Milestone 3 - Launcher-Based Inference with Sleep/Wake (Test Release #6)

What's Changed

• Reword GPU-ful to GPU-bearing to pass typo checker by @MikeSpreitzer in #395
• Consider port when selecting launcher by @waltforme in #396
• Use EnvVars map instead of copying it by @MikeSpreitzer in #401
• Add annotations to instances in launcher by @MikeSpreitzer in #399
• Controll the GPU assignment for e2e test on OpenShift by @waltforme in #403
• Begin to use annotations in VllmConfig by @waltforme in #404

Full Changelog: v0.5.1-alpha.5...v0.5.1-alpha.6

Milestone 3 - Launcher-Based Inference with Sleep/Wake (Test Release #5)

What's Changed

• Null out serverDat.Sleeping when no vLLM instances associated yet by @waltforme in #359
• deps(actions): bump docker/login-action from 3.7.0 to 4.0.0 by @dependabot[bot] in #323
• deps(actions): bump actions/checkout from 4.2.2 to 6.0.2 by @dependabot[bot] in #324
• deps(actions): bump docker/setup-buildx-action from 3.12.0 to 4.0.0 by @dependabot[bot] in #325
• ci: fix actions/checkout version comments and pin by SHA by @MikeSpreitzer in #361
• deps(actions): bump docker/build-push-action from 6.18.0 to 7.0.0 by @dependabot[bot] in #326
• Add deploy_fma.sh and debug workflow for OCP E2E by @diegocastanibm in #357
• Discontinue the usage of LauncherGeneratedBy label by @waltforme in #365
• 🌱 Unify launcher unit testing by @MikeSpreitzer in #368
• Improve launcher logging - Part 2 by @diegocastanibm in #367
• Fix: Add enable-sleep-mode flag to enable sleep mode for vllm server by @aavarghese in #376
• deps(actions): bump actions/setup-go from 6.2.0 to 6.3.0 by @dependabot[bot] in #360
• Include creation parameters inline in launcher instance state replies by @MikeSpreitzer in #369
• 🌱 Hot fix to e2e test on Openshift by @MikeSpreitzer in #382
• Sync unbound launcher-based server-providing pods by @waltforme in #362
• Preserve the final state at the end of the e2e test in kind by @waltforme in #390
• Extract launcher E2E test scenarios into reusable script by @MikeSpreitzer in #386
• Pin ko base image to chainguard/static digest by @MikeSpreitzer in #392
• deps(actions): bump docker/setup-qemu-action from 3.2.0 to 4.0.0 by @dependabot[bot] in #370
• deps(actions): bump actions/cache from 5.0.3 to 5.0.4 by @dependabot[bot] in #371
• deps(actions): bump docker/metadata-action from 5.10.0 to 6.0.0 by @dependabot[bot] in #372
• deps(go): bump the kubernetes group across 1 directory with 3 updates by @dependabot[bot] in #373
• deps: bump code-generator from v0.34.2 to v0.34.6 by @MikeSpreitzer in #393
• Dump logs for every container in e2e test by @waltforme in #394
• Self-annotation on launcher pods to signal hosted instance changes by @waltforme in #391

Full Changelog: v0.5.1-alpha.4...v0.5.1-alpha.5

Milestone 3 - Launcher-Based Inference with Sleep/Wake (Test Release #4)

What's Changed

• Adjust Node viewing ClusterRole in E2E-on-OCP workflow by @MikeSpreitzer in #313
• fix: wait for CRDs to be Established in OpenShift E2E workflow by @MikeSpreitzer in #314
• Test cases for multiple instances sharing one launcher pod by @waltforme in #264
• deps(actions): bump actions/upload-artifact from 6.0.0 to 7.0.0 by @dependabot[bot] in #301
• ✨ Add workflow summary step showing gate decision by @MikeSpreitzer in #316
• Improvements to the launcher-based tests by @waltforme in #319
• GPU assignment for launcher-based server-providing Pods by @waltforme in #317
• [DOCS] More files to align FMA with LLM-d by @diegocastanibm in #283
• [DOCS] Adding governance documents by @diegocastanibm in #278
• 🌱 Remove per-repo gh-aw typo/link/upstream workflows by @clubanderson in #321
• Launcher log improvements by @diegocastanibm in #286
• Install kubernetes python library on launcher dockerfile by @manoelmarques in #328
• 🌱 Bump vllm-openai image to v0.15.1 by @MikeSpreitzer in #329
• Rework GetInferenceServerPort by @waltforme in #330
• Improve launcher's GPU mock by @waltforme in #322
• Fix misplaced envar names in launcher.md by @waltforme in #344
• Configure launcher for ConfigMap-based GPU UUID-to-index translation by @MikeSpreitzer in #341
• Revise type VllmConfig by @waltforme in #346
• Remove the dependency on the gpu-map ConfigMap for M3 code in production by @waltforme in #349
• 🌱 Pin E2E-on-OpenShift test to vllm-d cluster by @MikeSpreitzer in #350
• Fix typos and add typos config by @MikeSpreitzer in #351
• deps(go): bump the kubernetes group with 3 updates by @dependabot[bot] in #300
• ci: Use real requester and launcher in OpenShift e2e by @rubambiza in #343
• ci: Address review feedback on OpenShift e2e by @rubambiza in #354
• ✨ Add cleanup of launcher image by @MikeSpreitzer in #356
• ci: Enable launcher-populator in OpenShift E2E and local tests by @aavarghese in #348

Full Changelog: v0.5.1-alpha.3...v0.5.1-alpha.4

Milestone 3 - Launcher-Based Inference with Sleep/Wake (Test Release #3)

What's Changed

• ci: remove dual-pods finalizers before namespace deletion by @MikeSpreitzer in #288
• ci: use literal Go build version instead of go.mod value by @MikeSpreitzer in #290
• 🐛 Stop the E2E on OpenShift workflow from deleting the CRDs by @MikeSpreitzer in #293
• fix: upgrade vllm CPU build from v0.15.0 to v0.15.1 by @MikeSpreitzer in #294
• ✨ Reorg E2E on OCP workflow to always dump state by @MikeSpreitzer in #295
• deps(actions): bump actions/download-artifact from 6.0.0 to 8.0.0 by @dependabot[bot] in #303
• deps(actions): bump actions/github-script from 7.0.1 to 8.0.0 by @dependabot[bot] in #304
• ci: upgrade docker/login-action to v3.7.0 by @MikeSpreitzer in #305
• ✨ New management for ValidatingAdmissionPolicy[Binding] objects by @MikeSpreitzer in #297
• ci: upgrade actions/cache to v5.0.3 by @MikeSpreitzer in #306
• fix: replace ClusterRoleBinding to view with namespace-scoped pods permission by @MikeSpreitzer in #310

Full Changelog: v0.5.1-alpha.2...v0.5.1-alpha.3

Milestone 3 - Launcher-Based Inference with Sleep/Wake (Test Release #2)

What's Changed

• ✨ Add launcher and test workload E2E tests to OpenShift CI by @clubanderson in #262
• ✨ Add GitHub Agentic Workflows for typo, link, and upstream checks by @clubanderson in #255
• [FEATURE] - Fetch stdout in launcher by @diegocastanibm in #242
• docs: Update FMA docs with ecosystem context, milestone status, and dependencies by @rubambiza in #266
• Refining ValidatingAdmissionPolicy tests to match current implementation of DP controller by @aavarghese in #234
• deps(actions): bump docker/setup-buildx-action from 3.7.1 to 3.12.0 by @dependabot[bot] in #249
• deps(go): bump the kubernetes group with 5 updates by @dependabot[bot] in #253
• Bump docker/metadata-action, actions/setup-go, and ko-build/setup-ko by @MikeSpreitzer in #281
• Bump actions/setup-python from 4 to 6.2.0 by @MikeSpreitzer in #280
• deps(go): bump github.com/spf13/pflag from 1.0.6 to 1.0.10 in the go-dependencies group by @dependabot[bot] in #254
• Combine helm charts by @rubambiza in #263

Full Changelog: v0.5.1-alpha...v0.5.1-alpha.2

Milestone 3 - Launcher-Based Inference with Sleep/Wake (Test Release)

This is a test release for Milestone 3, introducing launcher-based inference server management with sleep/wake capabilities for efficient GPU resource utilization and quick start-up.

Run TAG="v0.5.1-alpha"
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Release v0.5.1-alpha completed successfully!
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Container Images:
• ghcr.io/llm-d-incubation/llm-d-fast-model-actuation/dual-pods-controller:v0.5.1-alpha
• ghcr.io/llm-d-incubation/llm-d-fast-model-actuation/launcher-populator:v0.5.1-alpha
• ghcr.io/llm-d-incubation/llm-d-fast-model-actuation/launcher:v0.5.1-alpha
• ghcr.io/llm-d-incubation/llm-d-fast-model-actuation/requester:v0.5.1-alpha

Helm Charts (version 0.5.1-alpha):
• oci://ghcr.io/llm-d-incubation/llm-d-fast-model-actuation/charts/dual-pods-controller
• oci://ghcr.io/llm-d-incubation/llm-d-fast-model-actuation/charts/launcher-populator

Install with:
helm install dpctlr oci://ghcr.io/llm-d-incubation/llm-d-fast-model-actuation/charts/dual-pods-controller --version 0.5.1-alpha
helm install launcher-populator oci://ghcr.io/llm-d-incubation/llm-d-fast-model-actuation/charts/launcher-populator --version 0.5.1-alpha
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Milestone 1: Dual pods without sleep/wake

Merge pull request #88 from MikeSpreitzer/source-reorg

Controller source reorg