ClusterForge Release v2.5.0

What's Changed

• docs: update 2.2.0 refs to 2.2.2 in migration from 2.1.x guide by @Q-Dub in #785
• EAI-7277: scoped OpenBao write credential for the API-key KV path by @johnl-amd in #784
• Add option to use local Helm values for apps by @Q-Dub in #787
• Minor correction by @rodrodsilo in #788
• EAI-7555_Overhaul_lgtm_stack by @woojae-siloai in #790
• Eai 7068 upgrade seaweedfs by @pwistbac in #796
• EAI-7854 Skip the keycloak CNPG PDB for a single instance by @pre in #798
• EAI-7854 Pin aiwb-infra-cnpg to one instance on small profile by @pre in #801
• EAI-7889 Add pod anti-affinity for SeaweedFS and fix keycloak-old CNPG topology by @silokimmo in #803
• feat: extraApps for argo by @lauri-amd in #791
• EAI-7854 Repair YAML block scalar in keycloak realm ConfigMap by @pre in #804
• Minor fixes for Openshift Radeon clusters by @rodrodsilo in #805
• Improving script and reducing required files by @rodrodsilo in #809
• Bump envoy + envoy-ai-gateway; cut ai-gateway over to native API-key auth by @johnl-amd in #794
• EAI-7650: Set explicit request timeout on SeaweedFS S3 route by @nowycondro in #800
• EAI-5958: Add kubeletstats collector to stamp workload_id on pod CPU/mem metrics by @johnl-amd in #810
• ai-gateway: raise drainTimeout so model-deploy churn doesn't kill live completions by @johnl-amd in #812
• fix: TOCTOU race in openbao secret-manager script by @lauri-amd in #813
• config(seaweedfs): pin to version 4.40 (docker-hub latest) by @Q-Dub in #797
• EAI-8208 Give the JVM the memory limit of its container by @pre in #814
• EAI-8004 remove deprecated AIMs by @brownzebra in #818
• update clusterruntimeconfig-override for aim-engine in root/values.yaml by @oskarasbrink in #819
• Declare the defaulted backendRef weight on ai-gateway-default-deny by @johnl-amd in #817
• Merge main into release/v2.4.x by @brownzebra in #820
• EAI-8232 Stop using PreSync hooks for seaweedfs ExternalSecrets by @silokimmo in #821
• EAI-8232: Merge main to release v2.4.x (SeaweedFS External Secret PreSync removal) by @silokimmo in #822
• Renew AIWB OpenBao token instead of rotating it by @johnl-amd in #816
• [EAI-7305]: Remove cluster-auth from the stack by @johnl-amd in #825
• Openshift custom values feature by @rodrodsilo in #826
• Minor fix by @rodrodsilo in #827
• Issue the ai-gateway webhook cert via cert-manager by @johnl-amd in #824
• feat(extra-apps): Add blueprint for vllm-sr by @lauri-amd in #792
• feat(EAI-8292): webhook heal script and platform gates (complements #824) by @Q-Dub in #823
• Merge main into release/v2.4.x by @lauri-amd in #828
• docs: document inter-release AIM catalog management by @Q-Dub in #811
• sanity check to get script if missing from release by @rodrodsilo in #830
• fix(blueprint): vllm-sr remove cluster-auth references by @lauri-amd in #829
• docs: align AIM catalog lifecycle with in-tree chart templates by @Q-Dub in #831
• Raise ai-gateway body-authz request-body cap to 32MiB (EAI-8489) by @fredespi in #832
• docs(aim-catalog): describe shipped catalog behaviour, not pending work by @Q-Dub in #833
• [EAI-8489]: Raise ai-gateway body-authz request-body cap to 4MiB by @johnl-amd in #834
• fix: aim-artifact-downloader version by @Q-Dub in #835
• EAI-8229 Fix CLUSTER_SIZE small: keep global.domain, hooks, drop gitea by @pre in #815
• feat: envoy-ai-gateway-ratelimit component by @lauri-amd in #837
• docs: rename AIM catalog guides to match their scope by @Q-Dub in #839
• docs: add AIM catalog lifecycle and model management guides by @Q-Dub in #840
• EAI-8713: move OpenShift to otel-lgtm-stack v1.0.8 by @silokimmo in #841
• Use values.yaml app versions as reference by @rodrodsilo in #843
• Release/v2.4.x by @oskarasbrink in #844
• EAI-8713: tune otel-lgtm-stack v1.0.8 — drop unused PVC, filter apiserver metrics, size tiers, GOMEMLIMIT by @silokimmo in #842
• EAI-8865 Add AGENTS.md with repository guidance by @pre in #846
• Adding airgap details by @rodrodsilo in #845
• EAI-7854 Spread CNPG instances across distinct nodes by @pre in #802
• Upgrading default gpu operator the latest chart by @rodrodsilo in #850
• EAI-9149: set large loki back to 50Gi and stop reconciling PVC size (2.5.x) by @silokimmo in #865
• EAI-9140 Trim node labels and restore the metrics-k8s batch cap by @silokimmo in #862

New Contributors

• @nowycondro made their first contribution in #800
• @fredespi made their first contribution in #832

Full Changelog: v2.2.2...v2.5.0