.11.0)
Security audits & remediation — audits/
audits/deep-sweep-2026-06-13.md— cross-repo deep audit (controller + agent) with remediation statusaudits/bughunt-reconcile-2026-06-13.md— reconciliation of the v0.30.3 BUGHUNT against current code + merged fix list
Spike & test findings — tests/
Per-slice spike/validation findings (phases 0–5, slices 7–10). See tests/.
Conventions
- Code-verified, not memory-derived. Architectural claims here are checked against the actual current source; if a claim can't be verified it is omitted and flagged, not guessed.
- Per-repo operational working files (
CLAUDE.md,CONTEXT.md,CHANGELOG.md,BUGHUNT.md,REPORT.md,TASK.md) live in their own repos — they are operational, not published docs. - Authoritative versions at last refresh: controller v0.59.0, agent v0.29.1, hub v0.11.0.