vxcontrol/pentagi · error

knowledge: compute embedding

Error message

knowledge: compute embedding: %w

What it means

CreateDocument computes an embedding for the new document content (truncated to maxEmbeddingBytes) before insertion. Failure of embedder.EmbedDocuments — nil embedder, provider auth/network errors — is wrapped as 'knowledge: compute embedding'.

Solutions

1. If the wrapped error says the embedder is not configured, set the embedding provider env vars and restart.
2. Validate provider credentials with a direct API call to the embeddings endpoint.
3. Check network reachability from the backend to the provider (DNS, proxy, ollama service name).
4. For 429/timeout causes, retry with backoff; for bulk imports, throttle creation rate.

Example fix

// before
vecs, err := ks.embedder.EmbedDocuments(ctx, []string{embeddingText})
// after
if ks.embedder == nil {
    return nil, fmt.Errorf("knowledge: embedding provider not configured; set embedding env vars")
}
vecs, err := ks.embedder.EmbedDocuments(ctx, []string{embeddingText})

if embedder == nil {
    return errors.New("embedding provider not configured")
}
if strings.TrimSpace(content) == "" {
    return errors.New("document content must not be empty")
}
if len(content) > maxEmbeddingBytes*100 { // absurd input guard
    return errors.New("document content too large")
}

func isEmbedderConfigErr(err error) bool {
    return strings.Contains(err.Error(), "not configured")
}

doc, err := store.CreateDocument(ctx, userID, input)
if err != nil {
    if strings.Contains(err.Error(), "compute embedding") {
        // check provider status/key before surfacing to user
        return fmt.Errorf("document indexing temporarily unavailable: %w", err)
    }
    return err
}

When it happens

Trigger: CreateDocument called when no embedding provider is configured, the provider API key is wrong/expired, the endpoint is unreachable, or the request is rate-limited.

Common situations: Deploying without embedding env vars (nil embedder by design); Ollama not running in the compose network; provider outage or 429 under bulk document creation.

Related errors

• knowledge: embed query
• failed to load document
• knowledge: embedding provider is not configured
• adviser handler is required
• barrier (done) handler is required

AI-assisted analysis of vxcontrol/pentagi@ea665308ba (2026-09-01). Data as JSON: /api/errors/50cfae61685f083d. Report an issue: GitHub.

Appendix: source

Thrown at backend/pkg/database/knowledge/knowledge.go:497

	if input.AnswerType != nil {
		meta.AnswerType = string(*input.AnswerType)
	}
	if input.CodeLang != nil {
		meta.CodeLang = *input.CodeLang
	}

	content := strings.TrimSpace(input.Content)
	meta.PartSize = len(content)
	meta.TotalSize = len(content)

	// Truncate to embedding size limit for vector computation; full content goes to DB.
	embeddingText := content
	if len(embeddingText) > ks.maxEmbeddingBytes {
		embeddingText = embeddingText[:ks.maxEmbeddingBytes]
	}
	vecs, err := ks.embedder.EmbedDocuments(ctx, []string{embeddingText})
	if err != nil {
		return nil, fmt.Errorf("knowledge: compute embedding: %w", err)
	}
	if len(vecs) == 0 {
		return nil, fmt.Errorf("knowledge: embedder returned no vectors")
	}

	cmJSON, err := metaToJSON(meta)
	if err != nil {
		return nil, fmt.Errorf("knowledge: marshal cmetadata: %w", err)
	}

	id := uuid.New()
	docID, err := ks.db.InsertKnowledgeDocument(ctx, database.InsertKnowledgeDocumentParams{
		Uuid:      id,
		Document:  nsOf(content),
		Embedding: formatVector(vecs[0]),
		Cmetadata: cmJSON.RawMessage,
	})
	if err != nil {

View on GitHub (pinned to ea665308ba)