oli 0 Report post Posted April 25, 2019 Hi @Physicsman - I've got another question for you... it's been a while I've recently started getting "Error 429 - Too Many Requests" when submitting multiple newstoplimitorder requests at market opening times. Would you be able to shed any light on - what limits are typically in place on an account (do they differ between consumer / professional accounts?) - what counts towards a 'request' (e.g. is a historical bars request treated equally to a newstoplimitorder request?) - how to find out the current level remaining calls available at any point in time. I couldn't find any mention in the documentation.. - whether it's possible to increase that limit? thank you so much, as always Oliver Share this post Link to post
Physicsman 25 Report post Posted April 30, 2019 Hi Oli, 1) There is a throttle of 500 requests over a 5 second window, which should be sufficient for most users. This applies to the API as a whole and does not differentiate between account types. 2) Any call to the API counts as a request. Yes, a request for historical price data is treated as a request equally to a newstoplimitorder request. 3) We do not have a call to show a counter of requests remaining within the 5 second window. 4) A request could be made to the appropriate engineering team for a change in the throttle limit, but of course I cannot say in advance what their response would be. How many requests are you submitting at market open? I would be rather surprised if you are submitting >500 within 5 seconds. Kind Regards, PM Share this post Link to post
oli 0 Report post Posted May 14, 2019 Hi PM - thank you so much for coming back to me and, yeah, it seemed strange to me too. However, I think I identified the problem as I haven't had the error since the original post. It appears to have been linked to an underpowered server making the calls to the CI api. I make multiple calls at market opening times and it seems that this activity was pushing the server to its limits, potentially causing repeated or malformed api queries. Certainly no part of the code could have generated 500 requests over 5 seconds... Since doubling the processors and memory it appears to have solved the problem. No errors since then. Bit strange, but hopefully resolved. All the best Oliver Share this post Link to post
Physicsman 25 Report post Posted May 14, 2019 Hi Oli, Appreciate the update - glad to hear everything is working fine now! Kind Regards, PM Share this post Link to post