The vLLM detour (a2b39b1, same day) is retired: the Q4_K_M llama.cpp setup
is the requested shape and the vLLM KV-quant options on sm_86 all measured
dead (fp8 hardware-impossible, int8 page-size unification, turboquant
mm-prefix). Root unit restored, vLLM quadlet/Dockerfile/fetch script moved
to .AGENTTRASH, runbook listed as same-day-reverted. Rollback target:
frank-small-vllm-gemma4.
Runtime swap on the RTX 3090 (port 7771):
- vLLM gemma-x86_64-cu129 dev build, digest-pinned, audio extra pip-installed
into localhost/vllm/vllm-openai:frank-small (Dockerfile, no COPY lines)
- google/gemma-4-12B-it-qat-w4a16-ct (10.26 GiB int4) + Google's bf16 MTP
assistant (845 MB) — the w4a16-ct assistant cannot draft in vLLM (drafter
path skips Marlin; acceptance 1.00), bf16 measures 2.80-3.56
- root systemd unit -> rootless Quadlet (dankclaw/frank-small/quadlet/);
manual CUDA device+lib mounts (no nvidia-container-toolkit on host)
- bf16 KV (fp8 hardware-dead on sm_86), 289,647-token pool, 2 concurrent
full-context sessions, CUDA graphs on
- native text+image+audio: the fleet's multimodal endpoint
- measured: pp@67k 1293 t/s, pp@213k 608 t/s, tg@213k ~13.6 t/s,
tg@85k ~17.6 t/s
- host driver 610.57.04 (Ubuntu 26.04); sm_86 is correct, no arch override
- planner/scoping: frank-small now selected by dankclaw/frank-small/ +
env/serve/download trio; old unit retired to scripts/.AGENTTRASH/
- docs: new runbook frank-small-vllm-gemma4.md; AGENTS.md, LLM index,
top docs index updated; old runbook moved to retired
The two R9700s sit on unequal PCIe paths: ROCm0 (03:00.0) is CPU-root-port
Gen4 x16 (~31.5 GB/s); ROCm1 (0b:00.0) rides the B650 chipset through
bridges negotiating Gen4 x1 (~1.97 GB/s). With the MTP drafter, output
layers and heavier compute buffers pinned to the last-layers card, live
rocm-smi read 98% busy on the x1 card vs 40% on the x16 card. Moving one
DeltaNet layer (34) to ROCm0 shifts ~0.43 GiB off the bottlenecked card;
the seam fit lands 30.59/30.01 with gap 0.58 GiB, inside floor and slack.
36:29 fails both gates, so this lever is spent.
The bridge-subnet-only pair dropped the host's own health probe
(source 192.168.0.143/127.0.0.1) and failed a healthy deploy. The
deployer now installs the full set: ACCEPT 127.0.0.1, ACCEPT the host
IP, ACCEPT 10.89.0.0/16, then DROP.
169.254.1.2 is not a local address on this host (netavark runs its
bridges in aardvark-dns' netns), so the relay cannot bind it — and a
loopback bind refuses container traffic. Probe-socket proof: a host
socket on 192.168.0.143:18094 answers the container's 169.254.1.2:18094
dial while the LAN IP refuses from the container. The relay binds
192.168.0.143:18093 and the deployer installs an iptables INPUT pair
(ACCEPT 10.89.0.0/16, DROP rest) so the unauthenticated endpoint stays
off the LAN.
A socket bound to 127.0.0.1 refuses container traffic that arrives via
the podman bridge: the first live webhook failed with 'dial tcp
169.254.1.2:18093: connection refused'. The relay now binds the
host-gateway endpoint — link-local to the podman bridge, off the LAN —
and the deployer's health probe follows it there.
- runbook: identity now 4x262144, audio verified, throughput table with
the OOM and negative-result rows, deployed flags updated to reality.
- AGENTS.md frank-small paragraph matches the deployed shape.
- log.md: the sweep entry.
- scripts/frank-small.env: my earlier line-range edit had clobbered
NGL=999 and FLASH_ATTN=on out of the file; the seams caught it. The
live service was unaffected because the script's defaults match the
declared values exactly, but the file contract is restored.
Suite: ALL TESTS PASSED (321), vault 125 pages 0 unresolved.
The 393216/34:31/MTP deploy measured 30.78 GiB on ROCm1 against the
fit's 28.32 — the MTP draft buffers (one nextn layer's KV, 4096 B/pos,
plus draft buffers) came in 2.46 GiB above the estimate. Recalibrate
the seam's MTP constant to the measured 4.20 GiB total charge on card
1 and drop the floor 1.5 -> 1.0 GiB for this shape (ROCm1's 1.08 GiB
headroom is real; 35:30/36:29 fail card 0's floor at this pool, so
34:31 is the only split that fits 2x196k).
AGENTS.md and the runbook move to the MTP drafter: 34:31, 393216 pool
(196608 per conversation), the DFlash2 sidecar documented as a
one-line revert, and a dated part-2 section with the fit/measured
table and the recalibration.
AGENTS.md topology, Active Services, the CUDA-quant-KV lesson (f16 is no
longer why), and the runbook index. New frank-small-gemma4-12b runbook:
payload digests, the SWA/q8 arithmetic behind --fit's 262k window, the
MTP-sidecar + ngram-mod chain, the always-accepted reasoning_effort, and
the fit-probe error that looks like a fault at boot. The Qwen runbook and
both indexes carry the retirement; work log updated.
AGENTS.md topology row and section, and the runbook: the pool is 327680
positions (163840 per conversation over two slots), the split is 36:29 -- one
layer past the 34:31 balance at this context, on purpose, for headroom on the
drafter card (fit 29.19/26.04, measured 28.40/26.66 GiB at load, 3.46/5.20
headroom) -- and vision is on via the explicit --mmproj flag, because this
build has no autoload (the first deploy came up text-only and 500'd image
requests). The rope caveat is written down where the cap lives: the GGUF has
no scaling keys, so the top 65k of the pool extrapolates on the base rope.
New page [[frank-medium-batch-ubatch]]: the sweep's method (one invocation per
cell, a VRAM sampler per cell, 64k prefill measured AT 64k depth), the full
batch x ubatch matrix, the depth-0 control, the deployed-shape VRAM table, and
the limitations — including the one that cost me two runs, that amdgpu does not
return ~26 GiB at once, so a reload after a kill fails with "unable to allocate
ROCm0 buffer" and reads exactly like a configuration that does not fit.
AGENTS.md: the batch lesson now carries the measured curve (the peak at 512, the
10 % loss at 2048, the irrelevant logical batch, the invariant decode) instead
of the "~3.3 vs ~1.7 GiB" figure, which was measured at the bench's shape; at
the deployed shape the asymmetry is 2.41 against 1.30. Fleet row and the
frank-medium section now say 35:30 and batch 6144x512.
Runbook: deployed flags, the split history with the ubatch-512 table, the fault
note, and the verified VRAM (6.44 / 6.90 idle, 6.23 / 6.64 under two concurrent
requests).
Dual-R9700-Cooperation: the live capability table (card 0 ReqEn+ against card 1
ReqEn-, x1 effective link on card 1) and an honest label on the M.2_2 riser
premise — it is ASUS's spec sheet, not a measurement, and CPU-attached does not
imply Routing+ on this board (00:08.3 is CPU-side and Routing-).
log.md: the round trip recorded, newest first.
pcie-link-test.cu (CUDA, pinned buffers, cudaEvent timing, both services stopped):
pinned H2D 6.72 GB/s and D2H 6.60 GB/s, i.e. 85 % of the Gen4 x4 ceiling. Async
is no better than sync and pageable sits within 1 % of pinned, so these copies are
link-limited rather than staging-limited — the usual "pageable costs 2x" does not
apply on this host. A 3 MiB slab costs 474 us, essentially 3 MiB / 6.72 GB/s plus
~27 us of launch.
That settles the expert-streaming question quantitatively: 1.70 GiB of active
experts per token (A3B @ Q4_K_M, 75 GiB expert mass) is 272 ms of link time ⇒
3.7 tok/s, 4.6 with a fifth of the experts resident. The same bytes computed in
host RAM at 256 GB/s cost 7.1 ms — the link is 31x slower than the memory behind
it, which is why llama.cpp's -ot exps=CPU path measured 20.46 tok/s while an
engine that streams loses by 3-5x. The port under the card is x8-capable and the
card is x16-capable, so the x4 is the adapter; doubling it still leaves streaming
well behind.
Also records FreeToken's own constraint: it keeps experts in host RAM with an LRU
GPU cache (exactly this streaming shape) and loads only modelopt block-FP8
safetensors for qwen3_5_moe, so Q4_K_M GGUF is unloadable there regardless.
Operator instruction: purge the cache so no run can serve a stale payload. The two
live payloads were cleared and then re-fetched rather than left absent — deleting
them without a re-fetch means neither endpoint starts.
halogen's re-fetch proved the point: its sidecars had been stale *in place*
(overlay.hgn 2,477,677,120 -> 2,572,466,560 bytes, overlay-speed.hgn
2,383,306,048 -> 2,478,095,488) while the 115.55 GiB checkpoint returned
byte-identical. frank-small's 27B and mmproj re-fetched to Hub-matching hashes,
unchanged.
Trap recorded: podman refuses a bind mount whose source directory is missing
(statfs ... no such file or directory), so frank-large needs /models to exist
*empty* for its first-start fetch to run.
Q4_K_S against the recorded Q4_K_M numbers: +2.1/+1.6/+2.6 % (iGPU-only pp512 /
pp16384 / tg64), +2.7 % tg32@64k, +4.1 %/+2.0 % for the two layer-split configs,
~0 on both pp512@64k, host RAM bottoming at 10.0 GiB against ~5 GiB. A memory
lever with a small speed dividend. IQ4_XS (97,473,155,360 bytes, sha256
b4787265...d0574f9d) is pulled and verified for the next comparison.
Also repairs an AGENTS.md edit that landed in the wrong section: the peer
session's commits shifted that file between my read and my write, so the
frank-large text briefly split the Model Config section.
Checking whether a download was still running with pgrep -a curl printed the
whole argv, Authorization header included, into the session transcript. The
token is not in the repo and nothing was committed, but the practice is wrong:
secrets belong in a -K config file or the environment, not in argv, and no
pgrep -a should ever target a command carrying an auth header.
DSpark came off after a day. It was the better-looking file (1.27 GiB against
1.92) but it declares no sliding window, and the draft context is built with
n_ctx = llama_n_ctx(ctx_tgt) -- so its KV spanned the whole 262144-token window
(5.0 GiB) and cost ROCm1 four layers of split to carry. DFlash2's window bounds
its draft KV at 2048 positions (~40 MiB), which frees ~4.3 GiB on the
draft-carrying card.
The balance point moved with it, so it was re-measured at the deployed batch
(6144 x 1536, 2 slots, 262144) rather than interpolated:
39,26 2.42 / 7.20 37,28 4.28 / 5.34 35,30 5.14 / 4.48
38,27 3.85 / 5.77 36,29 4.71 / 4.91 34,31 6.57 / 3.05
36:29 is the minimum (0.20 GiB apart) and leaves 3.2 GiB above the operator's
1.5 GiB floor on each card, so the full window stays. One -ts step is one layer,
and qwen35's layers are not the same size -- 0.43 GiB on a DeltaNet boundary,
1.43 GiB on a full-attention one -- which is why the table exists.
Also in here:
- the fetch script pointed at the wrong file (DSpark) AND declared the wrong
size for it: 1359284737 against the Hub's 1455376576, 96 MB short, which made
--verify-only fail on a payload that was in fact correct. Repointed at z-lab
DFlash2 Q8_0, 2056414816, and both files' sha256 were checked against the
Hub's LFS oids (host == Hub for both).
- the seam's fit was recalibrated for a window-bounded draft KV and for the new
point; it now passes 36:29 and fails every other split measured, including
1:1. Its ubatch term is an extrapolation and it cannot model contiguity, which
the comment says out loud.
- the serve script's dead dspark-only tunables (n-min, p-min, p-split, poll)
are gone; the build's defaults apply.
The operator's revised spec: batch x3 (6144/1536), parallel 2, a 1.5 GiB per-card
floor, and the context maximised inside it.
Measured at 262144 with the old 38:27 split, 3x batch was brutally asymmetric:
ROCm1 (output layers + the drafter's graphs) paid ~3.3 GiB where ROCm0 paid
~1.7, leaving the draft-carrying card at 0.08 GiB free against ROCm0's 3.78.
One -ts step is worth ~1.5 GiB at this batch, so the balance is a measurement:
41:24 overshot (1.47 / 2.38), 40:25 landed at 1.92 / 1.96 GiB free.
That balance is also what maximises the context -- an unbalanced split would have
to buy its headroom out of CTX_SIZE, and CTX_SIZE is already at the model's
trained window (262144), now split over two slots of 131072. Verified under two
concurrent 1832-token prompts: 200/200, 1.87 / 1.86 GiB free, n_slots = 2,
n_ctx_slot = 131072, tool calls intact, and DSpark drafting on BOTH slots
(acceptance 0.38/0.22 and 0.52/0.43).
8192/2048 does not fit at this context at all: its 2020 MiB pp buffer fails to
allocate on ROCm1, and 4096/1024 fits but leaves 1.11 GiB -- both recorded in
the env's sizing table.
The unit seam now models both cards (share + draft + ubatch-scaled buffers,
calibrated to the rocm-smi figures to within 0.002%), asserts the 1.5 GiB floor
on each, and asserts the two cards stay within 0.5 GiB of each other. Both
assertions were confirmed to FAIL on a deliberately bad config (ubatch 8192, and
a 1:1 split) rather than pass by construction.
Correcting a claim I made from a bad query: Q4_K_S does exist for this finetune.
`mradermacher/Qwen3.8-Flash-Next-Uncensored-i1-GGUF` ships the large quants as
`.partNofM` fragments — Q4_K_M 37.00+37.00+36.97 = 110.97 GiB (which is how the
deployed file was assembled) and Q4_K_S 35.00+35.00+34.09 = 104.09 GiB — while
the smaller rungs are single files. A tree listing filtered on `.gguf` makes
Q4_K_S invisible, which is exactly how I concluded it did not exist.
Also records the fetch trap that hid it for a moment: `curl -o FILE URL` exits 0
on 401/403, so a gated repo produced three 153-byte "shards" with identical
sha256 values and a cheerful DOWNLOAD_DONE in 20 s for 104 GiB. -f plus the
Hub's declared size plus the LFS sha256 is the standard the fleet's own
download-frank-medium-gguf.sh already meets.
DSpark declares no sliding window, so its draft context -- built with
n_ctx = llama_n_ctx(ctx_tgt) -- caches the whole target window: 5 layers x 2 x 8
kv heads x 128 dim x 2 B = 20 KiB/position, i.e. 5.0 GiB at 262144, where
DFlash2's sliding_window 2048 cost ~40 MiB. At a 1:1 split that put ROCm1 at
33.9 GB of 34.2 GB and the first draft graph capture died inside
common_speculative_impl_draft_dflash::process with "ROCm error: out of memory"
-- after the model had loaded and /health had answered, which is why the deploy
looked like a crash-loop rather than a sizing failure.
Handing three target layers to ROCm0 restores headroom on both cards at the full
window. Measured at 262144: 28.52 GB / 30.71 GB, i.e. 5.30 GiB and 3.26 GiB free
(was 2.7 GB free on the draft-carrying card with 1:1 + DFlash2). The unit seam
now models the busier card as a skewed share plus the drafter's full-context KV,
calibrated to the rocm-smi figure.
AGENTS.md gains the durable lesson: read sliding_window out of the drafter's
GGUF header before sizing a spec-decode deployment.
I had written 'memory pressure from the 116 GiB iGPU runs' as the cause of the
pause-process death. journalctl -k records no OOM kill for that window and the
kernel says nothing else about it, so the claim was a guess wearing a finding's
clothes — in AGENTS.md, the x4 note, and the log. All three now state the
trigger is not established, and the note's 'co-tenants' patience' bullet no
longer asserts that causality either. AGENTS.md also gains the real lesson:
journalctl -k is the check, and an unknown cause does not need a story.
Asked whether Vulkan on both devices unlocks the split. It does not: the
per-tensor design aborts byte-identically, because the gate is not the driver
mix but `--override-tensor` — `has_tensor_overrides()` is a required term of
the scheduler's `pipeline_parallel` flag (llama-context.cpp:433, the only place
in the tree that reads it), and that flag sets `n_copies` in
`ggml_backend_sched_new`. With `n_copies == 1` a pre-allocated weight on a
non-primary device has no backend and the graph aborts.
Chasing it exposed a harness artifact that invalidated the earlier conclusion:
llama-bench splits a device LIST with `/` and SWEEP values with `,`, so
`-dev CUDA0,Vulkan1` was two single-device runs all along (and `-ts 0.85,0.15`
two sweep values, i.e. device 0 takes everything — which is where the bogus OOM
came from). With the correct syntax two-device splits work in both flavours:
pure Vulkan -dev Vulkan1/Vulkan0 -ts 0.85/0.15 -> 27.55 tok/s @64k (12.8 GiB on the 3090)
cross-vendor -dev CUDA0/Vulkan1 -ts 0.15/0.85 -> 27.23 tok/s @64k, pp512 268.5 (15.0 GiB)
iGPU alone 25.77 tok/s @64k, pp512 231.8
+7 % decode, +16 % prefill — real but not a regime change, and it costs the
3090 ~15 GiB, which this fleet spends on its only vision endpoint. The design's
premise still fails: whole layers carry their experts (1.62 GiB each), the 12
attention layers are spread every 4th, and the 32.8 GiB PLE must stay on the
main device, so KV cannot be consolidated on the 3090.
Also records the evening's incident (a dead rootless podman pause process took
all container management down; `podman system migrate` panics in 5.7.0; reboot
repaired it) and the netavark host-address lesson from the hermes 502.
The fleet now runs draft-dflash,ngram-mod, so the rule needs to be written down
where the next agent will look: draftless types (ngram family) are added to the
spec configuration unconditionally, while draft-mtp/dflash/dspark/eagle3 all
require a draft CONTEXT -- therefore one context-backed drafter plus any number
of ngram types is legal, and two context-backed ones never are ('draft-mtp,draft-dflash'
dies at load with 'model doesn't contain MTP layers'). Plus the operational
trap: ngram-mod prints no startup banner, so a working chain looks unengaged in
the journal.
Context: 262144 was measured before it was deployed -- 24.72 GB on GPU 0 and
29.14 GB on GPU 1 of 34.21 GB each, ~5 GB of headroom on the busier card, and a
140,069-token prompt prefilled through the single slot in 134 s to prove the
window is real. The 114688 wall was the vLLM predecessor's arithmetic, not this
runtime's.
Throughput, measured through the service: 1.0-1.4k tok/s prefill from 4k to
140k, 43.8 tok/s decode on a 512-token generation, 41-71 % DFlash2 acceptance,
and the prefix cache re-prefilling 4 of 998 tokens on a repeated prompt.
The signals move with the runtime, so the fleet's rules and dashboard are
re-cut rather than renamed: vllm: series do not exist any more, and two of the
old rules watched things llama.cpp does not export at all (KV-pool fill, p95
latency). Those are retired with the reason written down, and LlmSpecDecodeOff
takes one of the slots -- a lost drafter is the failure no client can see. The
observability seam now fails if any rule names a series the endpoint does not
export, which is what would have caught the stale vllm: expressions.
omp models.yml (outside the repo) is updated in the same pass: contextWindow
262144, and the full reasoning-effort map, because Qwen3.8's template answers
500 on 'high' -- verified against the live endpoint, low/medium/xhigh answer
200. Subagents run at high.
halogen-flash-server is ACTIVE again with its 2026-09-16 measured ledger (budget
33.5 GiB against a need of 27.2 — no clamp — and 14.7 GiB left for everything
else), ik_llama.cpp is retired with its measurements kept as the record of why,
and the new frank-small runbook covers the -mtp file, the 64 KiB/token KV
arithmetic, the MTP contract, and the reasoning_effort trap that cost me every
subagent on the endpoint.
AGENTS.md: topology, Active Services, runbook index, and three new llama.cpp
lessons (MTP is a property of the file; --jinja is what `tools` needs; a model's
template can reject the harness's thinking level). Three pre-existing defects
found while editing it: a truncated sentence mid-file, a harness path that does
not exist on either server, and the benchmarking stop/mask recipe still calling
frank-large a root unit.
Truthfulness notes worth keeping: the co-tenant stall NOTE prints again, so the
claim that it "no longer prints" is gone and the three observed block counts are
recorded instead of a guessed threshold; and the Observability page says out
loud why frank-large has no scrape job rather than leaving a silent absence.
New page: LLM/Dual-R9700-Cooperation.md — registered in the vault index, with
the measured evidence rather than the theory.
**The answer.** Card #2 sits behind the B650 chipset, whose bridges advertise
`AtomicOpsCap: Routing-` (read-only), so ROCr refuses to fill the hostcall
buffer RCCL's kernels declare and no RCCL op can dispatch on it. Two paths fix
it: an M.2→PCIe x4 riser into the free, CPU-attached M.2_2 slot (~$30; per
ASUS's spec page M.2_1 and M.2_2 are CPU-attached, PCIEX16_2 is chipset in x1
mode, and there is no x8/x8 bifurcation on this board), or a board whose CPU x16
bifurcates to x8/x8. Bandwidth is not the constraint: ≈10 MB per token-step ≈
1.4 ms at Gen4 x4.
**New measured results.** A bare 2-rank torch.distributed run: rank 0 (CPU-lane
card) passes broadcast/all_reduce/barrier; rank 1 (chipset card) fails all
three, and `send`/`recv` fails too — which kills pipeline parallelism as an
escape, since vLLM's PP handoff is exactly that primitive.
**Correction.** I previously said no software path exists. That overshot. One is
open and unverified: recompiling RCCL so its kernels stop declaring a hostcall
buffer (`-mprintf-kind=buffered` is the claim). Counter-evidence recorded: RCCL's
device source contains no live printf/assert/__assertfail, so the trigger is
unexplained, and the claim traces to search summaries, not a primary source. The
page carries the two-command test that would settle it. Also corrected: 0.9.3 is
radiance's version, not vLLM's — the base is vLLM 0.27.1 / torch 2.11.0 / ROCm
7.14, and its shipped RCCL exposes no `ncclDevComm*`, so torch's symmetric-memory
path cannot run.
**Rule added to the benchmarking policy**: it covers ANY GPU experiment, not just
benchmarks — including a bare 2-rank probe. Running one against the live service
on 2026-09-16 preceded the host going hard down (no ARP, no ICMP, power cycle
required); a hung AQL dispatch wedging the driver is the plausible mechanism.
The policy also now says how to tell off from wedged: `ip neigh` from a peer.
Retiring dankclaw/compose.yaml and adding alert rules invalidated claims across
the repo, including in the rule file agents read first. Clean cutover includes
the prose.
**Tests that had gone hollow.** Four assertions grepped dankclaw/compose.yaml
for a service name. With the file gone they passed by *absence* — a gate that
can no longer fail. Each now asserts the real invariant ("no dankclaw compose
file competes with the quadlet units") and keeps a live check for a re-added
entry, so it still fails when one appears: verified by recreating a compose file
with a comfyui service and watching three tests go red, then removing it.
**AGENTS.md, which was wrong about ComfyUI.** It said ComfyUI runs CPU-only
("no ROCm wheels for gfx1151, CUDA images wrong"). That was true in the
compose era and is false now: the restored unit runs with /dev/kfd + /dev/dri and
must NOT pass --cpu. An agent trusting the rule file would have re-broken it.
Also corrected: the observability paragraph ("RUNNING BUT UNCONFIGURED, no
datasource, no rules" — all three have been configured since 2026-09-16), the
Active runbook row that still named halogen for frank-large, halogen's absence
from the Retired table, the benchmark policy naming halogen's harness, and the
seam list.
**Unit and page claims.** prometheus.container said "no alerting rules anywhere
in this stack"; node-exporter.container said Dankserver runs from its compose
stack; Prometheus.md documented three jobs as working (they scrape HTML: they
can never report up) and said no dashboard existed; index.md and the three
authentik service pages pointed at the retired file; Authentik.md's upgrade
instruction told you to bump an image tag in it.
One bad edit of my own on the way (a stale line number clobbered a See Also
entry in Prometheus.md) — caught by reading back, fixed, and the vault link
checker now proves the cross-references hold.
The two replicas behind :7770 were on plain round-robin, which throws away
prefix-cache hits: an agent loop re-sends its system prompt and file context
every turn, and alternating replicas means every other turn pays a cold
prefill. Measured with 90k-token contexts, warm-turn TTFT was 2676 ms.
Pinning everything to one replica is not the answer either — a cold batch
(the shape a benchmark or a burst of parallel subagents produces) then loses a
card's worth of prefill: 16 requests took 30.1 s pinned against 18.6 s spread.
So the policy routes by request shape. A continuation (more than one message —
the agent-loop signature) is pinned to the replica that served its prefix;
anything else spreads across the cards by counter. Measured: 19.97 s for the
burst and 1095 ms warm-turn TTFT — the best of both, without a regression on
either workload.
The policy lives in proxy/routing.py, separate from the HTTP plumbing in
proxy/affinity-proxy.py, so it can be tested without aiohttp, a network or a
GPU. tests/frank_medium_proxy.bats covers it, and it earned its keep: the
first version hashed fresh prefixes to pick a card, which sends every member of
a same-prefix burst to one card (29.8 s, measured), and a stale index in the
pin path returned 500 on every continuation. Both are now assertions.
ROUTING=roundrobin and ROUTING=affinity stay available to re-measure the
choice on this host; the default is adaptive.
The radiance MXFP4 stack cannot run TP=2 on this host. RCCL's collective
kernels carry a hidden hostcall-buffer parameter, and ROCr refuses to
dispatch them on a device without PCIe atomics: the AQL dispatch fails with
hipErrorIllegalState ("the operation cannot be performed in the present
state"), which is what every two-rank collective hit. The second R9700 sits
behind the chipset bridges 05:00.0/06:09.0, which advertise
`AtomicOpsCap: Routing-` — a read-only capability, so no software can enable
atomic routing on that path and the driver's
pci_enable_atomic_ops_to_root() fails for that card. The first card, on CPU
root port 00:01.1, has atomics enabled.
Ruled out: kernel version (7.0.0-28/-30/-31 identical, after a full upgrade),
iommu=pt, P2P hardware (45.3 GB/s between the cards), the shipped librccl (a
from-source rebuild reproduces the failure at the same call path), vLLM and
pynccl (a bare two-rank torch.distributed allreduce fails the same way), all
NCCL transport knobs, and single-rank tests (NCCL short-circuits at
world_size=1, so one-card success proves nothing).
Deployment: three rootless Quadlets — frank-medium-a (:7771, card 1),
frank-medium-b (:7772, card 2), both TP=1 at 114688 context (262144 and
131072 do not fit one card's KV) with vLLM-profiled KV — and
frank-medium-proxy (:7770), a round-robin that preserves the fleet contract
of one endpoint and one served model name. frank-medium.container stays as
the TP=2 target for a host whose second card is atomic-capable; the deploy
handler does not install it and trashes any stale copy.
Measured through the proxy (vllm bench serve, random dataset, --ignore-eos):
conc 1 42.8 tok/s, conc 4 147.6 tok/s, conc 8 230.8 tok/s, 32k prefill at
~2.8k tok/s per card, TPOT p50 23.6-28.8 ms.
Also: healthcheck reads PORT so one probe serves all three units, the hermes
example and the harness models.yml drop to the real 114688 context, and the
runbook records the full evidence chain.
- Runbook with the measured envelope (prefill ~192 t/s at the VNNI ceiling;
decode ~3.8-4.3 t/s gated by 100 backend splits/token, GPU 95% idle), the
root-cause evidence chain, tuning levers, and co-tenancy traps
- AGENTS.md: fleet topology row + Active Services section for the ik
deployment; halogen demoted to rollback; bench policy unit type fixed
- LLM + root indexes relinked (frank-large-ik-uncensored ACTIVE,
halogen-flash-server retired)
New runbook page serverAppConfigs-docs/LLM/frank-medium-rocmfpx.md records
the hardware swap, the self-contained gfx1201 build, the measured layer-vs-
tensor decision (llama-bench multi-device syntax gotcha included), the
tensor-mode draft abort, served-path numbers, and the post-cleanup host ROCm
posture. AGENTS.md, README, docs indexes, Dankserver System Specs,
deployment/Hermes/OpenWebUI/Plex/Stack pages follow the rename;
freetoken-frank-small is marked RETIRED and kingjones-q8agent SUPERSEDED.
Work log entry added.
Predicted ~10.8 GiB KV / ~33.4 need / ~10.5 host-left; the boot ledger
printed ~11.0 / ~31.5 / 100.0 GiB held with 9.1 GiB + 1396 contiguous
2 MiB blocks left. Stall NOTE no longer prints at this footprint. Also
documents the ~45-60 s first-request PING silence after a restart (the
healthcheck's HealthStartPeriod covers it; do not restart the unit).
The 09-12 doubled pool fit only at reserve 15, leaving 7.1 GiB / 310
contiguous 2 MiB blocks for the ~28 co-tenant containers and making the
engine print its minutes-long-stall warning on every boot. 401,408
positions (~10.8 GiB KV) hands ~3.4 GiB back while keeping MAX_TOK=32768
and reserve 15: modelled need ~33.4 GiB vs budget 38.5, so the fit check
should not clamp. Capacity becomes one full 262k conversation resident
beside a ~139k one, or three concurrent 131k chats (three slots).
Docs: AGENTS.md fleet table + halogen section, runbook flags table /
co-tenancy tense / admission-arithmetic paragraph updated for the trim.
Ledger confirms the change landed: "3 slot(s) over one 524288-position pool",
"slots: 1 -> 3 POOLED, 111.0 MiB each", held 103.5 -> 103.4 GiB.
Recorded the oversubscription nuance that the old row missed: the slots were
never the binding constraint on context. 4 x 262144 exceeds the 524288 pool,
so admission was pool-bound all along -- two full-length conversations, or
four at 131k. Lowering the cap trims simultaneous requests, not capacity.
Raised /swap.img from 8 GiB to 64 GiB on Dankclaw. Host-level state with no
repo file, so it was invisible to the repo until now: fstab line 12 drives the
generated swap.img.swap unit, ext4, fallocated hole-free, swappiness 60,
zswap off. 748 GiB still free on /.
The reusable part is the ordering. The old file held 4 GiB of live pages and
MemAvailable read 73 GiB while only ~5 GiB was genuinely free -- the pinned
model weights are counted as reclaimable page cache, so the headline lies by
the size of the model. Swapping off before the replacement is online would have
pushed those pages into RAM that does not exist. Creating and enabling the new
swap first let them migrate instead: swapoff took 65 s and the host never
wobbled.
New page serverAppConfigs-docs/Dankclaw/Swap Configuration.md, linked from
the index, plus a pointer in the AGENTS.md memory-model section so nobody
reasons about headroom without checking swapon.
Old 8 GiB file staged at dankclaw:~/.AGENTTRASH/20260912-swapfile-8g/
pending approval to purge.
Seven restarts measured. New section "The reserve cannot buy air": reserve 16
was deployed and held exactly what 15 holds — 103.5 GiB, 7.1 GiB left to the
host. The flag is a term in the fit-check inequality, not a reservation, so it
has only two regimes: below the pool's requirement it changes nothing
observable, above it the pool silently disappears.
Sizing rule now defers to the engine's printed requirement (budget >= 36.5 for
524288 at max_tok 32768, from "need ~35.0 GiB plus 1.5 of margin") over my
modelled 37.0, and the ledger table carries the reserve-16 row.
Ledger confirms the doubled pool is actually reserved (524288, 13312.0 MiB)
at max_tok 32768 with no clamp and no WARNING: 38.5 budget vs 37.0 needed.
Recorded the cost honestly. The server now holds 103.5 of 121.2 GiB and the
engine reports only 310 contiguous 2 MiB blocks / 7.1 GiB left for everything
else (24-reserve baseline: 2081 blocks / 12.1 GiB). All 28 co-tenants stayed
up, but the engine volunteers a new note: when memory runs out, frank-large
AND its neighbours "can stall for minutes at a time at full CPU with no disk
activity and no output, which is not a crash and needs no restart". That is
the least diagnosable of the three failure modes seen today, and `free` hides
it — MemAvailable still reads ~74 GiB because the pinned weights count as
reclaimable page cache.
Sizing rule now expresses the budget as a function of reserve, shows both
endpoints (24 -> 29.5, 15 -> 38.5), and the max_tok ceiling is labelled as
reserve-conditional (<= 18,130 at 24; <= 35,700 at 15).
Deployed HALOGEN_MAX_TOK=32768 by operator call (upstream default, full prefill
speed) and confirmed the prediction from the ledger:
kv pool: 524288 positions need ~35.0 GiB ... LOWERING THE POOL TO 262144 (~27.8 GiB)
memory: 68.0 weights + 7.2 KV pool + 21.1 working = 96.3 GiB in all
so the effective pool is 262144 again and the doubling this task set out to do
is not in force. Healthy, completion verified. Left the pool requesting 524288
deliberately: with KV_POOL_FIT on that is a grow-if-RAM-frees-up declaration,
not a mistake.
Two corrections to what I wrote earlier today:
1. The engine's sizing block, and its "WARNING even 262144 ... host RAM cannot
spare it" line, print ONLY when the fit check had to clamp something. The
09-11 baseline boot prints no kv pool: sizing lines at all. So the absence
of a warning in old logs was never evidence of headroom.
2. The budget needs a vision term. Actual need is pool + arena + 1.97 GiB of
vision tower, not pool + a 1.5 GiB nominal margin. That makes the ceiling
for a doubled pool max_tok <= 18,130 rather than 19,115, and it means the
32768/262144 corner this box has run in since 09-11 is ~0.3 GiB OVER its own
stated budget -- and has been the whole time. It boots and serves with
12-14 GiB genuinely free because the 24 GiB reserve is generous. A soft
floor, not a wall.
The arena model also reproduces upstream's three published "on the device"
figures exactly (35 / 42 / 41 GiB), confirming they are pool+arena totals at
the default max_tok and not pool-only -- which is the misreading that made the
doubling look affordable in my first estimate.
HALOGEN_KV_POOL_POSITIONS 262144 -> 524288, which is upstream's own
default (2 x HALOGEN_CTX). The old pool was the binding constraint on
concurrency, not context: a request reserves prompt + max_tokens against
the pool, so with pool == CTX four concurrent 65k chats were contending
for one full-length conversation's worth of positions. Two full-length
conversations, or four at 131k, now stay resident.
Cost is ~7.2 GiB more device RAM (measured 28 KiB/position), taken from
the page cache the 47.7 GiB n-gram lookup table reads through. Live
startup ledger for the prior config: 68.0 weights + 7.2 pool + 21.1
working = 96.3 GiB held, 12.1 GiB left on the host. The doubled pool
puts the remainder near 5 GiB against a ~4.3 GiB co-tenant working set,
and cache at ~69 GiB -- still above the lookup table, so the table can
stay hot. Compensating lever if long prompts start crawling:
HALOGEN_MAX_TOK=16384, which frees 8.3 GiB of prefill arena for ~8%
of prefill throughput.
Docs updated to match, including the admission-trap paragraph, which was
only true while pool == CTX.
The 2026-09-12 homepage outage slipped through because the live-fleet
Playwright suite was a manual 'MUST invoke' that deploys skipped. Wire it
into deploy.sh as the [3/3] gate: after a behavior-affecting deploy,
run tests/e2e/run.sh against the live fleet; a failure exits DEPLOY
INCOMPLETE. Add --skip-e2e as an emergency escape hatch. Update the
AGENTS.md deployment-completeness contract and flow diagram to match.
Verified: bash -n clean, --skip-e2e parsed, unknown flags rejected,
seam suite green (only the pre-existing comfyui test 73 fails).
Rewrites the [[Observability]] umbrella page and the five component pages for
what is actually deployed: rootless Quadlet units instead of compose services,
the loopback-only publish posture (node-exporter's LAN binding explained), the
loki healthcheck living in its image because the image has no shell, and the
Home=/root trap that killed the first grafana start.
Adds the note that matters most to a reader: the stack is installed, healthy
and UNCONFIGURED. Grafana reports zero datasources and zero dashboards with an
empty provisioning directory; Prometheus has no rule files and no Alertmanager;
Loki is configured but has no reader; node-exporter runs its default
collectors. Alloy is the only configured piece. Metrics and logs are collected
and retained, and nothing is visualized or alerted on.
AGENTS.md's boot-persistence gap note is replaced by the resolution.
Rebooting Dankclaw for the amd_iommu=off change exposed five services that
exist only as compose.yaml services with no Quadlet unit (prometheus, grafana,
loki, alloy, node-exporter): they did not come back and had to be restored by
hand. The boot-persistence rule already requires units for them; this records
the concrete gap so the next reboot is not a surprise.
frank-large moves from the kingjones30/ROCmFPX llama.cpp build to
peonist-ai/halogen-flash-server 0.5.8 serving Qwen3.8-Flash-Next (w4b
.hgn 115.55 GiB + quality overlay). Upstream measures this engine at
roughly 4x the prefill throughput of the llama.cpp family on the same
gfx1151 silicon, and prefill dominates any prompt with real context.
Shape: a rootless Quadlet container (dankclaw/frank-large/quadlet/),
engine on container loopback :8730 (never published), OpenAI front-end
:8731 published as host 7770, HALOGEN_MODEL_ID=frank-large so no client
renamed anything. Context 262144, KV pool 262144 positions (~7.2 GiB,
conservative for a shared 121 GiB host), 4 slots, host reserve 24,
vision tower on, weights pinned, all other flags at vendor defaults.
Weights are fetched by HALOGEN_DOWNLOAD on first start into a writable
/home/dank/models/halogen.
The llama.cpp era (unit, env, serve script) is staged in .AGENTTRASH/
rather than deleted; the STRIX payload and the ROCmFPX build stay on
disk as the rollback target. deploy_llm_large becomes a user-quadlet
handler: it retires the old ROOT unit (moving its checkout symlink into
the home trash, never rm) and gates the first start on the download
rather than reporting a failure for a fetch that is working exactly as
designed.
frank-npu is retired in the same change. The kernel cmdline moves from
iommu=pt to amd_iommu=off for the 13-16% of prefill upstream measured,
and that takes the XDNA2 NPU with it: without PASID/SVA the driver
cannot open /dev/accel/accel0 (err=-19). Its five dependent routes are
repointed (Hermes vision -> frank-large, Hermes title generation ->
frank-small, harness vision -> frank-large, tiny -> frank-small, task ->
frank-large:medium) and its provider blocks are removed from both the
harness config and Hermes's live config. Audio input leaves the fleet
with no replacement; image input survives on halogen's vision tower.
Docs: new halogen-flash-server runbook, new IOMMU decision record, and
frank-npu rewritten as a retirement record. AGENTS.md topology, runtime
sections, benchmarking policy (frank-large is a user unit now) and the
harness model-config section are corrected.
FrankBook-Pro becomes the only *arr/qBittorrent deployment. Every fleet
container (qB + five arrs) now shares gluetun's netns, so all egress —
torrent, indexer APIs, metadata lookups — exits the WireGuard tunnel by
construction (AllowedIPs=0.0.0.0/0); the arrs' own ports publish on the
gluetun unit, none on the arr units themselves.
- dankserver/compose.yaml: remove the six arr services (the Aug 20 compose
copy — never healthy, superseded; config dirs trashed on the host)
- dankclaw/Caddyfile + shared/homepage/services.json: arr vhosts and
probes repoint 192.168.0.235 -> 192.168.0.112 (FrankBook-Pro)
- frankbook-pro units: five arrs join gluetun netns (Wants/After/Network);
gluetun publishes 9696/8989/7878/8686/8787 alongside 8080/6881
- tests/frankbook_quadlet.bats: new VPN egress seam — every arr+qB unit
must share gluetun's netns and publish no port; gluetun must publish the
whole fleet's ports
- deploy-plan.py: gluetun changes carry the whole netns family (a gluetun
recreation kills all six joiners); dankserver compose set loses the arrs
- AGENTS.md: deployment is complete only when unit tests AND the Playwright
E2E suite both pass
Mechanical identity rename across the knowledge base, AGENTS.md, README, the
hermes config example, and homepage services.json. No behavior change; served
aliases on the LLM endpoints change with the deploy of commit 69f87c8.
Every Caddy-proxied service is now either behind the auth chain or
explicitly internal-only, with the provider-first flip procedure codified.
The comfyui GPU-access seam test now asserts the deployed contract
(group_add keep-groups + udev MODE 0666 on kfd/renderD*) instead of the
pre-fix 'video' group assumption.
Env file is the source of truth (unit ExecStart runs the serve script which
sources it); serve-script default and the deployer health-gate URL flipped in
lockstep so the post-restart health gate polls the new port. Docs, fleet table,
and both index bullets updated to match.
Prior docs claimed unknown tags silently serve llama3.2:1b. Live probes on the
live endpoint disprove it: /v1/chat/completions answers HTTP 200 with a flat
{"error":"Failed to load <tag> model!"} body and no choices; /v1/completions
answers a nested OpenAI-shaped server_error (code 500) under HTTP 200. Valid
tags route per-request to the named model (response echoes the requested tag);
no restart needed to switch between registry tags. Corrected in the runbook,
the AGENTS.md runtime section, the serve-script header comment, and the
historical FastFlowLM incident note.
Frank