vh/esh-pfi-infrastructure
1
0
Fork 0
Code Issues 7 Pull Requests Actions Packages Projects Releases Wiki Activity

docs(nh3-pve): record GPU-install post-boot state — RTX 2000E at 01:00.0, iGPU hidden by BIOS, uplink moved to enp2s0f1np1

Browse Source
This commit is contained in:
vh
2026-09-25 11:22:00 -07:00
parent 58390a4785
commit 2bf418c510
2 changed files with 35 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files
persistent-memory.md
+19 -4
View File
@@ -1,6 +1,6 @@
# Persistent memory — eshpfi-management
_Last updated: 2026-09-25 ~1105 PT (⭐⭐ nh3-pve going DOWN for a GPU install, Prime on site; this session ends with nh3-dev — post-boot checklist below. Today: TEI = fleet embed/rerank engine, esh-ml1 sole backend + reward seat moved there, fv-ml1 GPU 1 freed ~15 GB; esh-ml1 monitoring wired; homepage wedge cleared; NIC names pinned on nh3-pve. 34 old entries archived.)_
_Last updated: 2026-09-25 ~1122 PT (⭐⭐ nh3-pve back from the GPU install, checklist passed; ⚠ iGPU hidden by BIOS → NanoKVM blind, needs on-site BIOS fix. Today: TEI = fleet embed/rerank engine, esh-ml1 sole backend + reward seat moved there, fv-ml1 GPU 1 freed ~15 GB; esh-ml1 monitoring wired; homepage wedge cleared; NIC names pinned on nh3-pve. 34 old entries archived.)_
> **Always check for `/tmp/infra-ops-handoff.md`** — if it exists and its
> `Written:` stamp is under **8 hours** old, read it (it carries the in-flight
@@ -117,10 +117,25 @@ no longer deployed sidecars here. See Recent decisions.)
_As of 2026-09-24 ~2245 PT._
### ⭐⭐ IN PROGRESS 2026-09-25 ~1100 PT: nh3-pve shut down for a GPU install (Prime on site)
### ⭐⭐ nh3-pve GPU install — host back 2026-09-25 1116, checklist run 1122
This session runs ON nh3-dev and dies with it. **Successor: work this list
first.** (No /tmp handoff: nh3-dev wipes /tmp at boot.)
**Result:** checklist items 1 and 3–6 pass, and nothing needed restarting. Guests,
NFS, the post office (push), svos, hermes-gateway, Booth, talk, ttyd, zellij,
backup timers, pbs-nh3 datastore, the mesh (irv-ml1 direct), the egress proxy and
NH3 DNS are all fine. Open items:
- ⚠ **iGPU gone → NanoKVM has no picture.** The BIOS Auto primary display picked
the RTX and hid `00:02.0`. It needs a BIOS change on site. See
`servers/nh3-pve/README.md`.
- The uplink cable moved from `enp2s0f0np0` to `enp2s0f1np1`. It is harmless
because both ports are bridge members. Recorded in the README.
- **pfi-gx10 booted 1108:54** after being off since 09-24 1435. That fits the AC
restore working, but only Prime knows whether he pressed the button. The README
keeps "NOT YET VALIDATED" until he says so.
- Pre-existing and not caused by the reboot: `althing-pump-infra-hermes` and
`althing-seat-daemon-jekyll` are both disabled (stopped 09-19 and 09-24). The
`infra-hermes` handle shows `pull`.
The pre-shutdown record follows, kept for the next GPU visit:
Pre-flight done before shutdown:
- NIC names pinned by MAC (a GPU renames every NIC on this MS-01), initramfs