vxcontrol/pentagi · error

failed to create flow msg log: %w

Error message

failed to create flow msg log: %w

What it means

Second step of newFlowProviderWorkers: the message log worker is created via cnts.mlc.NewFlowMsgLog(ctx, flowID, pub). A failure here is wrapped with this message and stops flow worker construction, meaning the flow cannot start without message logging persistence.

Source

Thrown at backend/pkg/controller/flow.go:1174

	}

	return nil
}

func newFlowProviderWorkers(
	ctx context.Context,
	flowID int64,
	cnts *flowProviderControllers,
	pub subscriptions.FlowPublisher,
) (*flowProviderWorkers, error) {
	alw, err := cnts.alc.NewFlowAgentLog(ctx, flowID, pub)
	if err != nil {
		return nil, fmt.Errorf("failed to create flow agent log: %w", err)
	}

	mlw, err := cnts.mlc.NewFlowMsgLog(ctx, flowID, pub)
	if err != nil {
		return nil, fmt.Errorf("failed to create flow msg log: %w", err)
	}

	slw, err := cnts.slc.NewFlowSearchLog(ctx, flowID, pub)
	if err != nil {
		return nil, fmt.Errorf("failed to create flow search log: %w", err)
	}

	tlw, err := cnts.tlc.NewFlowTermLog(ctx, flowID, pub)
	if err != nil {
		return nil, fmt.Errorf("failed to create flow term log: %w", err)
	}

	vslw, err := cnts.vslc.NewFlowVectorStoreLog(ctx, flowID, pub)
	if err != nil {
		return nil, fmt.Errorf("failed to create flow vector store log: %w", err)
	}

	tclw, err := cnts.tclc.NewFlowToolCallLog(ctx, flowID, pub)

View on GitHub (pinned to ea665308ba)

Solutions

1. Read the wrapped cause in backend logs to identify the failing SQL
2. Check PostgreSQL health and message_logs table status
3. Apply pending migrations and retry flow creation
4. Re-check that the flow row still exists (concurrent deletion)

Example fix

// before: sequential creation, partial failure leaves nothing behind
mlw, err := cnts.mlc.NewFlowMsgLog(ctx, flowID, pub)
if err != nil { return nil, fmt.Errorf("failed to create flow msg log: %w", err) }
// after: ensure ctx deadline so failures surface as timeouts, and log flowID
ctx, cancel := context.WithTimeout(ctx, 30*time.Second)
defer cancel()
mlw, err := cnts.mlc.NewFlowMsgLog(ctx, flowID, pub)
if err != nil {
    return nil, fmt.Errorf("failed to create flow msg log (flow=%d): %w", flowID, err)
}

var exists int
if err := db.QueryRowContext(ctx, `SELECT COUNT(1) FROM flows WHERE id=$1`, flowID).Scan(&exists); err != nil || exists == 0 {
    return fmt.Errorf("flow %d missing before creating msg log", flowID)
}

func hasValidFlowID(flowID int64) bool { return flowID > 0 }

mlw, err := cnts.mlc.NewFlowMsgLog(ctx, flowID, pub)
if err != nil {
    if errors.Is(err, context.Canceled) { return nil, err }
    return nil, fmt.Errorf("failed to create flow msg log (flow=%d): %w", flowID, err)
}

When it happens

Trigger: NewFlowMsgLog fails during flow startup — DB error creating/accessing the message log storage for this flow, invalid flowID, or context cancellation after the agent log step succeeded.

Common situations: Postgres connection dropped between the agent-log and msg-log steps; flow was deleted concurrently; oversized/backlogged message_logs table causing timeouts; missing migration for the message log schema.

Related errors

• failed to create flow agent log: %w
• failed to create flow search log: %w
• failed to create flow term log: %w
• failed to create flow vector store log: %w
• failed to create flow tool call log: %w

AI-assisted analysis of vxcontrol/pentagi@ea665308ba (2026-09-01). Data as JSON: /api/errors/3cb7dc6b5f0e5442. Report an issue: GitHub.