Ilker ManapandClaude Opus 5.5 8df78a4d41 The firewall setting reaches the mirror machine and the store's members ...

The leading control plane tells the mirror over its API (PUT /v1/firewall,
kept as .firewall.json beside the models and applied by the same path unit
control planes use), and keeps the state in the store's cluster_state for
members that run no llmgw, which a root timer reads from the member's own
copy. The mirror also lets in the network of every node and control plane
the leader knows. Tried on the PoC: off, on and a network added, on every
kind of machine.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>

2026-10-09 08:53:17 +02:00

..

0001_init.sql

Run the store on PostgreSQL, chosen by what the address looks like

2026-09-25 16:18:10 +02:00

0002_claims.sql

Run the store on PostgreSQL, chosen by what the address looks like

2026-09-25 16:18:10 +02:00

0003_users.sql

Run the store on PostgreSQL, chosen by what the address looks like

2026-09-25 16:18:10 +02:00

0004_admins.sql

Run the store on PostgreSQL, chosen by what the address looks like

2026-09-25 16:18:10 +02:00

0005_models.sql

Run the store on PostgreSQL, chosen by what the address looks like

2026-09-25 16:18:10 +02:00

0006_threads.sql

Run the store on PostgreSQL, chosen by what the address looks like

2026-09-25 16:18:10 +02:00

0007_memory.sql

Stop the text-search comment naming one language

2026-09-27 08:09:49 +02:00

0008_leadership.sql

Stop a second control plane from placing work behind the first's back

2026-09-25 18:24:06 +02:00

0009_capacity.sql

Keep what a benchmark found out, instead of leaving it in a terminal

2026-09-27 11:51:58 +02:00

0010_memberships.sql

Let a person belong to many places, or to none

2026-09-28 19:03:46 +02:00

0011_cluster_secrets.sql

Give every engine a key only the router holds

2026-09-30 00:38:29 +02:00

0012_usage_diagnostics.sql

Follow one request from its id to the replica that served it

2026-09-30 02:36:45 +02:00

0013_capacity_quotas.sql

Guarantee and limit capacity per part of the tenancy tree

2026-09-30 18:15:38 +02:00

0014_node_engines.sql

Give each GPU node a vLLM build it can run, chosen on the cluster page

2026-10-01 12:39:14 +02:00

0015_usage_rollup.sql

Show each team's first-token and latency percentiles, errors and refusals

2026-10-03 19:50:16 +02:00

0016_control_planes.sql

Add a standby control plane with server add, and draw the screens as windows

2026-10-03 22:29:12 +02:00

0017_revoked_nodes.sql

Revoke a removed node's certificate, and require --api-client with a mirror's mutual TLS

2026-10-04 10:34:35 +02:00

0018_usage_guard.sql

Hook a guard into the router: judge requests before dispatch and filter answers as they are relayed

2026-10-04 12:20:24 +02:00

0019_guard_attachments.sql

Attach guard policies to scopes: store, admin API and ctl guard; run the gate when server.yaml defines policies

2026-10-04 12:34:47 +02:00

0020_audit.sql

Audit log: every change through the admin API and the administration pages, with ctl audit

2026-10-04 16:25:56 +02:00

0021_staff_role.sql

Staff roles (viewer, operator, billing) and the console's Audit view

2026-10-04 16:30:56 +02:00

0022_key_ttl.sql

Key lifecycle: lifetime limits by scope, rotation with overlap, kinds and expiry in the listing

2026-10-04 16:45:05 +02:00

0023_usage_tags.sql

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

2026-10-04 19:45:37 +02:00

0024_batches.sql

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

2026-10-04 21:00:01 +02:00

0025_request_log.sql

Requests kept and replayed: keys requests per project with retention and a share, ctl request --body and --replay

2026-10-04 22:51:21 +02:00

0026_residency.sql

Forgetting a person and keeping requests in the cluster: users forget, keys residency --in-cluster

2026-10-04 23:27:31 +02:00

0027_answer_cache.sql

Reusing answers: a response cache per project, exact match, opt-in with a TTL

2026-10-04 23:50:53 +02:00

0028_eval_runs.sql

Quality suites, and new weights held to one: ctl eval, model update/promote --eval, model gate

2026-10-05 00:04:43 +02:00

0029_federation_peers.sql

Federation: another llmgw cluster as a provider, its models as PEER/MODEL

2026-10-05 01:21:30 +02:00

0031_facility.sql

Facility-aware placement: racks, PDUs, holds and power caps

2026-10-05 01:17:49 +02:00

0032_cluster_settings.sql

The cluster's settings live in the store, and every change takes effect

2026-10-05 15:25:23 +02:00

0033_usage_guard_rule.sql

Guardrails turn off and on; a judge's refusal says which hazard; ctl guard judge ask

2026-10-06 12:41:08 +02:00

0034_password_resets.sql

A forgotten password is set through a one-time link

2026-10-07 13:20:11 +02:00

0035_knowledge.sql

Knowledge bases in the store: collections, grants and chunks kept apart per tenant

2026-10-08 12:47:57 +02:00

0036_kb_defaults.sql

Collections a part of the tree answers from when a request names none

2026-10-08 15:52:43 +02:00

0037_train_jobs.sql

LoRA adapters trained on the nodes: llmgw ctl train

2026-10-08 20:25:57 +02:00

0038_train_owner.sql

Trained adapters published, private to their owners like the knowledge bases they come from

2026-10-08 21:30:50 +02:00

0039_kb_vec_type.sql

Knowledge bases at scale: halfvec, a text index per collection, ef_search raised to the candidates wanted

2026-10-08 22:08:29 +02:00

0040_cluster_state.sql

The firewall setting reaches the mirror machine and the store's members

2026-10-09 08:53:17 +02:00