Closed aliaksei135 closed 3 years ago
Partly fixed in the above PR, however this does not allow for the risk map to be produced without the failed data. I will close this and #65 when #66 is merged and if failing data becomes an issue in future will look at implementing that idea.
The errors being received are <Response 500> (Internal server errors). Still looking to see if something has changed regarding request queries on their side as that's the only thing I can think of that's happened.
Okay so I have done a bit more research on this. Here are a few things to try:
POST
requests instead of GET
, like we are doing nowUser-Agent
header, something like seedpod-ground-risk v0.13.0 (Python 3.8/requests v<xxx>;)
Content-Type
header to what it actually is: application/x-www-form-urlencoded; charset=UTF-8
@Zach10a Good stuff that using POSTs fixed the error. I think it's good practice to implement the other 2 suggestions here while we are on it as well!
Closed by #66
Currently any failure in fetching OSM data results in an unhandled exception and complete failure to generate the map without any user notification (aside from the debug console). This should be handled gracefully and generate the risk map without the failed data if possible. Either way the user should be notified of the failure in the UI.