Closed rabernat closed 2 years ago
Locally I have pywren tests failing
There has not been a pywren release for over 3 years, since it has now become lithops. I wonder if we should remove or disable the pywren executor for this reason.
Merging #112 (ac9737f) into master (0a0d1eb) will decrease coverage by
0.48%
. The diff coverage is100.00%
.
@@ Coverage Diff @@
## master #112 +/- ##
==========================================
- Coverage 96.81% 96.32% -0.49%
==========================================
Files 12 11 -1
Lines 534 490 -44
Branches 121 112 -9
==========================================
- Hits 517 472 -45
Misses 11 11
- Partials 6 7 +1
Impacted Files | Coverage Δ | |
---|---|---|
rechunker/executors/dask.py | 100.00% <100.00%> (+5.45%) |
:arrow_up: |
rechunker/executors/prefect.py | 100.00% <100.00%> (ø) |
|
rechunker/executors/python.py | 100.00% <100.00%> (ø) |
|
rechunker/pipeline.py | 100.00% <100.00%> (ø) |
|
rechunker/types.py | 100.00% <100.00%> (ø) |
|
rechunker/api.py | 98.04% <0.00%> (-1.96%) |
:arrow_down: |
Continue to review full report at Codecov.
Legend - Click here to learn more
Δ = absolute <relative> (impact)
,ø = not affected
,? = missing data
Powered by Codecov. Last update 0a0d1eb...ac9737f. Read the comment docs.
Ok so in fb8b1ee I just removed the pywren executor. If it is unmaintained, it's not feasible for us to find workarounds to keep supporting it. @tomwhite - are you using the pywren executor these days?
I had a look at lithops, and it looks like it should be pretty straightforward to support. It would be nice to have a serverless option. If someone wants to implement and maintain a lithops executor, I would be happy to support that effort.
@tomwhite - are you using the pywren executor these days?
No, I'm not.
For some reason the test workflow is not running.
GitHub actions was having issues earlier, but nothing is reported at https://www.githubstatus.com/ at the moment.
Oh, https://github.com/pangeo-data/rechunker/actions/runs/2029808478 shows
The workflow is not valid. .github/workflows/ci.yaml (Line: 9, Col: 7): Unexpected value 'fast-fail'
This was a little coding project while stuck in a plane for a few hours. Here I am basically just copying over pangeo-forge-recipes' latest executors, which we spent a lot of time optimizing.
Locally I have pywren tests failing with a message
TypeError: cannot pickle 'weakref' object
. Was not able to debug this quickly as I'm not too familiar with pywren.