This week I’ve been getting a lot of errors when placing orders the first few minutes after market open.
Favorite error: “The remote server returned an error: (504) Gateway Timeout.” {“code”:50410000,“message”:“request timed out”}
The above condition may last for minutes. No mention of outages on Alpaca system status page. All is green.
Also, Alpaca likes to delay relaying my orders to exchanges with status “Order Not Yet Sent to Exchange”. Delays may last 10s of seconds!
I don’t know what type of order flow Alpaca gets but they better fix or beef up their infrastructure. How difficult could it be to take a message from one place and send it to another?
With the market selling off right after open the last few days the $ lost could’ve easily covered the IBKR account cost.
I’d like to hear what Dan from Alpaca has to say about that?
Same thing happened just now. The remote server returned an error: (504) Gateway Timeout
And system status is “No incidents reported today”
Well, i’m reporting incidents every day this week!.
how is there no response on this ticket yet? seems like this platform really is too good to be true… I want to get off IBKR but it still seems like the best choice by far
Yes, the issue didn’t go away but the error message changed. Now when my algo issues cancel/replace order instead of error: (504) Gateway Timeout it is now getting error: (422) Unprocessable Entity. code: 42210000, "order is not open"
Now that Alpaca has boasted about their spectacular order execution performance improvements here Alpaca Launches Next-Gen Order Management System that Makes Order Processing 100x Faster maybe it is time to reconsider the API rate limit. error code":42910000,“message”:“rate limit exceeded” Often times i need to sell 200 stocks on open and using Alpaca it takes a very long time
At least i’m not getting error: (504) Gateway Timeout and error: (422) Unprocessable Entity. code: 42210000, “order is not open” anymore. Those issues seem to be fixed. So i give credit where credit is due
IMHO they can’t just say “We’re here to help”
They should also actually help or at least acknowledge when there’s a fundamental technical issue with the product.
I can only give the benefit of the doubt to say that they must be overwhelmed?
A greater investment in a community liaison would probably pay big dividends.