coulomb/kaizen-agentic
4
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
kaizen-agentic/engagements/pilots/eng-coulomb-railiance01-ho-001/vault/baselines/railiance01.md
tegwick 6ca167ce19
Some checks failed
CI Smoke / host-smoke (push) Successful in 1s
Details
CI Smoke / container-smoke (push) Successful in 36s
Details
ci / test (push) Failing after 3m41s
Details
ops: complete host-operator ramp-up on railiance01 (WP-0009 T10) 
Live observe session verified SSH access, captured baseline and Critical
health/load findings (memory pressure, k3s API unavailable). RU checklist
closed; engagement phase operating; schedule enabled; no privileged changes.
2026-07-16 12:43:00 +02:00

2.3 KiB
Raw Blame History

Baseline — railiance01

Engagement: eng-coulomb-railiance01-ho-001 Status: captured Last captured: 2026-07-16 (observe session, T10 ramp-up) Access: ssh railiance01 (user tegwick, IdentityFile id_custodian_agent) — host_observe

Identity

Field Value
Hostname (kernel) 239.62.205.92.host.secureserver.net
Inventory name railiance01
Public IP (inventory) 92.205.62.239
Inventory ref railiance-hosts
Role Single-node k3s production (forgejo/apps, activity-core, platform services)

OS

Field Value
Distro Ubuntu 24.04.3 LTS (noble)
Kernel 6.8.0-87-generic #88-Ubuntu SMP PREEMPT_DYNAMIC
Arch x86_64
Uptime at capture 129 days, 13:34
CPU cores (nproc) 2
Memory 3.8 GiB total
Swap none
Root FS /dev/sda1 ext4 96G, 54% used (52G/45G free)

Capture commands (non-destructive)

ssh railiance01 'hostname; uptime; uname -a; nproc; free -h; df -hT; cat /etc/os-release | head -8'

Envelope seed (2026-07-16)

Metric Baseline Notes
Load 1/5/15 ~12 / 11 / 16 ≫ 2 cores — overloaded
MemAvailable ~0.1–0.5 GiB severe pressure; PSI full avg60 ~24%
Disk / 54% OK for capacity; watch growth
/var/log ~5.7G journal alone ~4.1G
Top host processes k3s server, gitea, temporal-server, activity-core/state-hub python
k3s API ServiceUnavailable at capture control plane unstable under memory pressure

Known quirks

  • Single-node production: no HA; memory headroom is the binding constraint
  • DinD / Actions runner privilege model (railiance-hosts ADR-004)
  • ops-bridge reverse tunnels to localhost ports (state-hub 18000/18001, etc.)
  • sshd journal noise: reverse-forward port 18765 already in use (tunnel contention)
  • Inventory hostname vs DNS: reverse DNS is HostEurope generic name

Access path (RU-01)

Method Result
ICMP to inventory IP reachable
ssh railiance01 (tegwick) success — non-interactive shell
ops-bridge tunnels to host connected (state-hub-railiance01, issue-core-railiance01, …)
Privileged ops not exercised; sudo -n available for read-only needrestart/ufw/k3s in this session