Commit Graph

16441 Commits

Author SHA1 Message Date
屈轩
8d0ee46764 fix(region): vendor update (#25203) 2026-07-22 11:19:05 +08:00
Jian Qiu
07b718eeb2 fix: allow empty ip on link-local nic (#25202)
Co-authored-by: Qiu Jian <qiujian@yunionyun.com>
2026-07-22 11:05:55 +08:00
屈轩
48204b5064 fix(region): vendor update (#25196) 2026-07-21 23:31:44 +08:00
屈轩
29a2eca906 fix(mcp-server): optimized mcp desc (#25193) 2026-07-21 16:19:09 +08:00
屈轩
8e1b9a4d11 fix(mcp-server): support forwarding mcp-server request (#25190) 2026-07-21 13:27:39 +08:00
屈轩
02c65a9d54 fix(mcp-server): optimized mcp server options (#25187) 2026-07-20 20:56:26 +08:00
屈轩
2a9be258f9 fix(mcp-server): mcp use climc struct (#25174) 2026-07-20 17:17:20 +08:00
wanyaoqi
76f035e4e3 feat(region,host,scheduler): isolated device refactor and hami support (#25182) 2026-07-20 14:00:04 +08:00
wanyaoqi
1d2c8733ad fix(host-deployer): bond netplan add vlan subinterface config (#25178) 2026-07-20 12:16:20 +08:00
wanyaoqi
142b0dabf5 fix(region): baremetal boot mode decided by selected host on image support both uefi&bios (#25154) 2026-07-20 12:07:22 +08:00
wanyaoqi
39e42cfa27 fix(host-deployer): windows disk extend separate partition and filesystem (#25138) 2026-07-20 12:05:36 +08:00
wanyaoqi
e15487dd68 fix(host): add nest virtualization options (#25125) 2026-07-20 12:03:26 +08:00
wanyaoqi
d5f113d64c fix(region): disk change driver support change aio_mode and cache_mode (#25095) 2026-07-20 11:39:13 +08:00
屈轩
57c858a6ea fix(mcp-server): query scope params (#25171) 2026-07-17 00:26:47 +08:00
屈轩
a9761ea558 fix(cloudid): batch detach policy (#25163) 2026-07-16 19:08:08 +08:00
屈轩
25ec1683c8 fix(region): support cas hypervisor (#25160) 2026-07-16 19:07:54 +08:00
Zexi Li
669fa8009c fix(aiproxy): emit done events before response.completed in stream converter (#25158)
Emit output_item.done, reasoning_summary_part.done, function_call_arguments.done,
and related finalize events for text, reasoning, and tool calls before the
response.completed event to match the Responses API stream contract.
2026-07-16 17:02:18 +08:00
Zexi Li
49c5cfa421 feat(aiproxy): auto-provision placeholder ai_key for upstream resolve (#25156) 2026-07-15 19:24:57 +08:00
cwz_eikoh
ed4a8dc049 feat(llm): add llm-router-agent (#25128) 2026-07-15 15:31:08 +08:00
cwz_eikoh
20620bb5a1 feat(llm): support llm-bench in llm (#25103) 2026-07-15 15:00:37 +08:00
Zexi Li
068cc8bdbd Automated cherry pick of #25145: feat(aiproxy): support streaming responses for visual messages (#25146)
* feat(aiproxy): support streaming responses for visual messages

* refactor(aiproxy): share API log helpers and support input/output token usage

Extract common chatlog record lifecycle across proxy handlers, and fall back
to Responses/Anthropic usage field aliases when filling token counts.
2026-07-15 10:57:49 +08:00
Zexi Li
cdc8ab78b5 feat(aiproxy): move visual provider to model columns and support streaming responses (#25144)
Promote visual_provider_id/visual_model_key to dedicated ai_model columns,
force non-stream upstream orchestration for visual paths, and synthesize SSE
chunks for streaming Responses requests that include images.
2026-07-14 18:39:30 +08:00
Zexi Li
73d7cbbb1f Automated cherry pick of #25132: refactor(aiproxy): rename chat log config to API log with S3 fields (#25141)
* refactor(aiproxy): rename chat log config to API log with S3 fields

* feat(aiproxy): add visual extension for Responses image tools

Wire visual model catalog, orchestration, and OpenAI Responses compat so chat/Codex can invoke image generation and editing tools.
2026-07-14 11:52:58 +08:00
wanyaoqi
5f1b76755b fix(region): reset guest quorum child index on guest shutdown (#25133) 2026-07-14 00:10:38 +08:00
屈轩
6a22bb4d2d fix(region): split sync and probe worker (#25100) 2026-07-13 14:00:02 +08:00
Jian Qiu
35b100b03c fix: allow security group cidr have multiple prefixes (#25122)
Co-authored-by: Qiu Jian <qiujian@yunionyun.com>
2026-07-13 13:58:46 +08:00
Zexi Li
40b3c82ab4 feat(aiproxy): add Z.AI / Zhipu provider with dual OpenAI and Anthropic API modes (#25131)
Register zhipu as a public OpenAI-compat provider with GLM catalog seeds,
Anthropic passthrough routing, and v4-style base URL path handling.
2026-07-13 11:25:22 +08:00
Zexi Li
afde072512 feat(aiproxy): add OpenAI Responses API with provider adapters and Codex support (#25127)
Implement /v1/responses endpoint with multi-provider translation (OpenAI, Azure,
Anthropic, Gemini), streaming compat layers, Codex config/tool integration, and
updated functional test docs.
2026-07-10 16:40:56 +08:00
Jian Qiu
05a55ed36b fix: allow deleting idp without checking project count (#25108)
Co-authored-by: Qiu Jian <qiujian@yunionyun.com>
2026-07-10 08:35:55 +08:00
屈轩
598381f9a5 fix(region): gcp network tags (#25119) 2026-07-09 17:11:21 +08:00
屈轩
df94e8049b fix(yunionconf): batch import tags (#25116) 2026-07-09 16:43:01 +08:00
Jian Qiu
7651726458 fix: host order by sql redundant group by (#25114)
Co-authored-by: Qiu Jian <qiujian@yunionyun.com>
2026-07-09 14:22:44 +08:00
cwz_eikoh
52d2ddcb45 feat(aiproxy): support log record in aiproxy through minio (#25078) 2026-07-09 13:01:21 +08:00
Jian Qiu
54b7c42d94 fix: lb listener rule specific certificates (#25106)
Co-authored-by: Qiu Jian <qiujian@yunionyun.com>
2026-07-08 12:03:53 +08:00
Zexi Li
167682c39b feat(aiproxy): add Anthropic Messages API and migrate functional tests to Go (#25107)
Add /v1/messages handler with Anthropic-to-OpenAI translation, upstream
failover, and probe endpoints. Replace shell-based functional test scripts
with pkg/aiproxy/ft and climc test commands; consolidate documentation.
2026-07-07 20:13:14 +08:00
屈轩
e6c9393955 fix(region): pve host driver compatibility (#25098) 2026-07-03 15:37:40 +08:00
Zexi Li
9cc4d76a81 feat(llm): add ModelScope instant model import and refactor model download (#25092)
Add ModelScope hub integration, download driver, and instant model import
flow parallel to HuggingFace. Extract shared HuggingFace download helpers
and reuse them across vLLM/SGLang container drivers.
2026-07-01 20:13:17 +08:00
Zexi Li
3431c10196 feat(llm): support local_path scheduling with prefer_hosts and deployment restart (#25090) 2026-07-01 13:11:19 +08:00
wanyaoqi
31e342881e fix(host-deployer): mount procfs before deploy guestfs (#25087) 2026-07-01 00:53:59 +08:00
Zexi Li
0f63315fc8 feat(scheduler): append GPU reserved resource hints on predicate failure (#25084) 2026-06-30 15:40:50 +08:00
Zexi Li
d9cf601841 feat(llm): backfill orphan LLM instances into deployments on init (#25081)
On service startup, find LLM instances without llm_deployment_id and
create a matching SLLMDeployment, then link the instance. Idempotent
across restarts; skipped on slave nodes.
2026-06-29 11:28:54 +08:00
wanyaoqi
1b742b49e9 fix(host-deployer): do base64 windows password (#25071) 2026-06-28 08:24:48 +08:00
屈轩
7802f909e0 fix(region): quota check (#25066) 2026-06-26 13:06:00 +08:00
wanyaoqi
75cc892bf7 fix(region,host): kvm guest change disk driver reset uefi vars (#25069) 2026-06-26 10:20:37 +08:00
wanyaoqi
be57fc2845 fix(host): qemu set multifd before start incoming migration (#25065)
use drive mirror do block migration
2026-06-26 10:18:58 +08:00
cwz_eikoh
b48c1be667 feat(llm): tool call defaults & auto gpu ultilization (#25074)
* feat(llm): upgrade auto-gpu-ultilization

* feat(llm): add vllm tool call defaults

* feat(llm): add sglang tool call defaults
2026-06-26 10:15:19 +08:00
Zexi Li
8f1d740b54 feat(llm): async instant model delete and improve import status tracking (#25073)
fix(aiproxy): set OpenAI handler process timeouts and skip upstream error
response when request context is already cancelled.
2026-06-25 19:52:46 +08:00
Zexi Li
ba39e80fd8 feat(llm): validate GPU devices on deployment create and improve error messages (#25063) 2026-06-24 14:27:25 +08:00
wanyaoqi
143ca823c8 fix(region): snapshot validate add check guest is quorum master guest (#25051) 2026-06-24 10:30:00 +08:00
cwz_eikoh
d4dd65cb24 Feature/http get+llm probe (#25016)
* hostman: probe http get in pod netns

* llm: watch pod status for service readiness
2026-06-23 19:40:30 +08:00