nautechsystems/nautilus_trader · error · anyhow::Error

partitioned cancel failed: {}

Error message

partitioned cancel failed: {}

What it means

This error aggregates all individual cancellation failures from a partitioned (chunked) batch cancel broadcast on the dYdX adapter. `broadcast_partitioned_cancels` collects per-partition error messages into a vector and, if any failed, bails with all messages joined by '; ' so the caller (cancel_all_orders or batch_cancel_orders) sees the complete failure set at once.

Source

Thrown at crates/adapters/dydx/src/execution/mod.rs:1179

                                &emitter,
                                clock,
                                &msg,
                            );
                        }
                        errors.push(msg);
                    }
                }
            }
            Err(e) => {
                let msg = format!("Failed to build long-term cancel messages: {e:?}");
                log::error!("{msg}");
                errors.push(msg);
            }
        }
    }

    if !errors.is_empty() {
        anyhow::bail!("partitioned cancel failed: {}", errors.join("; "));
    }

    Ok(())
}

fn emit_partitioned_cancel_rejections(
    orders: &[DydxCancelOrderRequest],
    emitter: &ExecutionEventEmitter,
    clock: &AtomicTime,
    reason: &str,
) {
    for order in orders {
        emitter.emit_order_cancel_rejected_event(
            order.strategy_id,
            order.instrument_id,
            order.client_order_id,
            order.venue_order_id,
            reason,

View on GitHub (pinned to 18893faf8b)

Solutions

1. Inspect the joined error messages to identify which partitions/orders failed
2. Retry cancellation for the failed orders once connectivity is restored
3. Verify the block time monitor has a current block height before mass-canceling
4. Check node/validator health and gas settings if broadcast rejections persist

Example fix

// before: blindly retrying everything
client.cancel_all_orders(cmd).await?;
// after: parse per-partition failures and retry selectively
if let Err(e) = client.cancel_all_orders(cmd) {
    for part in e.to_string().split("; ") {
        log::warn!("partition failure: {part}");
    }
    // re-issue cancels for the affected orders
}

if !client.is_connected() { skip_mass_cancel(); }

match client.cancel_all_orders(cmd) {
    Err(e) => {
        let failed: Vec<&str> = e.to_string().split("; ").collect();
        for f in failed { log::warn!("retry needed: {f}"); }
    }
    Ok(_) => {}
}

When it happens

Trigger: Calling `cancel_all_orders` or `batch_cancel_orders` when one or more partitioned broadcast chunks fail (e.g. TX broadcast rejected on-chain, block height stale, order encoding missing for some orders).

Common situations: Canceling many open orders during network congestion where some tx broadcasts are rejected; canceling orders whose client order IDs were never encoded by this client instance; node outages affecting only some partitions.

Related errors

• Cannot cancel order: not connected
• Client order ID not found in cache
• Cannot cancel orders: not connected
• Transaction broadcast failed: code={}, log={}
• Broadcast semaphore closed: {e}

AI-assisted analysis of nautechsystems/nautilus_trader@18893faf8b (2026-09-08). Data as JSON: /api/errors/b7b76ff5a3389202. Report an issue: GitHub.