janhq/jan · error · Error

Vector DB extension not available

Error message

Vector DB extension not available

What it means

Thrown by RAGExtension.ingestAttachments() (thread-level ingestion) when the resolved VectorDB extension is missing OR lacks createCollection/insertChunks. This is a different capability check than project-level: it requires the foundational vector-db methods. Without them, no thread collection can be created or populated, so the call aborts before the file loop.

Source

Thrown at extensions/rag-extension/src/index.ts:445

  ): Promise<{
    filesProcessed: number
    chunksInserted: number
    files: AttachmentFileInfo[]
  }> {
    if (!threadId || !Array.isArray(files) || files.length === 0) {
      return { filesProcessed: 0, chunksInserted: 0, files: [] }
    }

    // Respect feature flag: do nothing when disabled
    if (this.config.enabled === false) {
      return { filesProcessed: 0, chunksInserted: 0, files: [] }
    }

    const vec = window.core?.extensionManager.get(
      ExtensionTypeEnum.VectorDB
    ) as unknown as VectorDBExtension
    if (!vec?.createCollection || !vec?.insertChunks) {
      throw new Error('Vector DB extension not available')
    }

    // Load settings
    const s = this.config
    const maxSize = (s?.enabled === false ? 0 : s?.maxFileSizeMB) || undefined
    const chunkSize = s?.chunkSizeChars as number | undefined
    const chunkOverlap = s?.overlapChars as number | undefined

    let totalChunks = 0
    const processedFiles: AttachmentFileInfo[] = []

    for (const f of files) {
      if (!f?.path) continue
      if (maxSize && f.size && f.size > maxSize * 1024 * 1024) {
        throw new Error(
          `File '${f.name}' exceeds size limit (${f.size} bytes > ${maxSize} MB).`
        )
      }

View on GitHub (pinned to fad3f12a14)

Solutions

1. Enable and load vector-db-extension (verify via window.core.extensionManager.get(ExtensionTypeEnum.VectorDB)).
2. Reinstall/update vector-db-extension so createCollection and insertChunks are present.
3. Check the extension's startup log for a native-binding load failure.
4. Guard the caller: skip RAG ingestion and degrade to plain LLM context when the extension is unavailable.

Example fix

// before
await rag.ingestAttachments(threadId, files)

// after
const vec = window.core?.extensionManager.get(ExtensionTypeEnum.VectorDB) as any
if (!vec?.createCollection || !vec?.insertChunks) {
  logger.warn('Vector DB unavailable; attachments will not be indexed')
  return { filesProcessed: 0, chunksInserted: 0, files: [] }
}
await rag.ingestAttachments(threadId, files)

const vec = window.core?.extensionManager.get(ExtensionTypeEnum.VectorDB) as any
const ok = vec && typeof vec.createCollection === 'function' && typeof vec.insertChunks === 'function'
if (!ok) {
  // degrade RAG or prompt to enable/update the extension
}

function hasCoreVectorDB(vec: unknown): vec is { createCollection: Function; insertChunks: Function } {
  const v = vec as any
  return !!v && typeof v.createCollection === 'function' && typeof v.insertChunks === 'function'
}

When it happens

Trigger: Calling ingestAttachments(threadId, files) when vector-db-extension is not registered, crashed, or is a stripped-down build missing createCollection/insertChunks; ExtensionTypeEnum.VectorDB resolves to a different extension than expected.

Common situations: vector-db-extension disabled while rag-extension is enabled; extension failed to load due to a missing native dependency (e.g. sqlite-vec); partial upgrade left an incompatible extension installed.

Related errors

• Vector DB extension does not support project-level ingestion
• llamacpp extension not available
• llamacpp extension not available
• File '${file.name}' has already been attached to this thread
• RAG extension not available

AI-assisted analysis of janhq/jan@fad3f12a14 (2026-08-12). Data as JSON: /api/errors/9f947a6436d2a2d8. Report an issue: GitHub.