[2026-08-22 20:54:50.879] [debug] [mapillary:tools] 20:54:50.875 - ERROR - HTTPError: POST https://graph.mapillary.com//finish_upload => 400 Bad Request: {“error”: {“message”: “It appears that you are abusing this feature because you are using it too often. You are temporarily unable to use the feature.”, “type”: “OAuthException”, “code”: 368, “error_data”: {“sentry_block_data”: "AettA_1uSm0dmGgWRrGxKj…(503 chars truncated)
I’m getting this error. I’m uploading the data from the holiday, maybe I overloaded the server a bit (around 500 GB uploaded today I think) Any idea how to fix this @boris ?
Hi @TheWizard, thanks for the report. Not much of a useful answer, and I hope that works for you, as I would suggest waiting at least 12h (or better 24h) and just uploading later. The reason is that we rarely can do something about this fast, and especially during the weekend.
Some details: I don’t have the number by hand, but the number of uploads you have to make to reach it is significant, yet we’ve seen it being triggered before. The unit of tracking in this case is one sequence (not one image). So, if you have plenty of small sequences, you can hit it, and the easiest option is just to wait 24h.
Thanks for the quick answer! There where multiple uploads at the same time, but this is not the first time for this amount of data (partially video, partially sampled images in standard chunks of 1000).
Could you check on Monday what the limits are? Of the cooldown time and the upload limit. Just so that I (we als users) can take into account and not hit the limits in the future.
I spent some time digging, and it looks like that’s my limit on what is possible to dig. In short, there are two main types of blockers (they are more granular, that’s just a high level): fixed and adjustable.
Hard blockers on this endpoint are extremely high values (they are per minute, and those were not triggered, they are irrelevent for all the cases).
But there are more complex, adjustable rules which are generated based on some algorithms I just don’t know in detail. They are triggered when there’s a spike in uploading which can catch systems’ attention, and even the cooldown is changing adjustably.
If I’m not mistaken, you uploaded around 4k sequences (a bit more) in approx 12 hours. So, that’s the best (empirical) limit I can think of: 4k in 12 hours.
Thanks for the digging! . What is a sequence in Mapillary terms, the 1000 samples per block sequence? I previous uploaded also a lot at the same time, maybe I had 1 upload too much running… I’ll try to maximize my upload line, maybe it’s too much…
I didn’t test it, I first want to wait your answer, but the test was good, no blockage!
. What is a sequence in Mapillary terms, the 1000 samples per block sequence? I previous uploaded also a lot at the same time, maybe I had 1 upload too much running… I’ll try to maximize my upload line, maybe it’s too much…