Commit Graph
88 Commits
Author SHA1 Message Date
Ilker ManapandClaude Opus 5.5 6289b01ed6 The cluster's settings live in the store, and every change takes effect
Everything in server.yaml that is not one machine's own is now a versioned
document in the store (cluster_settings), the same on every control plane.
The first start against a store imports the file's cluster keys, with the
baseline guard policy when there is none; after that those keys in a file
are ignored, with a warning.

Each control plane polls the document's version and its own server.yaml's
hash. A change is validated and applied: guard live (Gate.Replace), anything
else by a graceful restart that systemd brings back, standbys first and the
leader 30s later. A version that does not load on a machine is not applied;
at start the newest one that loads is used, so a bad edit cannot keep every
control plane down.

llmgw ctl config show/get/set/unset/edit/history/rollback, the admin API
under /api/v1/settings, and the console's Settings screen, which layers the
store's document over the machine's file, write it. Commands beside the
control plane that read cluster keys (init, node add, server add, backup,
upgrade) read the layered configuration; init --bundle writes offline to
the store once it holds the settings.

A standby that becomes the leader re-reads models, engine choices and
facility holds from the store.

Retires appendGuardIfMissing: the import does it.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-05 15:25:23 +02:00
Ilker ManapandClaude Opus 5.5 481523ca46 serve: tool calls declared, slots that hold the context, longer answers
A --context-length 32768 --ctx-pool 32768 model got llama.cpp's default four
slots of 8192, and refused what the gateway admitted; --parallel is now
what the pool holds of the context. max_output was 1024 for every model,
which cut an editor's agent off mid-file; it is now a quarter of the
context, from 1024 to 8192. --tools writes capabilities.tools, which the
Continue configuration now reads, as it reads max_output.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-05 09:30:40 +02:00
Ilker ManapandClaude Opus 5.5 df6d243ee7 The Continue configuration says what the catalog says of the model
It told Continue every model uses tools and nothing of its context: a model
whose calls come back as text stalled Continue's agent, and Continue sent
prompts far over a CPU replica's 4096 tokens, each refused. tool_use is now
written for a model whose catalog entry says tools, and the context length
with a fitting maxTokens whenever the catalog knows it.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-05 09:16:34 +02:00
Ilker ManapandClaude Opus 5.5 33334c8f33 Merge C24 federation: another llmgw cluster as a provider
Conflicts with facility-aware placement (config, control, admin API routes,
store test tables) all keep both; the generated reference and OpenAPI
document regenerated.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-05 01:29:35 +02:00
Ilker ManapandClaude Opus 5.5 40e9fd2ab1 Federation: another llmgw cluster as a provider, its models as PEER/MODEL
ctl federation add/list/remove/sync keeps peers in the store (migration
0029); their /v1/models are offered here as external models named
PEER/MODEL and kept in step, a model still named by a fallback kept
withdrawn. Peers are checked at /ready, timed, and taken out of routing
while down. Requests to a peer carry x-llmgw-hop and are refused (508) when
they come back to a cluster they passed through; residency and no_store
keep requests off peers; X-Request-Id and, with --send-tags, attribution
tags go along. ctl model fallbacks sets overflow to a peer. Console view,
metrics, admin API routes and docs.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-05 01:21:30 +02:00
Ilker ManapandClaude Opus 5.5 e8a62b23b7 Merge features-c2: the docs portal and its links in English and Turkish
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-05 00:15:25 +02:00
Ilker ManapandClaude Opus 5.5 11231d8c2b The portal in English and Turkish: a catalogue per language, a switcher at the foot of every page, landing.language, Turkish numbers and dates
Every template and handler message in internal/web goes through a JSON
catalogue under internal/web/locales/<code>/, with English as the fallback.
The language is the person's choice (a cookie set by /_lang/{code}), else
Accept-Language, else landing.language. Tests refuse a key some language
lacks, a translation with other arguments than English, and template text
left in English.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-05 00:12:53 +02:00
Ilker ManapandClaude Opus 5.5 f48f55a866 Merge C17 response cache
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-04 23:51:19 +02:00
Ilker ManapandClaude Opus 5.5 44f3a9e31d Reusing answers: a response cache per project, exact match, opt-in with a TTL
llmgw keys cache --project P --ttl 1h [--any] | --off, and Cache answers… on
the console's project menu. The same request — project, endpoint, canonical
model and the body without stream, stream_options, user, metadata, store and
service_tier, keys sorted — is answered from the gateway's memory for the
TTL. Only reproducible answers (temperature 0 or a seed; embeddings, rerank,
score) unless --any; only a whole 200 from the model asked for, not a
fallback's, a canary's, or one a guardrail changed.

Looked up after the limits and the input guardrails; the output policy
judges a reused answer; no_store and stored conversations bypass it.
Streamed answers are kept put back together and streamed back as chunks.
x-llmgw-cache: hit|miss|bypass, Age on a hit, Cache-Control: no-cache or
x-llmgw-cache: no to bypass. A hit is recorded with no tokens and no cost,
marked cache_hit (migration 0027) and shown by ctl request. An LRU bounded
by router.cache.max_bytes and max_entry_bytes; llmgw_cache_requests_total,
llmgw_cache_bytes, llmgw_cache_entries.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-04 23:50:53 +02:00
Ilker ManapandClaude Opus 5.5 e81f1cba4c Merge C24 docs portal and client guide
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-04 23:47:35 +02:00
Ilker ManapandClaude Opus 5.5 bd439aca29 A docs portal at /docs and a client developer's guide
The manual the binary carries, rendered on the client port: a sidebar of
the documents in the console's reading order (now one list, llmgw.Manual,
for both), heading anchors, links between documents kept in the portal and
links to files outside the binary sent to the repository, no script.
Public by default; landing.docs_sign_in sends visitors to sign in first.

docs/clients.md shows the official OpenAI and Anthropic SDKs against
llmgw in Python, TypeScript, Go and curl, with llmgw's headers; the
portal's index, the landing page and the account page link to it.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-04 23:46:42 +02:00
Ilker ManapandClaude Opus 5.5 30f7f3474a Forgetting a person and keeping requests in the cluster: users forget, keys residency --in-cluster
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-04 23:27:31 +02:00
Ilker ManapandClaude Opus 5.5 35657c1417 Secrets in Vault: env: and vault: references wherever a secret file is named, and the store's own secrets sealed under a master key — Vault transit or a local key
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-04 23:21:44 +02:00
Ilker ManapandClaude Opus 5.5 5b92ae0277 Token-level timing and admission by SLO: TPOT, inter-token and tokens a second by model and tenant; a model past its SLO is saturated for the lower bands
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-04 23:13:18 +02:00
Ilker ManapandClaude Opus 5.5 e502996198 Anthropic's Messages API: /v1/messages translated to chat completions and back, streamed or not, with tools and images; keys in x-api-key
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-04 22:59:55 +02:00
Ilker ManapandClaude Opus 5.5 9136bc71c2 More of OpenAI's API: Responses structured output, images and previous_response_id; rerank, score and audio passed through; moderations by the judge
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-04 22:56:25 +02:00
Ilker ManapandClaude Opus 5.5 32ff1dbd6f Requests kept and replayed: keys requests per project with retention and a share, ctl request --body and --replay
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-04 22:51:21 +02:00
Ilker ManapandClaude Opus 5.5 841ee21c12 Batches: OpenAI's Files and Batch APIs, served at the lowest priority under a store lease, each line a request of the key that made it
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-04 21:00:01 +02:00
Ilker ManapandClaude Opus 5.5 0a4076928a Canary routing: a model's canary takes its share of requests, a session stays on one side, x-llmgw-variant pins one, and the two are compared
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-04 19:56:16 +02:00
Ilker ManapandClaude Opus 5.5 29a7347df2 Requests can be tagged: metadata, x-llmgw-tags and the user field on the usage record, and counted by the tags router.tag_labels names
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-04 19:45:37 +02:00
Ilker ManapandClaude Opus 5.5 178964cf6a gofmt
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-04 16:40:27 +02:00
Ilker ManapandClaude Opus 5.5 14ea59d83f Provisioning through SCIM 2.0: accounts and groups kept in step by the identity provider
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-04 16:40:15 +02:00
Ilker ManapandClaude Opus 5.5 5cff456b9b Let /login/oidc through without an API key: /login alone was exempt
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-04 16:37:24 +02:00
Ilker ManapandClaude Opus 5.5 4cf40644c9 Single sign-on through OpenID Connect: accounts, staff roles and memberships from the provider
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-04 16:36:15 +02:00
Ilker ManapandClaude Opus 5.5 f8fcecd640 Audit log: every change through the admin API and the administration pages, with ctl audit
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-04 16:25:56 +02:00
Ilker ManapandClaude Opus 5.5 02aac84ca5 The /admin/guard page, a Guardrails link, and the policies in force on the profile page
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-04 15:15:38 +02:00
Ilker ManapandClaude Opus 5.5 47e847423d Guardrail metrics and alerts: checks, latency, judge failures, fail-open, redactions, stopped streams, tenant blocks
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-04 15:11:50 +02:00
Ilker ManapandClaude Opus 5.5 43f6a585a6 A judge model on the cluster: role guard in the catalog, judge in a guard policy
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-04 15:09:31 +02:00
Ilker ManapandClaude Opus 5.5 ec737cffb5 no_store: refuse conversations, notes and models outside the cluster under a policy that keeps nothing
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-04 12:36:02 +02:00
Ilker ManapandClaude Opus 5.5 9b9349183b gofmt the guard hooks
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-04 12:23:02 +02:00
Ilker ManapandClaude Opus 5.5 979aa44804 Hook a guard into the router: judge requests before dispatch and filter answers as they are relayed
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-04 12:20:24 +02:00
Ilker ManapandClaude Opus 5.5 ba5158d463 List in /v1/models only the models the key may use
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-04 10:49:54 +02:00
Ilker ManapandClaude Opus 5.5 36cdbce125 Merge ha: replicated store, second control plane, join codes, setup and console screens
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-03 22:59:49 +02:00
Ilker ManapandClaude Opus 5.5 aa3efcebbe Add a standby control plane with server add, and draw the screens as windows
server add gives another machine this llmgw, a certificate, the
configuration and store address, lets it into the store through Patroni, and
starts it; --vip keeps a floating address on the leader with keepalived,
following a new /leader endpoint. Nodes learn every control plane from the
leader and notice a dead one by HTTP/2 ping instead of after the kernel's
fifteen minutes. The setup, console, node and init screens are drawn full
screen in one frame: title bar, sidebar, bordered panel, key bar, dialogs.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-03 22:29:12 +02:00
Ilker ManapandClaude Opus 5.5 ce006a6e6f Publish per-tenant request and budget series, and alert on them
Requests by outcome, tokens and first-token time per tenant, at the level
router.tenant_label names (team by default), and every budget's share spent
this period. Rules alert on budgets at 80% and 100%, tenants hitting limits,
failing, spiking past their weekly average, or waiting on first tokens.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-03 19:55:24 +02:00
Ilker ManapandClaude Opus 5.5 9f479b9a70 Show each team's first-token and latency percentiles, errors and refusals
Percentiles are read from the histograms kept with the usage sums and shown
per group on the usage tab and in the people page's panels. Requests refused
by a limit are recorded as 429s against their owner, free of charge, so a team
hitting its limits shows as such. Usage can be grouped by model.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-03 19:50:16 +02:00
Ilker ManapandClaude Opus 5.5 372ef64889 Count the GPU time each owner's models held, on the usage page
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-01 16:07:59 +02:00
Ilker ManapandClaude Opus 5.5 4f280a3546 Give each GPU node a vLLM build it can run, chosen on the cluster page
A node that cannot run the CUDA 13 build installs the same version's
CUDA 12.9 build from vLLM's GitHub release, with PyTorch's cu129 torch;
the install computes on the GPU before it counts. The build can be
chosen per node — auto, cu13, cu129, none — on the cluster page, with
ctl engine-build, or over the admin API; the control plane keeps the
choice and tells the node at once.

Refs #2.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-01 12:39:14 +02:00
Ilker ManapandClaude Opus 5.5 e40b9ac202 Load LoRA adapters at scale zero
--lora with --lora-init-without-apply leaves an adapter's scale at 1,
and llama-server applies it at that scale to any request without a
"lora" list of its own: every plain request for the model was served
with its adapters on. Found live on qwen2.5-coder-7b-instruct. Each
adapter is now loaded with --lora-scaled FILE:0.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-01 07:39:12 +02:00
Ilker ManapandClaude Opus 5.5 cf53904cc5 Serve LoRA adapters on llama.cpp models, asked for by name
A model's adapters are fetched by each node with its weights and loaded
unapplied at start; a request naming an adapter reaches the model's
replicas with llama.cpp's per-request lora field. Adapter names share the
model namespace, are listed, and follow allowlists on their own. The
router's config seed and pushed catalog now build the name map in one place.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-09-30 23:40:47 +02:00
Ilker ManapandClaude Opus 5.5 94eec557ad Serve a model from an external OpenAI-compatible provider
A catalog model with external {url, model, api_key_file} and no profiles
is published as an endpoint by the control plane, the key read from its
file each time; the router rewrites the request's model to the provider's
name and sends its key. It is never scraped nor stale, and nothing is
placed for it. Named in fallbacks, it serves when the cluster has no room.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-09-30 23:29:01 +02:00
Ilker ManapandClaude Opus 5.5 3d3c96e422 Displace the newest lowest-band request when the gateway queue is full
A total bound across bands (admission.queue_capacity_total, default half of
all bands' capacity together); reached, a request is not refused while a
lower band has requests queued, and the newest of the lowest band gets a 503
with Retry-After instead.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-09-30 23:07:53 +02:00
Ilker ManapandClaude Opus 5.5 f288f3c423 Set capacity quotas, class, owner and reservation from the admin pages
The quotas page gains a capacity section over the same API methods the CLI
uses; a model's card sets its class, owner, memory and cores, applying only
what changed. Headroom counts the replicas a model could get by preempting
what its class outranks, and the card and ctl model target use it.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-09-30 22:40:45 +02:00
Ilker ManapandClaude Opus 5.5 62657f4c2a Release v0.10.4: CPU threads, tracing, and the rest of the review
CPU replicas start on half the machine's CPUs, not all of them: two busy
llama.cpp servers each spinning on all 8 vCPUs fell to 0.8 tok/s apiece,
where at 4 threads they keep 5.5. llamacpp.threads sets it per node.

Inference requests can be traced over OTLP/HTTP (tracing.endpoint), with
admission, schedule and upstream spans and traceparent sent to the engine.

Security: onboarding no longer puts the cluster root in the system trust
store and removes it where it was; the mirror's mutual-TLS API can be
limited to named certificates (--api-client); init.log no longer keeps the
administrator's password; a warning when auth is off on a public listen;
the recorded actor is sanitised and sudo-aware; Ray's dashboard is off, and
the remaining engine gaps are documented.

reconciler.go and agent.go are split by topic, the status mapping of each
failure is tested, and the plan's checklist is brought up to date.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-09-30 04:00:35 +02:00
Ilker ManapandClaude Opus 5.5 3bf4b7b709 Tell clients how much of their rate limit is left
SDKs and editors pace themselves from OpenAI's x-ratelimit-* headers, and
the gateway sent none: only Retry-After, and only on the refusal they would
have avoided. A limit can be set at several levels of the tenancy tree, each
counted in its own pool, so the headers report the one with least left —
the one that will refuse first — after this request's own charge, on
refusals and successes alike. The windows slide, so the reset figure is the
next minute boundary, when a whole window has moved on. A limit that is not
set is not advertised.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-09-30 02:42:21 +02:00
Ilker ManapandClaude Opus 5.5 f8572fb5ee Follow one request from its id to the replica that served it
"Why was this request slow?" had no answer. Nothing tied a request to the
replica that served it, the time admission held it went out in a response
header and was never kept, and the only per-request log lines were for
disconnects, early stream ends and upstream retries.

Every request now has an id: the client's X-Request-Id when it is a plain
string of reasonable length, otherwise one the gateway makes. It is set on
the response before anything else, so a request refused as malformed can be
named too, and forwarded to the engine. One access log line per request
carries it with the model asked for and served, the replica, status, time in
the queue, time to the first token, total time, tokens, and the key,
project and user.

Usage records gain the id, the replica, queue time and time to first token,
tried against a restored copy of the live store before being shipped: 591
rows kept, columns and index in place. `llmgw ctl request ID` reads one back
through the admin API.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-09-30 02:36:45 +02:00
Ilker ManapandClaude Opus 5.5 91fe24a105 Close what a second security review found still open
A review of v0.10.4-rc.4, and a look at the live network, found:

- Alertmanager listening on every interface: from any node its API took
  silences, with no login. Bound to loopback on the monitoring host, and the
  monitoring README now installs it that way; Prometheus beside it is its only
  client, and one Alertmanager needs no cluster port.
- Login and signup unbounded in memory. Each attempt runs argon2 with 64 MiB,
  and nothing limited how many ran at once; the per-address throttle kept one
  entry per address ever tried. Hashing now takes one of a few slots and
  waits for one otherwise. The throttle expires entries whose window has
  passed when it fills, and if it is still full it refuses new addresses:
  evicting live ones instead would let a flood of made-up addresses reset the
  throttle on the address actually being guessed.
- Metric labels taken from the request. A refused request counted the model
  name it asked for, so any key holder could mint a series per request.
  Requests are counted under the catalog's name, or "unknown".
- Two in code written this week. A SQLite backup made its snapshot with the
  default mode in the listable backup directory, a full copy of the store
  readable while the backup ran; it is now made in a private directory, and
  the unit sets UMask=0077. And archive links were checked by name only, so
  "a -> ." then "a/a/b -> ../.." passed and a file written through them landed
  above the install directory, as root. Archives now unpack through os.Root,
  and a link may only point at a file beside it — which every llama.cpp
  b10991 archive, CPU and CUDA, was checked to follow.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-09-30 02:25:07 +02:00
Ilker ManapandClaude Opus 5.5 e751c89937 Serve the project's website at /about, with its fonts in the binary
The website was a file built by hand outside the repository. It is now a
template in internal/web: every gateway serves it at /about, without a key,
linked from the landing page's footer, and the published site is rendered
from the same template.

The fonts are vendored (Latin subsets, OFL) so the page sends nobody's
browser to Google. The proxy configurations it shows are copies of
deploy/proxy, kept equal by a test. A development build says so instead of
linking to release files that do not exist.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-09-29 10:44:45 +02:00
Ilker ManapandClaude Opus 5.5 f989e67e1a Connect an editor from the account page, and serve the Responses API
The account page makes a key for Continue, Codex or Cline and shows
what the tool needs with the key, this gateway's address and a served
model in it: Continue's assistant file, written to its own place under
~/.continue/assistants so the person's config.yaml is not touched;
Codex's provider and profile, added to ~/.codex/config.toml without
changing its default model; and for Cline, whose settings live in VS
Code's secret storage, the four fields to paste. Continue and Codex get
a one-line command whose script, served without a session, holds no
secret - the key comes from the environment, never a URL - keeps a copy
of whatever it replaces, and leaves the rest of the file as it was.

POST /v1/responses translates the Responses API, which Codex speaks, to
chat completions and back: instructions, messages, function tools and
Codex's free-form custom tools, their calls and outputs, and usage, as a
response object or as its stream of typed events. It goes down the same
path as any request - quotas, admission, routing, usage. Server-side
tools and stored responses have no equivalent and are ignored.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-09-29 10:04:17 +02:00
Ilker ManapandClaude Opus 5.5 20479a2d5f Find multi-node NVLink domains, place across them, and show them as racks
A GB200 NVL72 rack or an NVLink Switch system is a scale-up island: its
GPUs reach each other over NVLink whatever node they are in. Each GPU
now reports its fabric from nvidia-smi -q - state, status, ClusterUUID,
CliqueId, bandwidth - at registration and once a minute after, since a
fabric manager restarting drops every card out at once; and each node
reports nvidia-imex and its channels, without which the domain's GPUs
cannot share memory across nodes. A single machine's NVSwitch, which
reports a cluster of zeros, is no domain.

A replica spanning nodes is placed inside one domain first, on nodes
whose every card is joined and whose IMEX is up, and its shards are
told NCCL_MNNVL_ENABLE=1; InfiniBand is the fallback.

/admin/nvlink draws each domain as a rack - the NVLink Switch band on
top, light running down a backplane, trays of GPU chips that glow when
joined, pulse red when not, turn amber on degraded bandwidth - and
lists what keeps a tray out. A node's page shows its GPU-by-GPU NVLink
matrix, each card's NUMA node and nearest adapter, and its domain and
IMEX. doctor checks joining, bandwidth, IMEX and its channel; metrics
llmgw_gpu_nvlink_joined and llmgw_imex_up; /api/v1/nvlink.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-09-29 09:07:59 +02:00