LLMKube

v0.9.11 Feature

This release adds 3 notable features for engineering teams evaluating rollout.

Published 2d Containers & Orchestration

✓ No known CVEs patched

✓ No known CVEs patched in this version

Topics

ai apple-silicon autoscaling edge-computing gguf gpu

+12 more

self-hosted inference kubernetes llama-cpp llm local-llm metal mlx multi-gpu nvidia tgi vllm

Summary

AI summary

Updates 0.9.11, Bug Fixes, and 2026-07-24 across a mixed release.

Full changelog

cache prep subcommand with direct chown/chmod syscalls (#890 item 3) (#1258) (bde1a87)
charts/foreman: expose the agent --accelerator flag (#1242) (#1261) (a39fd71)
declarative bindAddress on InferenceService (#1240) (#1247) (bf01598)
federation: FederatedCluster registry + fleet status rollup (#1234, #1235) (#1241) (1b5bf08)
foreman: tiered str_replace matching (exact, trailing-ws, uniform-indent) (#1231) (0372e24)
grafana: chart the InferenceService metrics with no panel (#1245) (a5ce923)
Kueue prerequisites: spec.suspend, GPUQuota deferral, pkg/apiutil GPU mapping (#1254) (c95e06f)
metrics: publish operator state metrics from observed state (#1229) (5db0915)
prefetch Model artifacts into the shared cache without an InferenceService (#1218) (7c1161b)

chart: make PrometheusRule select the chart's own scrape targets (#1239) (ffc697a)
delete GPUQuota metric series when the CR is deleted (#1230) (#1246) (1f58c3c)
foreman: edits disarm a nudged RepeatedToolCall hash (#1215) (#1216) (60df2ad)
gate bite check reverts to the upstream merge-base, never the fork tip (#1260) (8a6928f)
grafana: delete dashboard panels for metrics nothing emits (#1244) (9c586d9)
metrics: report GPU queue depth per namespace (#1243) (5779c05)
recycle inference pods without invalid deadlines (#1232) (0b716d7)
waitForIdle proceeds past crashlooping pods; truthful deferral reason on mixed state (#1250) (#1262) (88c87f9)

envtest test-craft guide (status subresource, probe stubs, asset hygiene) (#1263) (#1264) (ea4eee6)

Weekly OSS security release digest.

The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.

No spam, unsubscribe anytime.

Share this release

Track LLMKube

Get notified when new releases ship.

About LLMKube

Kubernetes operator for llama.cpp-native LLM inference with GPU scheduling, Apple Silicon Metal support, and OpenAI-compatible API.

v0.8.1 foreman: requestTimeoutSeconds now sets loop-wide budget, default changes from 600 to 3600.