You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
[Binance] Futures stop algo order stuck in TRIGGERING is reported as Accepted; strategy cannot detect the unprotected position
#5378
On the Binance USD-M Futures demo environment, a reduce-only STOP_MARKET algo order was triggered but never produced an order, and stayed in TRIGGERING. The position it was meant to close stayed open, and the strategy had no way to tell from Nautilus events or the cache.
As far as I can tell from the code this is by design, so I'm opening it as a discussion rather than a bug.
What happened (2.0.0rc5, demo, times UTC, 2026-10-19)
14:00:03 a BUY 1.103 ETHUSDT STOP_MARKET reduceOnly stop was accepted (OrderAccepted, algo id 1000000246753436).
15:31:17 the venue marked it triggered. Querying GET /fapi/v1/algoOrder the next day still returned:
The 1.103 short stayed open with no working protection. In Nautilus the stop was still ACCEPTED, so a strategy checking "are my open stops covering the position?" sees nothing wrong.
16:00:04 cancelling it after placing a replacement stop returned -2011 Unknown order sent.
The trigger itself looks like a demo venue glitch (production ETHUSDT high for that minute was 2410.43, below the trigger), so I don't expect this often on production. But a stop that triggers and then does not execute leaves a position unprotected, and today that state is invisible to the strategy.
Why it is invisible (v2.0.0rc7)
REST status reports map TRIGGERING to Accepted (models.rs#L1716), so reconciliation also reports it as a normal working order.
An ALGO_UPDATE with TRIGGERING is only logged at debug level (dispatch.rs#L1409). No event reaches the strategy.
Question
Would it make sense to surface this state, for example emitting OrderTriggered when the algo order reports TRIGGERING, or exposing the venue algo status on the order? Then a strategy could notice a stop that triggered but did not fill within some time and close the position itself.
Our plan is to work around it in our strategy by polling openAlgoOrders and by closing the position if price has passed the stop while the position is still open. If there is a better existing way to detect this in Nautilus, I'd be glad to hear it.
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
Uh oh!
There was an error while loading. Please reload this page.
On the Binance USD-M Futures demo environment, a reduce-only
STOP_MARKETalgo order was triggered but never produced an order, and stayed inTRIGGERING. The position it was meant to close stayed open, and the strategy had no way to tell from Nautilus events or the cache.As far as I can tell from the code this is by design, so I'm opening it as a discussion rather than a bug.
What happened (2.0.0rc5, demo, times UTC, 2026-10-19)
BUY 1.103 ETHUSDT STOP_MARKET reduceOnlystop was accepted (OrderAccepted, algo id1000000246753436).GET /fapi/v1/algoOrderthe next day still returned:{"algoId": 1000000246753436, "orderType": "STOP_MARKET", "side": "BUY", "quantity": "1.103", "algoStatus": "TRIGGERING", "actualOrderId": "", "triggerPrice": "2423.40", "reduceOnly": true, "createTime": 1792418403391, "triggerTime": 1792423877699}ACCEPTED, so a strategy checking "are my open stops covering the position?" sees nothing wrong.-2011 Unknown order sent.The trigger itself looks like a demo venue glitch (production ETHUSDT high for that minute was 2410.43, below the trigger), so I don't expect this often on production. But a stop that triggers and then does not execute leaves a position unprotected, and today that state is invisible to the strategy.
Why it is invisible (v2.0.0rc7)
TRIGGERINGtoAccepted(models.rs#L1716), so reconciliation also reports it as a normal working order.ALGO_UPDATEwithTRIGGERINGis only logged at debug level (dispatch.rs#L1409). No event reaches the strategy.Question
Would it make sense to surface this state, for example emitting
OrderTriggeredwhen the algo order reportsTRIGGERING, or exposing the venue algo status on the order? Then a strategy could notice a stop that triggered but did not fill within some time and close the position itself.Our plan is to work around it in our strategy by polling
openAlgoOrdersand by closing the position if price has passed the stop while the position is still open. If there is a better existing way to detect this in Nautilus, I'd be glad to hear it.All reactions