Repository navigation
Open
Labels
area/infraInfrastructure, deployment, and operationsInfrastructure, deployment, and operationskind/choreRoutine maintenance, housekeeping, dependency hygieneRoutine maintenance, housekeeping, dependency hygienetask/triageQueue packet kind: verify, reproduce, and route this itemQueue packet kind: verify, reproduce, and route this item
Description
Drew's home lab build-out story (2026-11-04 ~23:30 CST), filed as a ticket set. These are Drew's hands-on physical tasks — the tickets structure his work; any config/doc prep named in a ticket is agent work.
Proposed sequence
Physical plant first, then switches, then routing/mesh on top of the plant, then interconnect, then the cluster on a stable network:
- Racking + units placed
- ops(lab): another 4 each of racking (as dictated) #6070 — another 4 each of racking (as dictated)
- ops(lab): install 3 more DXO10 units (model name as dictated) #6071 — install 3 more DXO10 units (model name as dictated)
- Dell switches (×3) — each: rails → firmware → verify → initial config → config stored → labels
- ops(lab): rack Dell switch 1 — rails, firmware, initial config, config stored, labels #6073 — Dell switch 1
- ops(lab): rack Dell switch 2 — rails, firmware, initial config, config stored, labels #6074 — Dell switch 2
- ops(lab): rack Dell switch 3 — rails, firmware, initial config, config stored, labels #6075 — Dell switch 3
- Edge routing + mesh
- ops(lab): reconfigure the edge routing and rewire #6067 — reconfigure the edge routing and rewire
- ops(lab): rewire the mesh Wi-Fi #6068 — rewire the mesh Wi-Fi
- Cluster interconnect
- ops(lab): cluster interconnect — connect the dozen cluster machines + 2–3 mini PCs #6077 — connect the dozen cluster machines + 2–3 mini PCs (physical links, addressing, switch ports, full-mesh reachability)
- Cluster bring-up — TALM + Talos + Cozystack: already ticketed as epic(cluster): TALM + Talos bootstrap #1086 (TALM + Talos bootstrap) and epic(cluster): Cozystack platform #1087 (Cozystack platform) under program epic EPIC: Basement Talos/Cozystack cluster program #1817; no new ticket filed (cross-link comment on epic(cluster): TALM + Talos bootstrap #1086). This is also the execution home for the tracked goal "Stand up lab cluster (talm, talos, cozystack, postgres, rustfs)" (goal_ba17a5777c21).
- Generic containers for ID + fleet — software-side, can be prepared in parallel; deployed once network/cluster are ready
- impl(lab): generic containers for ID and fleet purposes (Nix-built) #6076 — generic containers for ID and fleet purposes (Nix-built)
Side research (no sequence dependency):
- research(lab): are Raspberry Pis / SBCs worth anything in the lab? measured perf-per-watt vs cluster nodes + mini PCs #6078 — are Raspberry Pis / SBCs worth anything in the lab? Measured perf-per-watt vs cluster nodes + mini PCs, verdict per role (quorum witness / light web serving / free ID sandbox tier)
Notes
- Fourth switch: Drew said "3 or 4 Dell switches". Three switch tickets are filed. If a 4th switch exists, file it identically to the other three and add it here — no phantom hardware ticket is filed in advance.
- Verbatim items: two items are recorded exactly as dictated and flagged on their own tickets: "another 4 each of racking" (ops(lab): another 4 each of racking (as dictated) #6070) and the model name "DXO10" (ops(lab): install 3 more DXO10 units (model name as dictated) #6071). Both carry interpretation notes for Drew to correct on the ticket if the dictation garbled them.
- Related: EPIC: Basement Talos/Cozystack cluster program #1817 (basement Talos/Cozystack program), epic(cluster): network fabric #1084 (cluster network fabric), Epic: deployment architecture — services layer + deploy tool (tracking) #5765 (deployment epic), Containers: Nix-build audit — mcpx Containerfile, duplicate lab-ops Dockerfile, amd-crd definition-or-waiver #5771 (containers Nix-build audit), id: design: compile economics — compiled-world/DLL caching, compile-cost penalties, premium gating #5965 (ID compile economics / premium gating — SBC free-tier tie-in).
Metadata
Metadata
Assignees
Labels
area/infraInfrastructure, deployment, and operationsInfrastructure, deployment, and operationskind/choreRoutine maintenance, housekeeping, dependency hygieneRoutine maintenance, housekeeping, dependency hygienetask/triageQueue packet kind: verify, reproduce, and route this itemQueue packet kind: verify, reproduce, and route this item
Activity
developing-today commentedon Nov 5, 2026
CODE-1729
dezren39 commentedon Nov 5, 2026
Power dependency for this build-out set: #1105 (task(cluster): install and verify the dedicated 20A circuits, under the basement power epic #1083) — scope updated 2026-11-04 with panel review, wiring check, and 4–16 dedicated 20A circuits (count decided by the panel review + full-build-out load calc). Heavy racking (#6070) and the switch installs (#6073/#6074/#6075) sequence after or alongside the electrical work; the interconnect/cluster bring-up assumes the circuits it plugs into exist. Circuit → load mapping will be recorded on #1105 as circuits land.
dezren39 commentedon Nov 5, 2026
🤖 Agent note · done (triage)
Triage findings (verified fresh 2026-11-05 via
gh):kind/chore,task/triage,status/needs-triage,queue/queued,area/infra. No hard-stop labels (held,user/owned,do-not-merge,do-not-auto-edit) and nouser/review/user/escalated— but per the body this epic structures Drew's hands-on physical work; the agent-owned share is only the config/doc prep named inside individual child tickets. There is nothing for a queue worker to implement on the epic itself.Triage complete for task 10429; issue #6066 stays open as the tracking epic.