GoogleCloudPlatform/ai-on-gke

Issues

• [Ray TPU Webhook] Pod Informer inconsistency for large RayCluster sizes

  #858 opened 2 months ago by ryanaoleary
  1

• Include Ray cluster details in outputs.tf

  #888 opened 2 months ago by andrewsykim
  0

• Terraform in benchmarking has impossible provider requirements

  #834 opened 2 months ago by christopher-hendrich-sada
  4

• [benchmarking] profile generator metrics scraping improved resilience

  #884 opened 2 months ago by annapendleton
  0

• Ray on GKE depends on unavailable sourcerepo.googleapis.com

  #743 opened 6 months ago by rti
  2

• Dependency on sourcerepo failing the terraform script

  #878 opened 2 months ago by msathe-tech
  0

• tools/gke-disk-image-builder: Fails with Message: Quota 'CPUS_ALL_REGIONS' exceeded.

  #849 opened 3 months ago by katilp
  1

• Remove 60s sleep waiting for GMP

  #502 opened 3 months ago by andrewsykim
  1

• Terraform apply not working when IAP_DOMAIN is empty

  #699 opened 4 months ago by laurentgrangeau
  1

• RAG frontend is getting OOMKilled

  #518 opened 4 months ago by andrewsykim
  1

• JupyterHub default profile options should only be "CPU" and "L4 GPU"

  #433 opened 4 months ago by andrewsykim
  1

• Update the README to reflect the correct sample dataset & surface this info earlier in the README

  #438 opened 4 months ago by kumar-dhanagopal
  2

• RAG application leaks DB connection objects

  #725 opened 6 months ago by bhlim
  1

• Only set CLOUDSQL_INSTANCE_CONNECTION_NAME env var in Ray clusters when deploying RAG

  #424 opened 4 months ago by andrewsykim
  1

• CLOUDSQL_INSTANCE_CONNECTION_NAME env in RayCluster should only be configured for RAG deployments

  #435 opened 4 months ago by andrewsykim
  0

• Auto enable APIs in the infrastructure module

  #474 opened 4 months ago by andrewsykim
  0

• [benchmarks] enable TGI to use Hugging Face DLC

  #821 opened 4 months ago by annapendleton
  0

• create gke cluster CI is broken

  #820 opened 4 months ago by arueth
  0

• benchmark locust tool feature request: update locust requests to match LPG requests

  #818 opened 4 months ago by annapendleton
  0

• Add me-central2 to the list of supported locations.

  #802 opened 4 months ago by alaliaa
  0

• Logs from CI should be visible to all contributors

  #786 opened 4 months ago by roberthbailey
  0

• Failed Build: Solution Deployment RAG on GKE

  #791 opened 5 months ago by HGPai
  0

• Add flag to customize metric scrape interval.

  #783 opened 5 months ago by raywainman
  0

• Error: "POST /generate HTTP/1.1" 404 Not Found when running Locust tool against vLLM model server

  #777 opened 5 months ago by Edwinhr716
  0

• ai-on-gke benchmark locust load inference feature request: implement multi-threading on locust worker processes

  #768 opened 5 months ago by annapendleton
  0

• ai-on-gke benchmark locust tool feature request: run locust worker and master on separate CPU nodes

  #767 opened 5 months ago by annapendleton
  0

• ai-on-gke benchmark locust load inferencer hits 90%+ cpu usage with master at 200+ users

  #766 opened 5 months ago by annapendleton
  0

• ERROR: Clonning repo

  #760 opened 5 months ago by german-grandas
  0

• RAG tf apply fail on AP cluster due to AP not scale up fast enough to deploy GMP

  #750 opened 6 months ago by yiyinglovecoding
  0

• Service Management API has not been used in project when creating playground

  #700 opened 7 months ago by laurentgrangeau
  0

• Kueue/DWS in helm chart

  #688 opened 8 months ago by krisero
  3

• Enable Image Streaming by default

  #509 opened 7 months ago by andrewsykim
  2

• RAG Application - release-1.1 - Failing running Terraform

  #649 opened 8 months ago by vmasilva
  1

• Configure TPU Provisioner tests to run in CI

  #659 opened 8 months ago by danielvegamyhre
  0

• TPU provisioner should be configurable to stop new nodepool create

  #661 opened 8 months ago by kyle-google
  0

• Switch all quick start deployments to use Autopilot by default

  #573 opened 8 months ago by andrewsykim
  2

• oauth2/google: invalid token JSON from metadata: EOF

  #634 opened 9 months ago by kzos
  4

• JupyterHub fails to start with proxy 599 timeout

  #565 opened 8 months ago by artemvmin
  0

• Error 400: Autopilot clusters must be regional clusters

  #628 opened 9 months ago by kzos
  2

• CI flakes sometimes when waiting to pull Ray images

  #520 opened 9 months ago by andrewsykim
  1

• RAG deployment sometimes need a 3rd g2-standard-24 instance

  #572 opened 9 months ago by andrewsykim
  0

• Resolve SQLAlchemy warning from RAG notebook

  #555 opened 9 months ago by andrewsykim
  0

• GCSFuse permission issues when deploying RAG on existing cluster

  #488 opened 9 months ago by andrewsykim
  1

• VPC networks cannot be deleted for RAG deployments if a private connection to CloudSQL is established

  #550 opened 9 months ago by andrewsykim
  0

• warning: Secret value is no longer used with k8s 1.24+

  #490 opened 10 months ago by soorena776
  1

• RAG cannot be deployed on an existing cluster due to CloudSQL requiring at least 1 private service connection

  #529 opened 9 months ago by andrewsykim
  0

• Use Ray interactive client in the example notebooks

  #425 opened 9 months ago by andrewsykim
  0

• JupyterHub service account is missing GCS related roles

  #428 opened 10 months ago by andrewsykim
  1

• Autopilot e2e tests are flaky due to GMP webook

  #402 opened 10 months ago by andrewsykim
  0

• Jupyter authenticator: Path requires update

  #415 opened 10 months ago by hsachdevah
  0