惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

罗磊的独立博客
L
LangChain Blog
aimingoo的专栏
aimingoo的专栏
IT之家
IT之家
B
Blog
博客园_首页
博客园 - 司徒正美
有赞技术团队
有赞技术团队
博客园 - 聂微东
I
InfoQ
美团技术团队
GbyAI
GbyAI
阮一峰的网络日志
阮一峰的网络日志
H
Help Net Security
大猫的无限游戏
大猫的无限游戏
MyScale Blog
MyScale Blog
WordPress大学
WordPress大学
The GitHub Blog
The GitHub Blog
A
About on SuperTechFans
人人都是产品经理
人人都是产品经理
Microsoft Azure Blog
Microsoft Azure Blog
Engineering at Meta
Engineering at Meta
钛媒体:引领未来商业与生活新知
钛媒体:引领未来商业与生活新知
The Cloudflare Blog

Recent Commits to openclaw:main

test: merge chat side-result checks · openclaw/openclaw@ddd2c2a test: merge cron history checks · openclaw/openclaw@f7eb746 test: merge responsive navigation shell checks · openclaw/openclaw@c2e4b47 docs(changelog): add codex oauth fixes · openclaw/openclaw@628e6cd test: merge navigation routing cases · openclaw/openclaw@5d8cecb Tests: mock channel registry bundled fallback · openclaw/openclaw@2b08233 Secrets: avoid broad web search discovery for single plugin config · openclaw/openclaw@a464f59 test: merge config view browser checks · openclaw/openclaw@20cf511 fix(status): align oauth health with runtime · openclaw/openclaw@eed7116 feat: add macOS screen snapshots for monitor preview (#67954) thanks … · openclaw/openclaw@f377db1 fix: report shared auth scopes in hello-ok (#67810) thanks @BunsDev · openclaw/openclaw@0b6c39b Auto-reply: avoid eager bundled route fallback · openclaw/openclaw@3ea1bf4 Tests: narrow session binding contract setup · openclaw/openclaw@54e4e16 fix(macOS): enable undo/redo in webchat composer text input (#34962) · openclaw/openclaw@00951dc Tests: speed up channel setup promotion · openclaw/openclaw@82b529a Docs: refresh agent instructions · openclaw/openclaw@5775fe2 fix(auth): serialize OAuth refresh across agents to fix #26322 (#67876) · openclaw/openclaw@8e79080 test: allow ollama public surface boundary test · openclaw/openclaw@7d4f1a6 Docs: add test performance guardrails · openclaw/openclaw@89706d3 Tests: restore context-engine usage proof · openclaw/openclaw@e4c4f95 Tests: slim context engine runtime coverage · openclaw/openclaw@74c198f ci: retry failed custom checkouts · openclaw/openclaw@0ee5baf test: trim duplicate provider auth onboarding cases · openclaw/openclaw@1ffc02e matrix: fix sessions_spawn --thread subagent session spawning (#67643) · openclaw/openclaw@1ce2596 test: reduce auth choice fixture churn · openclaw/openclaw@857b9cd test: mock health status config boundaries · openclaw/openclaw@9d5ab4a test: mock onboard config io boundary · openclaw/openclaw@299694d test: mock legacy state plugin boundaries · openclaw/openclaw@2713089 test: mock channel install boundaries · openclaw/openclaw@b945248 test: mock doctor preview channel boundaries · openclaw/openclaw@b1a3ad4
fix(memory): cap ollama non-batch embedding concurrency ·...
steipete · 2026-04-28 · via Recent Commits to openclaw:main
Original file line numberDiff line numberDiff line change

@@ -20,6 +20,7 @@ Docs: https://docs.openclaw.ai

2020

### Fixes

2121
2222

- Agents/ACPX: stop forwarding Codex ACP timeout config controls that Codex rejects while preserving OpenClaw's run-timeout watchdog for ACP subagents. Fixes #73052. Thanks @pfrederiksen and @richa65.

23+

- Memory/Ollama: add `memorySearch.remote.nonBatchConcurrency` for inline embedding indexing, default Ollama non-batch indexing to one request at a time, and keep batch concurrency separate from non-batch concurrency so local embedding backfills avoid timeout storms on smaller hosts. Carries forward #57733. Thanks @itilys.

2324

- Docs/tools: clarify that `tools.profile: "messaging"` is intentionally narrow and that `tools.profile: "full"` is the unrestricted baseline for broader command/control access. Carries forward #39954. Thanks @posigit.

2425

- Control UI/Agents: redact tool-call args, partial/final results, derived exec output, and configured custom secret patterns before streaming tool events to the Control UI, so tool output cannot expose provider or channel credentials. Fixes #72283. (#72319) Thanks @volcano303 and @BunsDev.

2526

- Agents/sessions: keep `sessions_history` recall redaction enabled even when general log redaction is disabled, and clarify that safety-boundary UI/tool/diagnostic payloads still redact independently of `logging.redactSensitive`. Carries forward #72319. Thanks @volcano303 and @BunsDev.

Original file line numberDiff line numberDiff line change

@@ -1,4 +1,4 @@

1-

5ffabe5ff76d8e4a0d121e89f74f84917b919447e63bf12e0e5b0e4c0211d451 config-baseline.json

2-

7dcb21e47ddd5de98e2af1ecbc41e11ac0c5742819c359e6d851fbc39c0226e9 config-baseline.core.json

1+

0f57fb6d20b9d300c4325b227e49f17f04349b0f3c27dd218397fe7a3b5001dc config-baseline.json

2+

9d1815981dc3f89d1dfdc72f0a4723d4fd5efca8e5b8a1a1cbf6a053c50c937d config-baseline.core.json

33

c4f07c228d4f07e7afafa5b600b4a80f5b26aaed7267c7287a64d04a527be8e8 config-baseline.channel.json

44

6938050627f0d120109d2045b4300aa8b508b35132542db434033ed0fe3e2b3a config-baseline.plugin.json

Original file line numberDiff line numberDiff line change

@@ -885,7 +885,13 @@ For the full setup and behavior details, see [Ollama Web Search](/tools/ollama-s

885885

{

886886

agents: {

887887

defaults: {

888-

memorySearch: { provider: "ollama" },

888+

memorySearch: {

889+

provider: "ollama",

890+

remote: {

891+

// Default for Ollama. Raise on larger hosts if reindexing is too slow.

892+

nonBatchConcurrency: 1,

893+

},

894+

},

889895

},

890896

},

891897

}

@@ -899,10 +905,11 @@ For the full setup and behavior details, see [Ollama Web Search](/tools/ollama-s

899905

defaults: {

900906

memorySearch: {

901907

provider: "ollama",

908+

model: "nomic-embed-text",

902909

remote: {

903910

baseUrl: "http://gpu-box.local:11434",

904-

model: "nomic-embed-text",

905911

apiKey: "ollama-local",

912+

nonBatchConcurrency: 2,

906913

},

907914

},

908915

},

Original file line numberDiff line numberDiff line change

@@ -386,6 +386,7 @@ Prevents re-embedding unchanged text during reindex or transcript updates.

386386
387387

| Key | Type | Default | Description |

388388

| ----------------------------- | --------- | ------- | -------------------------- |

389+

| `remote.nonBatchConcurrency` | `number` | `4` | Parallel inline embeddings |

389390

| `remote.batch.enabled` | `boolean` | `false` | Enable batch embedding API |

390391

| `remote.batch.concurrency` | `number` | `2` | Parallel batch jobs |

391392

| `remote.batch.wait` | `boolean` | `true` | Wait for batch completion |

@@ -394,7 +395,9 @@ Prevents re-embedding unchanged text during reindex or transcript updates.

394395
395396

Available for `openai`, `gemini`, and `voyage`. OpenAI batch is typically fastest and cheapest for large backfills.

396397
397-

This is separate from `sync.embeddingBatchTimeoutSeconds`, which controls inline embedding calls used by local/self-hosted providers and hosted providers when provider batch APIs are not active.

398+

`remote.nonBatchConcurrency` controls inline embedding calls used by local/self-hosted providers and hosted providers when provider batch APIs are not active. Ollama defaults to `1` for non-batch indexing to avoid overwhelming smaller local hosts; set a higher value on larger machines.

399+
400+

This is separate from `sync.embeddingBatchTimeoutSeconds`, which controls the timeout for inline embedding calls.

398401
399402

---

400403
Original file line numberDiff line numberDiff line change

@@ -89,11 +89,17 @@ export function resolveEmbeddingTimeoutMs(params: {

8989
9090

export function resolveMemoryIndexConcurrency(params: {

9191

batch: { enabled: boolean; concurrency: number };

92-

configuredConcurrency?: number;

92+

configuredNonBatchConcurrency?: number;

93+

providerId?: string;

9394

}): number {

94-

return params.configuredConcurrency != null || params.batch.enabled

95-

? params.batch.concurrency

96-

: EMBEDDING_INDEX_CONCURRENCY;

95+

if (params.batch.enabled) {

96+

return params.batch.concurrency;

97+

}

98+

const configured = params.configuredNonBatchConcurrency;

99+

if (typeof configured === "number" && Number.isFinite(configured)) {

100+

return Math.max(1, Math.floor(configured));

101+

}

102+

return params.providerId === "ollama" ? 1 : EMBEDDING_INDEX_CONCURRENCY;

97103

}

98104
99105

export abstract class MemoryManagerEmbeddingOps extends MemoryManagerSyncOps {

@@ -509,7 +515,8 @@ export abstract class MemoryManagerEmbeddingOps extends MemoryManagerSyncOps {

509515

protected getIndexConcurrency(): number {

510516

return resolveMemoryIndexConcurrency({

511517

batch: this.batch,

512-

configuredConcurrency: this.settings.remote?.batch?.concurrency,

518+

configuredNonBatchConcurrency: this.settings.remote?.nonBatchConcurrency,

519+

providerId: this.provider?.id,

513520

});

514521

}

515522
Original file line numberDiff line numberDiff line change

@@ -46,11 +46,35 @@ describe("memory index concurrency resolution", () => {

4646

).toBe(4);

4747

});

4848
49-

it("respects configured concurrency even when batch mode is disabled", () => {

49+

it("respects configured non-batch concurrency when batch mode is disabled", () => {

5050

expect(

5151

resolveMemoryIndexConcurrency({

5252

batch: { enabled: false, concurrency: 1 },

53-

configuredConcurrency: 1,

53+

configuredNonBatchConcurrency: 1,

54+

}),

55+

).toBe(1);

56+

});

57+
58+

it("clamps configured non-batch concurrency to a positive integer", () => {

59+

expect(

60+

resolveMemoryIndexConcurrency({

61+

batch: { enabled: false, concurrency: 2 },

62+

configuredNonBatchConcurrency: 2.8,

63+

}),

64+

).toBe(2);

65+

expect(

66+

resolveMemoryIndexConcurrency({

67+

batch: { enabled: false, concurrency: 2 },

68+

configuredNonBatchConcurrency: 0,

69+

}),

70+

).toBe(1);

71+

});

72+
73+

it("uses conservative non-batch concurrency for Ollama by default", () => {

74+

expect(

75+

resolveMemoryIndexConcurrency({

76+

batch: { enabled: false, concurrency: 2 },

77+

providerId: "ollama",

5478

}),

5579

).toBe(1);

5680

});

Original file line numberDiff line numberDiff line change

@@ -127,11 +127,15 @@ describe("memory search config", () => {

127127

function expectMergedRemoteConfig(

128128

resolved: ReturnType<typeof resolveMemorySearchConfig>,

129129

apiKey: unknown,

130+

extras?: { nonBatchConcurrency?: number },

130131

) {

131132

expect(resolved?.remote).toEqual({

132133

baseUrl: "https://agent.example/v1",

133134

apiKey,

134135

headers: { "X-Default": "on" },

136+

...(typeof extras?.nonBatchConcurrency === "number"

137+

? { nonBatchConcurrency: extras.nonBatchConcurrency }

138+

: {}),

135139

batch: {

136140

enabled: false,

137141

wait: true,

@@ -535,6 +539,18 @@ describe("memory search config", () => {

535539

expectMergedRemoteConfig(resolved, "default-key"); // pragma: allowlist secret

536540

});

537541
542+

it("merges remote non-batch concurrency from defaults with agent overrides", () => {

543+

const cfg = configWithRemoteDefaults({

544+

apiKey: "default-key", // pragma: allowlist secret

545+

headers: { "X-Default": "on" },

546+

nonBatchConcurrency: 1,

547+

});

548+
549+

const resolved = resolveMemorySearchConfig(cfg, "main");

550+
551+

expectMergedRemoteConfig(resolved, "default-key", { nonBatchConcurrency: 1 }); // pragma: allowlist secret

552+

});

553+
538554

it("preserves SecretRef remote apiKey when merging defaults with agent overrides", () => {

539555

const cfg = configWithRemoteDefaults({

540556

apiKey: { source: "env", provider: "default", id: "OPENAI_API_KEY" }, // pragma: allowlist secret

Original file line numberDiff line numberDiff line change

@@ -22,6 +22,7 @@ export type ResolvedMemorySearchConfig = {

2222

baseUrl?: string;

2323

apiKey?: SecretInput;

2424

headers?: Record<string, string>;

25+

nonBatchConcurrency?: number;

2526

batch?: {

2627

enabled: boolean;

2728

wait: boolean;

@@ -165,9 +166,11 @@ function mergeConfig(

165166

overrideRemote?.baseUrl ||

166167

overrideRemote?.apiKey ||

167168

overrideRemote?.headers ||

169+

overrideRemote?.nonBatchConcurrency != null ||

168170

defaultRemote?.baseUrl ||

169171

defaultRemote?.apiKey ||

170-

defaultRemote?.headers,

172+

defaultRemote?.headers ||

173+

defaultRemote?.nonBatchConcurrency != null,

171174

);

172175

const includeRemote =

173176

hasRemoteConfig ||

@@ -191,6 +194,8 @@ function mergeConfig(

191194

baseUrl: overrideRemote?.baseUrl ?? defaultRemote?.baseUrl,

192195

apiKey: overrideRemote?.apiKey ?? defaultRemote?.apiKey,

193196

headers: overrideRemote?.headers ?? defaultRemote?.headers,

197+

nonBatchConcurrency:

198+

overrideRemote?.nonBatchConcurrency ?? defaultRemote?.nonBatchConcurrency,

194199

batch,

195200

}

196201

: undefined;

Original file line numberDiff line numberDiff line change

@@ -4368,6 +4368,14 @@ export const GENERATED_BASE_CONFIG_SCHEMA: BaseConfigSchemaResponse = {

43684368

description:

43694369

"Adds custom HTTP headers to remote embedding requests, merged with provider defaults. Use this for proxy auth and tenant routing headers, and keep values minimal to avoid leaking sensitive metadata.",

43704370

},

4371+

nonBatchConcurrency: {

4372+

type: "integer",

4373+

exclusiveMinimum: 0,

4374+

maximum: 9007199254740991,

4375+

title: "Remote Non-Batch Embedding Concurrency",

4376+

description:

4377+

"Limits concurrent inline embedding requests during non-batch memory indexing. Lower this for local or small self-hosted providers such as Ollama; batch embedding concurrency is configured separately under remote.batch.",

4378+

},

43714379

batch: {

43724380

type: "object",

43734381

properties: {

@@ -6358,6 +6366,11 @@ export const GENERATED_BASE_CONFIG_SCHEMA: BaseConfigSchemaResponse = {

63586366

type: "string",

63596367

},

63606368

},

6369+

nonBatchConcurrency: {

6370+

type: "integer",

6371+

exclusiveMinimum: 0,

6372+

maximum: 9007199254740991,

6373+

},

63616374

batch: {

63626375

type: "object",

63636376

properties: {

@@ -26081,6 +26094,11 @@ export const GENERATED_BASE_CONFIG_SCHEMA: BaseConfigSchemaResponse = {

2608126094

help: "Adds custom HTTP headers to remote embedding requests, merged with provider defaults. Use this for proxy auth and tenant routing headers, and keep values minimal to avoid leaking sensitive metadata.",

2608226095

tags: ["advanced"],

2608326096

},

26097+

"agents.defaults.memorySearch.remote.nonBatchConcurrency": {

26098+

label: "Remote Non-Batch Embedding Concurrency",

26099+

help: "Limits concurrent inline embedding requests during non-batch memory indexing. Lower this for local or small self-hosted providers such as Ollama; batch embedding concurrency is configured separately under remote.batch.",

26100+

tags: ["performance"],

26101+

},

2608426102

"agents.defaults.memorySearch.remote.batch.enabled": {

2608526103

label: "Remote Batch Embedding Enabled",

2608626104

help: "Enables provider batch APIs for embedding jobs when supported (OpenAI/Gemini), improving throughput on larger index runs. Keep this enabled unless debugging provider batch failures or running very small workloads.",

Original file line numberDiff line numberDiff line change

@@ -86,6 +86,7 @@ const TARGET_KEYS = [

8686

"agents.defaults.memorySearch.remote.baseUrl",

8787

"agents.defaults.memorySearch.remote.apiKey",

8888

"agents.defaults.memorySearch.remote.headers",

89+

"agents.defaults.memorySearch.remote.nonBatchConcurrency",

8990

"agents.defaults.memorySearch.remote.batch.enabled",

9091

"agents.defaults.memorySearch.remote.batch.wait",

9192

"agents.defaults.memorySearch.remote.batch.concurrency",