Mintplex-Labs/anything-llm
1
0
Fork 0
mirror of https://github.com/Mintplex-Labs/anything-llm.git synced 2026-07-19 14:13:35 -04:00
Code Issues 328 Packages Projects Releases 10 Wiki Activity

[GH-ISSUE #2159] [BUG]: Crash when try embed a document using milvus #1407

New Issue
Closed
opened 2026-02-22 18:24:38 -05:00 by yindo · 7 comments
yindo commented 2026-02-22 18:24:38 -05:00
Owner
Copy Link
Copy Source

Originally created by @Milor123 on GitHub (Aug 22, 2024).
Original GitHub issue: https://github.com/Mintplex-Labs/anything-llm/issues/2159

Originally assigned to: @shatfield4 on GitHub.

How are you running AnythingLLM?

Docker (local)

What happened?

I've tried search in the docs but nothing is here https://docs.useanything.com/setup/vector-database-configuration/local/milvus

I've used a docker image of milvus, and it allow me to access web panel from my pc, but anythingllm have a error when try embed a document.

My milvus is https://github.com/milvus-io/milvus/releases/tag/v2.4.9

  • milvus-standalone-docker-compose-gpu.yml

And I've use it with podman-compose.

Milvus console output:

[minio]      | Status:         1 Online, 0 Offline. 
[minio]      | API: http://10.89.0.6:9000  http://127.0.0.1:9000     
[minio]      | Console: http://10.89.0.6:9001 http://127.0.0.1:9001   
[minio]      |

Error from podman logs:

[collector] info: Collector hot directory and tmp storage wiped!
[collector] info: Document processor app listening on port 8888
Environment variables loaded from .env
Prisma schema loaded from prisma/schema.prisma

✔ Generated Prisma Client (v5.3.1) to ./node_modules/@prisma/client in 128ms

Start using Prisma Client in Node.js (See: https://pris.ly/d/client)

import { PrismaClient } from '@prisma/client'
const prisma = new PrismaClient()

or start using Prisma Client at the edge (See: https://pris.ly/d/accelerate)

import { PrismaClient } from '@prisma/client/edge'
const prisma = new PrismaClient()


See other ways of importing Prisma Client: http://pris.ly/d/importing-client

Environment variables loaded from .env
Prisma schema loaded from prisma/schema.prisma
Datasource "db": SQLite database "anythingllm.db" at "file:../storage/anythingllm.db"

21 migrations found in prisma/migrations


No pending migrations to apply.
[backend] info: [EncryptionManager] Loaded existing key & salt for encrypting arbitrary data.
[backend] info: [TELEMETRY ENABLED] Anonymous Telemetry enabled. Telemetry helps Mintplex Labs Inc improve AnythingLLM.
[backend] info: prisma:info Starting a sqlite pool with 29 connections.
[backend] info: [TELEMETRY SENT] {"event":"server_boot","distinctId":"7eecf98b-ccfd-40c5-a83b-5885cdac43b6","properties":{"commit":"--","runtime":"docker"}}
[backend] info: [CommunicationKey] RSA key pair generated for signed payloads within AnythingLLM services.
[backend] info: [EncryptionManager] Loaded existing key & salt for encrypting arbitrary data.
[backend] info: Primary server in HTTP mode listening on port 3001
[backend] info: [BackgroundWorkerService] Feature is not enabled and will not be started.
[backend] info: [MetaGenerator] fetching custome meta tag settings...
[backend] error: LMStudio:listModels Connection error.
[backend] info: [Event Logged] - update_vector_db
[backend] error: LMStudio:listModels Connection error.
[backend] info: Adding new vectorized document into namespace normal-es
[backend] info: [RecursiveSplitter] Will split with {"chunkSize":128,"chunkOverlap":20}
[backend] info: Chunks created from document: 317
[backend] info: [LMStudioEmbedder] fetch failed
[backend] error: addDocumentToNamespace LMStudio service could not be reached. Is LMStudio running?
[backend] error: Failed to vectorize TopfoodGuntry.txt
[backend] info: Adding new vectorized document into namespace normal-es
[backend] info: [RecursiveSplitter] Will split with {"chunkSize":128,"chunkOverlap":20}
[backend] info: Chunks created from document: 725
[backend] info: [LMStudioEmbedder] fetch failed
[backend] error: addDocumentToNamespace LMStudio service could not be reached. Is LMStudio running?
[backend] error: Failed to vectorize interpretacion_suenos_V2.md
[backend] info: [TELEMETRY SENT] {"event":"documents_embedded_in_workspace","distinctId":"7eecf98b-ccfd-40c5-a83b-5885cdac43b6","properties":{"LLMSelection":"lmstudio","Embedder":"lmstudio","VectorDbSelection":"milvus","TTSSelection":"native","runtime":"docker"}}
[backend] info: [Event Logged] - workspace_documents_added
[backend] info: Adding new vectorized document into namespace normal-es
[backend] info: [RecursiveSplitter] Will split with {"chunkSize":128,"chunkOverlap":20}
[backend] info: Chunks created from document: 386
[backend] info: [LMStudioEmbedder] Embedding 386 chunks of text with jina/jina-embeddings-v2-base-es/jina-embeddings-v2-base-es.gguf.
[backend] error: addDocumentToNamespace 13 INTERNAL: Received RST_STREAM with code 2 triggered by internal client error: Protocol error
[backend] error: Failed to vectorize Libretranslate.md
node:internal/process/promises:288
            triggerUncaughtException(err, true /* fromPromise */);
            ^

Error: 13 INTERNAL: Received RST_STREAM with code 2 triggered by internal client error: Protocol error
    at callErrorFromStatus (/app/server/node_modules/@grpc/grpc-js/build/src/call.js:31:19)
    at Object.onReceiveStatus (/app/server/node_modules/@grpc/grpc-js/build/src/client.js:193:76)
    at /app/server/node_modules/@grpc/grpc-js/build/src/call-interface.js:78:35
    at Object.onReceiveStatus (/app/server/node_modules/@zilliz/milvus2-sdk-node/dist/milvus/utils/Grpc.js:146:25)
    at Object.onReceiveStatus (/app/server/node_modules/@grpc/grpc-js/build/src/client-interceptors.js:360:141)
    at Object.onReceiveStatus (/app/server/node_modules/@grpc/grpc-js/build/src/client-interceptors.js:323:181)
    at /app/server/node_modules/@grpc/grpc-js/build/src/resolving-call.js:129:78
    at process.processTicksAndRejections (node:internal/process/task_queues:77:11)
for call at
    at ServiceClientImpl.makeUnaryRequest (/app/server/node_modules/@grpc/grpc-js/build/src/client.js:161:32)
    at ServiceClientImpl.<anonymous> (/app/server/node_modules/@grpc/grpc-js/build/src/make-client.js:105:19)
    at /app/server/node_modules/@zilliz/milvus2-sdk-node/dist/milvus/utils/Function.js:32:31
    at new Promise (<anonymous>)
    at /app/server/node_modules/@zilliz/milvus2-sdk-node/dist/milvus/utils/Function.js:29:16
    at Generator.next (<anonymous>)
    at fulfilled (/app/server/node_modules/@zilliz/milvus2-sdk-node/dist/milvus/utils/Function.js:5:58)
    at process.processTicksAndRejections (node:internal/process/task_queues:95:5) {
  code: 13,
  details: 'Received RST_STREAM with code 2 triggered by internal client error: Protocol error',
  metadata: Metadata { internalRepr: Map(0) {}, options: {} }
}

Node.js v18.20.4

Thank u very much guys!

Are there known steps to reproduce?

Upload a file, nice
Then search the file and move to the space, OK
but when i try "Save and Embed" the docker image crash

Originally created by @Milor123 on GitHub (Aug 22, 2024). Original GitHub issue: https://github.com/Mintplex-Labs/anything-llm/issues/2159 Originally assigned to: @shatfield4 on GitHub. ### How are you running AnythingLLM? Docker (local) ### What happened? I've tried search in the docs but nothing is here https://docs.useanything.com/setup/vector-database-configuration/local/milvus I've used a docker image of milvus, and it allow me to access web panel from my pc, but anythingllm have a error when try embed a document. My milvus is https://github.com/milvus-io/milvus/releases/tag/v2.4.9 - milvus-standalone-docker-compose-gpu.yml And I've use it with podman-compose. Milvus console output: ``` [minio] | Status: 1 Online, 0 Offline. [minio] | API: http://10.89.0.6:9000 http://127.0.0.1:9000 [minio] | Console: http://10.89.0.6:9001 http://127.0.0.1:9001 [minio] | ``` Error from podman logs: ``` [collector] info: Collector hot directory and tmp storage wiped! [collector] info: Document processor app listening on port 8888 Environment variables loaded from .env Prisma schema loaded from prisma/schema.prisma ✔ Generated Prisma Client (v5.3.1) to ./node_modules/@prisma/client in 128ms Start using Prisma Client in Node.js (See: https://pris.ly/d/client) import { PrismaClient } from '@prisma/client' const prisma = new PrismaClient() or start using Prisma Client at the edge (See: https://pris.ly/d/accelerate) import { PrismaClient } from '@prisma/client/edge' const prisma = new PrismaClient() See other ways of importing Prisma Client: http://pris.ly/d/importing-client Environment variables loaded from .env Prisma schema loaded from prisma/schema.prisma Datasource "db": SQLite database "anythingllm.db" at "file:../storage/anythingllm.db" 21 migrations found in prisma/migrations No pending migrations to apply. [backend] info: [EncryptionManager] Loaded existing key & salt for encrypting arbitrary data. [backend] info: [TELEMETRY ENABLED] Anonymous Telemetry enabled. Telemetry helps Mintplex Labs Inc improve AnythingLLM. [backend] info: prisma:info Starting a sqlite pool with 29 connections. [backend] info: [TELEMETRY SENT] {"event":"server_boot","distinctId":"7eecf98b-ccfd-40c5-a83b-5885cdac43b6","properties":{"commit":"--","runtime":"docker"}} [backend] info: [CommunicationKey] RSA key pair generated for signed payloads within AnythingLLM services. [backend] info: [EncryptionManager] Loaded existing key & salt for encrypting arbitrary data. [backend] info: Primary server in HTTP mode listening on port 3001 [backend] info: [BackgroundWorkerService] Feature is not enabled and will not be started. [backend] info: [MetaGenerator] fetching custome meta tag settings... [backend] error: LMStudio:listModels Connection error. [backend] info: [Event Logged] - update_vector_db [backend] error: LMStudio:listModels Connection error. [backend] info: Adding new vectorized document into namespace normal-es [backend] info: [RecursiveSplitter] Will split with {"chunkSize":128,"chunkOverlap":20} [backend] info: Chunks created from document: 317 [backend] info: [LMStudioEmbedder] fetch failed [backend] error: addDocumentToNamespace LMStudio service could not be reached. Is LMStudio running? [backend] error: Failed to vectorize TopfoodGuntry.txt [backend] info: Adding new vectorized document into namespace normal-es [backend] info: [RecursiveSplitter] Will split with {"chunkSize":128,"chunkOverlap":20} [backend] info: Chunks created from document: 725 [backend] info: [LMStudioEmbedder] fetch failed [backend] error: addDocumentToNamespace LMStudio service could not be reached. Is LMStudio running? [backend] error: Failed to vectorize interpretacion_suenos_V2.md [backend] info: [TELEMETRY SENT] {"event":"documents_embedded_in_workspace","distinctId":"7eecf98b-ccfd-40c5-a83b-5885cdac43b6","properties":{"LLMSelection":"lmstudio","Embedder":"lmstudio","VectorDbSelection":"milvus","TTSSelection":"native","runtime":"docker"}} [backend] info: [Event Logged] - workspace_documents_added [backend] info: Adding new vectorized document into namespace normal-es [backend] info: [RecursiveSplitter] Will split with {"chunkSize":128,"chunkOverlap":20} [backend] info: Chunks created from document: 386 [backend] info: [LMStudioEmbedder] Embedding 386 chunks of text with jina/jina-embeddings-v2-base-es/jina-embeddings-v2-base-es.gguf. [backend] error: addDocumentToNamespace 13 INTERNAL: Received RST_STREAM with code 2 triggered by internal client error: Protocol error [backend] error: Failed to vectorize Libretranslate.md node:internal/process/promises:288 triggerUncaughtException(err, true /* fromPromise */); ^ Error: 13 INTERNAL: Received RST_STREAM with code 2 triggered by internal client error: Protocol error at callErrorFromStatus (/app/server/node_modules/@grpc/grpc-js/build/src/call.js:31:19) at Object.onReceiveStatus (/app/server/node_modules/@grpc/grpc-js/build/src/client.js:193:76) at /app/server/node_modules/@grpc/grpc-js/build/src/call-interface.js:78:35 at Object.onReceiveStatus (/app/server/node_modules/@zilliz/milvus2-sdk-node/dist/milvus/utils/Grpc.js:146:25) at Object.onReceiveStatus (/app/server/node_modules/@grpc/grpc-js/build/src/client-interceptors.js:360:141) at Object.onReceiveStatus (/app/server/node_modules/@grpc/grpc-js/build/src/client-interceptors.js:323:181) at /app/server/node_modules/@grpc/grpc-js/build/src/resolving-call.js:129:78 at process.processTicksAndRejections (node:internal/process/task_queues:77:11) for call at at ServiceClientImpl.makeUnaryRequest (/app/server/node_modules/@grpc/grpc-js/build/src/client.js:161:32) at ServiceClientImpl.<anonymous> (/app/server/node_modules/@grpc/grpc-js/build/src/make-client.js:105:19) at /app/server/node_modules/@zilliz/milvus2-sdk-node/dist/milvus/utils/Function.js:32:31 at new Promise (<anonymous>) at /app/server/node_modules/@zilliz/milvus2-sdk-node/dist/milvus/utils/Function.js:29:16 at Generator.next (<anonymous>) at fulfilled (/app/server/node_modules/@zilliz/milvus2-sdk-node/dist/milvus/utils/Function.js:5:58) at process.processTicksAndRejections (node:internal/process/task_queues:95:5) { code: 13, details: 'Received RST_STREAM with code 2 triggered by internal client error: Protocol error', metadata: Metadata { internalRepr: Map(0) {}, options: {} } } Node.js v18.20.4 ``` Thank u very much guys! ### Are there known steps to reproduce? Upload a file, nice Then search the file and move to the space, OK but when i try "Save and Embed" the docker image crash
yindo added the possible bug label 2026-02-22 18:24:38 -05:00
yindo closed this issue 2026-02-22 18:24:38 -05:00
yindo commented 2026-02-22 18:24:39 -05:00
Author
Owner
Copy Link
Copy Source

@Milor123 commented on GitHub (Aug 23, 2024):

New updated (Not solve) but I've tried change from Ollama (local installed in my pc) to Docker Ollama, why? Because I also have other problems related with the normal ollama

And Now have:

  • docker.io/ollama/ollama:latest
  • docker.io/milvusdb/milvus:v2.4.9-gpu
  • And the docker.io/mintplexlabs/anythingllm:latest

But it not work while try use milvuys (my container of anythingllm crashs with it), when use a docker chroma instead of milvus, then it works very well.

It looks like a bug @shatfield4 and not SO problem

@Milor123 commented on GitHub (Aug 23, 2024): New updated (Not solve) but I've tried change from Ollama (local installed in my pc) to Docker Ollama, why? Because I also have other problems related with the normal ollama And Now have: - docker.io/ollama/ollama:latest - docker.io/milvusdb/milvus:v2.4.9-gpu - And the docker.io/mintplexlabs/anythingllm:latest But it not work while try use milvuys (my container of anythingllm crashs with it), when use a docker chroma instead of milvus, then it works very well. It looks like a bug @shatfield4 and not SO problem
yindo commented 2026-02-22 18:24:39 -05:00
Author
Owner
Copy Link
Copy Source

@timothycarambat commented on GitHub (Aug 23, 2024):

While the error for the milvus connection is not very helpful, we can at least handled the exception to prevent a full crash of AnythingLLM, since that is an annoyance

@timothycarambat commented on GitHub (Aug 23, 2024): While the error for the milvus connection is not very helpful, we can at least handled the exception to prevent a full crash of AnythingLLM, since that is an annoyance
yindo commented 2026-02-22 18:24:39 -05:00
Author
Owner
Copy Link
Copy Source

@Milor123 commented on GitHub (Aug 23, 2024):

While the error for the milvus connection is not very helpful, we can at least handled the exception to prevent a full crash of AnythingLLM, since that is an annoyance

How can i help you to depurate it? what logs or what do you need?

@Milor123 commented on GitHub (Aug 23, 2024): > While the error for the milvus connection is not very helpful, we can at least handled the exception to prevent a full crash of AnythingLLM, since that is an annoyance How can i help you to depurate it? what logs or what do you need?
yindo commented 2026-02-22 18:24:39 -05:00
Author
Owner
Copy Link
Copy Source

@timothycarambat commented on GitHub (Aug 23, 2024):

Oh there are no more logs to get, i was just saying how the logs from their library does not help us out much since it has to do with GRPC connection termination. The stack you provided should be enough to help us get started to replicate this - assuming your connection information is accurate and that is not what is causing the issue

@timothycarambat commented on GitHub (Aug 23, 2024): Oh there are no more logs to get, i was just saying how the logs _from their library_ does not help us out much since it has to do with GRPC connection termination. The stack you provided should be enough to help us get started to replicate this - assuming your connection information is accurate and that is not what is causing the issue
yindo commented 2026-02-22 18:24:39 -05:00
Author
Owner
Copy Link
Copy Source

@Milor123 commented on GitHub (Aug 23, 2024):

Oh there are no more logs to get, i was just saying how the logs from their library does not help us out much since it has to do with GRPC connection termination. The stack you provided should be enough to help us get started to replicate this - assuming your connection information is accurate and that is not what is causing the issue

Ahh I understand, thank u very much, hey bro I have a technical question not directly related, but the reason I was trying another vector database, is because I thought it would help me solve my problem.
you see, I was trying to embed a document with words as if they were dictionaries.

example: Cat. It has four legs and says meow
Dog Its a crazy zombie... etc etc
Jagar Its not humanoid

I want the LLM, when the user asks for cat, to tell him what it is like.
The problem is that I do not know if it is the embedding model or the vector database, but when I type the word, the search that is done does not find anything that has the keywords, when I search for dog, nothing appears. In many occasions it returns me a context or several useless contexts where the word is not there.

Is it that I misunderstand how all this works? or should I change the database or the embedding?
So far I have changed the size of the contexts, the size in which I part the vectors, tried some different embedding models, and changed to chroma, but when I ask the LLM using only one word as it should be, then I get results that are NOT related, they are useless, so the LLM will never know the answer.

What should I do for try solve this problem? Change Vector DB or what do you think?

@Milor123 commented on GitHub (Aug 23, 2024): > Oh there are no more logs to get, i was just saying how the logs _from their library_ does not help us out much since it has to do with GRPC connection termination. The stack you provided should be enough to help us get started to replicate this - assuming your connection information is accurate and that is not what is causing the issue Ahh I understand, thank u very much, hey bro I have a technical question not directly related, but the reason I was trying another vector database, is because I thought it would help me solve my problem. you see, I was trying to embed a document with words as if they were dictionaries. example: Cat. It has four legs and says meow Dog Its a crazy zombie... etc etc Jagar Its not humanoid I want the LLM, when the user asks for cat, to tell him what it is like. The problem is that I do not know if it is the embedding model or the vector database, but when I type the word, the search that is done does not find anything that has the keywords, when I search for dog, nothing appears. In many occasions it returns me a context or several useless contexts where the word is not there. Is it that I misunderstand how all this works? or should I change the database or the embedding? So far I have changed the size of the contexts, the size in which I part the vectors, tried some different embedding models, and changed to chroma, but when I ask the LLM using only one word as it should be, then I get results that are NOT related, they are useless, so the LLM will never know the answer. What should I do for try solve this problem? Change Vector DB or what do you think?
yindo commented 2026-02-22 18:24:39 -05:00
Author
Owner
Copy Link
Copy Source

@timothycarambat commented on GitHub (Aug 23, 2024):

Email me this question at team@mintplexlabs.com - I dont want to pollute this GH issue with something off topic!

@timothycarambat commented on GitHub (Aug 23, 2024): Email me this question at team@mintplexlabs.com - I dont want to pollute this GH issue with something off topic!
yindo commented 2026-02-22 18:24:39 -05:00
Author
Owner
Copy Link
Copy Source

@Milor123 commented on GitHub (Aug 23, 2024):

Thank u very much bro !

Email me this question at team@mintplexlabs.com - I dont want to pollute this GH issue with something off topic!

@Milor123 commented on GitHub (Aug 23, 2024): Thank u very much bro ! > Email me this question at [team@mintplexlabs.com](mailto:team@mintplexlabs.com) - I dont want to pollute this GH issue with something off topic!
yindo changed title from [BUG]: Crash when try embed a document using milvus to [GH-ISSUE #2159] [BUG]: Crash when try embed a document using milvus 2026-06-05 14:40:36 -04:00
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
Desktop
Docker
Integration Request
Integration Request
OS: Linux
OS: Mobile
OS: Windows
UI/UX
blocked
bug
bug
core-team-only
documentation
duplicate
embed-widget
enhancement
feature request
github_actions
good first issue
investigating
needs info / can't replicate
possible bug
pull-request
Mirrored from GitHub Pull Request
 question
stage: specifications
wontfix
No Label possible bug
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
yindo
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: Mintplex-Labs/anything-llm#1407
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?