Commits · 80955ef463b89cf8e943ab6b44deace394a69306 · OpenDAS / dynamo · GitLab

25 Feb, 2026 16 commits
- perf: Keep embeddings on GPU Embedding Sender in EPD pipeline + minor fixes (#6535) · 80955ef4
  Kris Hung authored Feb 25, 2026
  
  80955ef4
- chore: Update the reference in readme (#6298) · 5734f5c4
  Pavithra Vijayakrishnan authored Feb 25, 2026
```
Signed-off-by: pvijayakrish <pvijayakrish@nvidia.com>
```
  5734f5c4
- docs: embedding cache in vLLM and TRT-LLM (#6555) · d5add7ff
  Qi Wang authored Feb 25, 2026
  
  d5add7ff
- feat(mocker): add KV cache transfer latency for disagg serving (#6504) · 5d958623
  Yongming Ding authored Feb 24, 2026
```
Signed-off-by: Yongming Ding <yongmingd@nvidia.com>
```
  5d958623
- chore: bake sccache into dynamo base (#6532) · 8fbe7228
  Ran Rubin authored Feb 25, 2026
  
  8fbe7228
- ci: Accelerate dynamo frontend image build (#6479) · e5f97343
  Ran Rubin authored Feb 25, 2026
  
  e5f97343
- chore: add top_p=1 recommendation for gpt-oss to avoid token repetition (#5550) · c3bfbd20
  zhongdaor-nv authored Feb 24, 2026
```
Signed-off-by: zhongdaor <zhongdaor@nvidia.com>
```
  c3bfbd20
- fix: vllm launch script errors for disagg and spec decoding (#6562) · 35323da2
  Alec authored Feb 24, 2026
```
Signed-off-by: alec-flowers <aflowers@nvidia.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
```
  35323da2
- feat: MJPEG video streaming via /v1/videos/stream (experimental) (#6487) · 44d43d0c
  John Pohl authored Feb 25, 2026
```
Signed-off-by: Ryan McCormick <mccormick.codes@gmail.com>
Co-authored-by: Ryan McCormick <rmccormick@nvidia.com>
Co-authored-by: ishandhanani <82981111+ishandhanani@users.noreply.github.com>
```
  44d43d0c
- feat: add ZMQ KV event publishing to mocker [DYN-2221] (#6528) · 65dc451d
  Yan Ru Pei authored Feb 24, 2026
```
Signed-off-by: Ru Pei <rupei@nvidia.com>
Signed-off-by: PeaBrane <yanrpei@gmail.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
```
  65dc451d
- fix: add explicit --kv-transfer-config to disagg prefill workers (#6554) · e3dc1793
  Tushar Sharma authored Feb 24, 2026
```
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
```
  e3dc1793
- feat: add --kv-transfer-config NixlConnector to disagg scripts and recipes (#6560) · eac94322
  Alec authored Feb 24, 2026
```
Signed-off-by: alec-flowers <aflowers@nvidia.com>
```
  eac94322
- feat: Log discovery backend type (#6559) · 4bd4c165
  Ryan McCormick authored Feb 24, 2026
  
  4bd4c165
- docs: expand SGLang observability guide with tracing and dashboards (#6556) · 29c46fb7
  ishandhanani authored Feb 24, 2026
  
  29c46fb7
- feat: Support epp's "pods" interface in Dynamo fixes [DEP-424] (#6302) · c916cd42
  atchernych authored Feb 24, 2026
```
Signed-off-by: Anna Tchernych <atchernych@nvidia.com>
```
  c916cd42
- test: introduce multimodal benchmark toolkit (#6330) · 5a4c96db
  Qi Wang authored Feb 24, 2026
  
  5a4c96db
24 Feb, 2026 24 commits
- chore: document all env vars injected by operator (#6548) · fcdf6610
  Julien Mancuso authored Feb 24, 2026
```
Signed-off-by: Julien Mancuso <jmancuso@nvidia.com>
```
  fcdf6610
- chore: use aic release/0.7.0 (#6494) · 4ea80e95
  Jason Zhou authored Feb 24, 2026
  
  4ea80e95
- ci: enable GPU access for rust-checks in dynamo validation (#6545) · 7f9fdb6e
  Anant Sharma authored Feb 24, 2026
```
Signed-off-by: Anant Sharma <anants@nvidia.com>
```
  7f9fdb6e
- fix: restore E/P/D multimodal disagg serving and add Qwen3-VL-30B-A3B support (#6533) · 8e236375
  KrishnanPrash authored Feb 24, 2026
```
Signed-off-by: Krishnan Prashanth <kprashanth@nvidia.com>
Signed-off-by: KrishnanPrash <140860868+KrishnanPrash@users.noreply.github.com>
```
  8e236375
- docs: updated KVBM docs to specify correct configurations for the write-through property (#6506) · 5c64ffc3
  Kyle McGill authored Feb 24, 2026
```
Signed-off-by: Kyle McGill <kmcgill@nvidia.com>
```
  5c64ffc3
- feat: sglang to 0.5.9 + updated docs (#6518) · 6642e23e
  ishandhanani authored Feb 24, 2026
```
Co-authored-by: baihuitian <baihuitian.bht@gmail.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
```
  6642e23e
- feat: router metrics with dynamo_router_* {worker_id=...}. Update docs (#6227) · 1df620b4
  Keiven C authored Feb 24, 2026
```
Signed-off-by: Keiven Chang <keivenchang@users.noreply.github.com>
Co-authored-by: Keiven Chang <keivenchang@users.noreply.github.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
```
  1df620b4
- fix: vLLM Dockerfile need nvlink and crt for kvbm dependencies (#6521) · f8908800
  Keiven C authored Feb 24, 2026
```
Signed-off-by: Keiven Chang <keivenchang@users.noreply.github.com>
Co-authored-by: Keiven Chang <keivenchang@users.noreply.github.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
```
  f8908800
- chore: install vllm-omni in vllm container (#6458) · e94f4647
  Ayush Agarwal authored Feb 24, 2026
```
Signed-off-by: ayushag <ayushag@nvidia.com>
```
  e94f4647
- chore: de-async scheduler read paths and unblock decode output tracking (#6510) · efa89448
  Yan Ru Pei authored Feb 24, 2026
```
Signed-off-by: PeaBrane <yanrpei@gmail.com>
```
  efa89448
- test: add tests for lora aware kv cache events and routing (#6523) · 6fab12be
  Biswa Panda authored Feb 24, 2026
  
  6fab12be
- fix: frontend crash when using TRTLLM runtime image (#6481) · 35128b14
  Raghav Potluri authored Feb 24, 2026
```
Signed-off-by: Raghav Potluri <raghav.potluri21@gmail.com>
```
  35128b14
- chore: split JetStream subscriber into dedicated module and deprecate... · bc00ef38
  Yan Ru Pei authored Feb 24, 2026
```
chore: split JetStream subscriber into dedicated module and deprecate durable_kv_events [DYN-2203] (#6477)
Signed-off-by: PeaBrane <yanrpei@gmail.com>
```
  bc00ef38
- feat: make KV cache events and routing LoRA-aware (#6517) · 75fea787
  Biswa Panda authored Feb 24, 2026
  
  75fea787
- feat: add error_type label to reqs metric for fine-grained error classification (#5568) · 432fae67
  Vladislav Nosivskoy authored Feb 24, 2026
```
Signed-off-by: Vladislav Nosivskoy <vladnosiv@gmail.com>
Signed-off-by: Keiven Chang <keivenchang@users.noreply.github.com>
Co-authored-by: Keiven Chang <keivenchang@users.noreply.github.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
```
  432fae67
- feat(operator): operator config versioning and injected as configmap (#6464) · c0f34b15
  Thomas Montfort authored Feb 24, 2026
  
  c0f34b15
- chore: exclude docs/ Python files from CODEOWNERS review (#6473) · 59c5f6f1
  dagil-nvidia authored Feb 24, 2026
```
Signed-off-by: Dan Gil <dagil@nvidia.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
```
  59c5f6f1
- test: more stringent test for routing decisions (#6531) · 08db2844
  Yan Ru Pei authored Feb 24, 2026
```
Signed-off-by: PeaBrane <yanrpei@gmail.com>
```
  08db2844
- fix: prevent device_blocks double-add in TRT-LLM KVBM connector (#6406) · 6654a57d
  Yuewei Na authored Feb 24, 2026
```
Signed-off-by: Yuewei Na <nv-yna@users.noreply.github.com>
Co-authored-by: Yuewei Na <nv-yna@users.noreply.github.com>
```
  6654a57d
- feat: remove --connector flag for vLLM backend (LLM-90) (#6450) · 6d3b92f0
  Alec authored Feb 24, 2026
```
Signed-off-by: alec-flowers <aflowers@nvidia.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
```
  6d3b92f0
- chore: replace --enforce-disagg with --decode-fallback, default to enforcing disagg (#6515) · c9ff6235
  Yan Ru Pei authored Feb 24, 2026
```
Signed-off-by: PeaBrane <yanrpei@gmail.com>
```
  c9ff6235
- docs: Clarify DGDSA creation for services is disabled by default (#6389) · 0432aad8
  Mohamad authored Feb 24, 2026
```
Signed-off-by: Mohamad <159854973+maljazaery@users.noreply.github.com>
```
  0432aad8
- feat: add Planner schema to DGDR and Profiler input (#6463) · 659122ed
  hhzhang16 authored Feb 24, 2026
```
Signed-off-by: Hannah Zhang <hannahz@nvidia.com>
Co-authored-by: Sertaç Özercan <852750+sozercan@users.noreply.github.com>
```
  659122ed
- ci: switch container-validation-dynamo to runtime target (#6499) · 0f47842a
  Anant Sharma authored Feb 24, 2026
```
Signed-off-by: Anant Sharma <anants@nvidia.com>
```
  0f47842a