release: bump version to 1.6.0#48
Merged
Merged
Conversation
New direct vLLM integration + scatter-gather datapath since v1.5.2: - vLLM DfkvStoreConnector (KVConnectorBase_V1, GPUDirect RDMA, bypass LMCache) — dingodb#46 - Scatter-gather batch API (batch_put_sg/batch_get_auto_sg, QP max_sge 2->30): one multi-SGE RDMA per chunk, ~20x fewer keys/disk-reads — dingodb#46 - io_uring async GET serve loop (opt-in DFKV_SERVER_URING, default off) — dingodb#46 - 7 fresh-eyes review fixes (per-item SG failure, recv-thread hardening, empty-key skip, io_uring EINTR/short-read, true out_lens) + 2 regression tests — dingodb#46 - Docs: vLLM deploy guide + config reference, README multi-engine, HiCache boundary — dingodb#47 No wire change (kProtoVersion still 1); v1.5.x compatible. CI green incl. TSan + RDMA datapath.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Cuts v1.6.0 — the first release with the direct vLLM connector and the scatter-gather datapath, on top of v1.5.2.
Since v1.5.2
DfkvStoreConnector(feat(vllm): DfkvStoreConnector — direct vLLM KV connector (GPUDirect RDMA, bypass LMCache) [WIP] #46):KVConnectorBase_V1, stores/loads KV directly over GPUDirect RDMA, occupies theMooncakeStoreConnectorslot, bypasses LMCache. Validated on H100 + IB with DeepSeek-V4 (multi kv_cache_group / MLA + SWA), cross-restart + cross-DP prefix hit.batch_put_sg/batch_get_auto_sg, QPmax_sge2→30 — one multi-SGE RDMA per chunk (~20× fewer keys / disk reads). Additive; legacy 2-SGE path unchanged.DFKV_SERVER_URING(default off,-DDFKV_WITH_URING).out_lens.docs/vllm/DEPLOY.md(deploy + config reference + recommended settings), README multi-engine tagline,docs/DEPLOY.md§9 HiCache engine/feature boundary + corrected depth note.Compatibility
No wire change —
kProtoVersionstays1, so v1.6.0 is line-compatible with v1.5.x (mixed deployments not rejected). The HiCache/SGLang and LMCache paths are unchanged.Bumps
VERSION+CMakeLists.txtonly.