search

icesat2py / icepyx

Python tools for obtaining and working with ICESat-2 data
https://icepyx.readthedocs.io/en/latest/
BSD 3-Clause "New" or "Revised" License
210 stars 107 forks source link

calls to OpenAltimetry API 404 #454

Open JessicaS11 opened 12 months ago

JessicaS11 commented 12 months ago

Last week, OpenAltimetry transitioned their backend from its original domain to NSIDC and from grabbing data from a local server to the cloud-hosted copy of the data (ICESat-2 only so far). This has caused icepyx tests that utilize the OA programmatic API to start timing out and thus fail (see https://app.travis-ci.com/github/icesat2py/icepyx/builds/266433627 from #452).

JessicaS11 commented 12 months ago

After communicating with NSIDC on this, there are a few open tickets to deal with issues surrounding the migration of OpenAltimetry to NSIDC. Fixing the programmatic API and related functionality is one of them. Currently open and near-merging PRs in icepyx (e.g. #451, #452) are not time critical, so we will wait a few days to see if the issue is resolved and we can put through a patch to fix the failing tests (i.e. update the base_url for programmatic requests) before merging those.

mfisher87 commented 1 month ago

I believe this symptom has changed since the issue was opened from a timeout to a 404 (see #540).

mlampros commented 1 month ago

@JessicaS11, @mfisher87 the API endpoints that are used in the Python code base (http://openaltimetry.earthdatacloud.nasa.gov/data/api/icesat2) are not the same I use for the IceSat2R package (https://openaltimetry.earthdatacloud.nasa.gov/data/openapi/swagger-ui/index.html). The latter url currently works except for level3a

mfisher87 commented 1 month ago

Nice, thank you! I was looking all over trying to find if there was a working API URL someplace, and just couldn't find anything! :bow:

JessicaS11 commented 1 month ago

Sorry @mfisher87 that I referred you here and this issue isn't fully updated. I was working with Mikala via email on this. The url was the easy part once it stabilized(though at the time there wasn't any documentation, so I'm excited to see that - thanks @mlampros!). As noted in #495 the data format itself had changed, which would require rewriting things to parse (and plot) it. It shouldn't be too hard, but the bigger question was whether or not it was worth dev time and effort if nobody was using it...

mfisher87 commented 1 month ago

Right on, thanks for clarifying that, Jessica! That was exactly what I needed to know; updating #540 !