All pull requests

Pull requests

• [Bugfix][Router] Strip backend Date and X-Request-Id from proxied responses

  #1169·wguo7 opened Oct 20, 2026·3 tasks done·

  ·

• [Bugfix][Router] Avoid blocking the event loop when reading CPU usage in /metrics

  #1168·wguo7 opened Oct 20, 2026·3 tasks done·

  ·
  2

• [Bugfix][Router] Keep routing settings across a dynamic config reload

  #1167·rahulssv-ibm opened Oct 20, 2026·3 tasks done·

  ·
  1

  3

• [Bugfix] Use dashes instead of underscores for --max-loras flag

  #1166·rahulssv-ibm opened Oct 20, 2026·3 tasks done·

  ·
  1

  2

• [Bugfix][Router] Return 503 when failover ends on a backend client error

  #1164·Mazyod opened Oct 20, 2026·3 tasks done·

  ·
  1

• [Bugfix][Router] Resolve backend hostnames with getaddrinfo

  #1162·Mazyod opened Oct 20, 2026·3 tasks done·

  ·
  1

  3

• [Doc] Use the public router image in the secure tutorial

  #1160·git-jxj opened Oct 20, 2026·3 tasks done·

  ·

• [Router] Standardize route_request signature across all routers (rebase of #1023)

  #1159·chrikrah opened Oct 20, 2026Contributor·3 tasks done·

  ·
  1

  2

• [Bugfix][Helm] Preserve zero initial delays for engine probes

  #1158·git-jxj opened Oct 19, 2026·3 tasks done·

  ·

• [Bugfix][Router] Forward every transcription form field to the engine

  #1156·David-Wu1119 opened Oct 17, 2026Contributor·

  ·
  4

• [Bugfix] Fix KEDA default trigger query and examples

  #1155·zjjkee opened Oct 17, 2026Contributor·3 tasks done·

  ·
  1

• [Bugfix][Router] Accept single-label static backend URLs

  #1154·spicysharkco opened Oct 17, 2026·3 tasks done·

  ·
  2

• [Bugfix][Router] Forward the client's Authorization in the P/D routes

  #1153·David-Wu1119 opened Oct 17, 2026Contributor·

  ·
  7

• [Bugfix][Router] Route prefixaware requests that carry no prompt

  #1152·David-Wu1119 opened Oct 17, 2026Contributor·

  ·
  5

• [Bugfix][Router] Count partial prefix chunks at their actual length

  #1151·spicysharkco opened Oct 17, 2026·3 tasks done·

  ·

• [Bugfix][Feat] Fix EKS Terraform tutorial with an existing VPC

  #1150·zjjkee opened Oct 16, 2026Contributor·3 tasks done·

  ·
  18

• [Bugfix][Helm] Fix sharedPvc-only volumeMounts, pdb.maxUnavailable: 0 and lmcache without vllmConfig

  #1149·zjjkee opened Oct 16, 2026Contributor·3 tasks done·

  ·
  5

• [Doc] Drop --task transcription from the Whisper tutorial

  #1148·fjvbn2003 opened Oct 16, 2026·2 of 3 tasks·

  ·

• [Bugfix][Helm] Use vllm:prefix_cache_* in the GPU KV Cache Hit Rate panel

  #1147·fjvbn2003 opened Oct 16, 2026·2 of 3 tasks·

  ·

• [Bugfix][Router] Read current prefix-cache counters

  #1146·s04 opened Oct 15, 2026·3 tasks done·

  ·
  3

• [Bugfix][Router] Resolve model aliases on multipart routes

  #1144·git-jxj opened Oct 15, 2026·3 tasks done·

  ·
  1

  3

• [Bugfix] Match KEDA scalers to enabled model deployments

  #1142·bojiang-li opened Oct 15, 2026·

  ·
  5

• [CI/Build] Declare httpx as a test dependency

  #1138·AbdHajjar opened Oct 13, 2026·3 tasks done·

  ·
  2

• [Router][Feat] Advertise max_model_len for static models in /v1/models (keyed per model, YAML and CLI)

  #1137·sethforprivacy opened Oct 13, 2026·3 tasks done·

  ·
  1

  7

• [Bugfix][Router] Return 400 when the request model is not a string

  #1135·ptimizeroracle opened Oct 12, 2026·3 tasks done·

  ·