Morten Bendiksen
Morten Bendiksen
I took a look at the ComponentMapper. I must admit I struggled to see what it is doing that improves performance.
I think the actual storage in drogon is corrupted, so the warning might be valid?
I tried deleting storage, and running again the drogon_ahm.ert file. I don't see the warning any longer.
Do you mean remove all other folders under ".../ecl_test" that have a creation timestamp older than, say, 1 day?
So, I realized that the alternative is for the user to call the numpy_array function in stead of the [] indexing, and also change the expectations for the returned object....
I could not figure out a way to use the CLI compare commands and make it report failure. Tried something like this: `pytest-benchmark compare Linux-CPython-3.6-64bit/0001_my-benchmark.json Linux-CPython-3.6-64bit/0002_my-benchmark.json fail=mean:1%` I manually edited...
py.test-benchmark: error: unrecognized arguments: --fail=mean:1%
There are a few tests that can vary quite a bit in how long time it takes to execute. For those tests, I want to specify e.g. 50% variance in...
It seems this is what breaks our streaming endpoints, the lack of which kills throughput making everything slow.
A workaround for now is to use our connection pool directly in each endpoint doing streaming. Annoying, but works. Doing something like this for a psycopg pool: ```python @router.get("/stuff", response_model=list[Stuff])...