4.5 KiB
Session Log — 2026-05-28 — Discovery Deploy Verification + Server Rebuild Notice
User
- User: Howard Enos (howard)
- Machine: Howard-Home
- Role: tech
Session Summary
Following the discovery fixes committed earlier this session (ICMP ping fallback fcf5833, scan timeout/409 guard c6f1f73), Howard asked whether Mike had deployed the new builds. Checked the live server and agent versions against the git log to determine deployment status.
The CI auto-bumped the agent to 0.6.49 (commit c8c6952) after the ICMP fix landed. WIN-TG2STMODJG8 reported agent_version: "0.6.49", confirming the agent-side fix is live. A test scan was run and found 25 unique devices — up from 9 previously — confirming the ICMP ping fallback is working and discovering machines that had all TCP ports firewalled.
The server-side fix (scan timeout + concurrent scan 409 guard) is not deployed. The server Cargo.toml is at 0.3.36 but the live binary is still the old version. This was confirmed by triggering two back-to-back scans: both returned new scan IDs instead of the second getting a 409. The two test scans ran concurrently and both completed normally (WIN-TG2STMODJG8 was connected), but the new_devices counts were inflated because both scans claimed credit for the same newly-inserted IPs within their overlapping windows.
A coord message was sent to Mike (id: 88c733a8) requesting a server rebuild and service restart on 172.16.3.30 to activate the 0.3.36 binary.
Key Decisions
- Tested 409 guard by triggering two rapid scans — the cleanest live confirmation that the server fix is or isn't deployed. Two distinct scan IDs = old binary confirmed.
- Coord message rather than waiting — Howard doesn't rebuild the server; Mike owns that. Coord message is the right handoff.
Problems Encountered
- First coord POST returned parse error — newlines inside the JSON string literal caused the parse failure. Resolved by switching to heredoc (
--data-binary @-with<<'JSON').
Configuration Changes
None this session — no code changes, read-only verification work.
Credentials & Secrets
None created or modified this session.
Infrastructure & Servers
- GuruRMM server: 172.16.3.30:3001 — running old binary, needs rebuild to 0.3.36
- GuruRMM dashboard: https://rmm.azcomputerguru.com
- Coord API: http://172.16.3.30:8001/api/coord
- Discovery test site: eeb5f001-447b-4c1e-adc8-e18db2be9b5b
- Discovery node: WIN-TG2STMODJG8 — eee9f26d-0dbc-4b8e-8e42-3a901b4ff73a
Commands & Outputs
# Agent version check
GET /api/agents/eee9f26d... → agent_version: "0.6.49" (ICMP fix deployed)
# Server 409 guard test — two rapid scan triggers
POST /api/sites/eeb5f001.../discovery/scan → scan_id: b29bd012... (initiated)
POST /api/sites/eeb5f001.../discovery/scan → scan_id: fbd163b5... (initiated — NOT 409)
# → Server fix NOT deployed, old binary still running
# Both test scans completed normally
# devices_found: 10 + 25 (concurrent, inflated new_devices count due to overlap)
# Total unique devices in DB: 25 (up from 9 before ICMP fix)
# Coord message to Mike
POST /api/coord/messages → id: 88c733a8
# Subject: "Server rebuild needed — discovery timeout fix (0.3.36)"
Pending / Incomplete Tasks
- Server rebuild (Mike): 172.16.3.30 needs rebuild + restart to deploy server 0.3.36 (scan timeout + 409 concurrent guard). Coord message sent: 88c733a8.
- Discovery auto-deploy (P2): Not built — future feature.
- Discovery scheduling (P2): UI has options but backend not wired.
- SPEC-012 implementation: Sortable table headers, ready for sprint.
- SPEC-013 (P3): Deferred pending file transfer.
- SPEC-014 follow-up (Mike's): Policy tab for watch rules; push rules on agent connect.
- Cascades pending migration: M365 relicensing (time-sensitive), domain joins, HIPAA GPO, credential vault, Entra Connect expansion.
Reference Information
- Agent 0.6.49 CI bump commit: c8c6952 (0.6.48 → 0.6.49, after ICMP fix fcf5833)
- Server 0.3.36 CI bump commit: a0a00bf (0.3.35 → 0.3.36, after scan timeout fix c6f1f73)
- ICMP fix: fcf5833 — agent/src/discovery/mod.rs — ping_host() fallback
- Timeout fix: c6f1f73 — server/src/api/discovery.rs + server/src/db/discovery.rs
- Coord message: 88c733a8 — "Server rebuild needed — discovery timeout fix (0.3.36)"
- Discovered devices now in DB: 25 unique IPs (was 9 before ICMP fix)