Open corneliusroemer opened 1 year ago
@corneliusroemer What's the output of nextstrain version --verbose
?
Hm. This is coming from inside the aiobotocore client used by fsspec/s3fs. My first guess (probably wrong) is that you have an incompatible combination of s3fs and aiobotocore in whatever environment nextstrain
itself is running.
$ nextstrain version --verbose
nextstrain.cli 6.2.1
Python
/opt/homebrew/Caskroom/miniforge/base/envs/nextstrain/bin/python3.10
3.10.8 | packaged by conda-forge | (main, Nov 22 2022, 08:31:57) [Clang 14.0.6 ]
Runners
docker
nextstrain/base:build-20230411T103027Z (bb05df2db0ce, 2023-04-11 13:38:55 +0200 CEST)
augur 21.1.0
auspice v2.45.2
fauna e3ed8e1
sacra not present
conda
nextstrain-base 20230407T195218Z (h0dc7051_1_locked, nextstrain)
augur 21.1.0
auspice 2.45.1
singularity
docker://nextstrain/base (not present)
ambient (default)
augur 21.1.0
auspice 2.45.1
aws-batch
unknown
I ask because I wonder if https://github.com/bioconda/bioconda-recipes/pull/39711 is implicated. Though not certain if the expectation would be that you have a version of nextstrain-cli
from Bioconda before or after that…
Ah, the version is 6.2.1, which is after that Conda packaging change. Hmm.
This is my pip list
:
Is it ok for boto3 and botocore to have different versions?
boto3 1.26.111
botocore 1.29.111
This seems worth digging into, but I'd love if someone else in @nextstrain/core could pick it up.
Current Behavior
When trying to run the nextstrain/rsv workflow using
nextstrain build --aws-batch ...
I get an error:botocore.exceptions.ClientError: An error occurred (MaxMessageLengthExceeded) when calling the PutObject operation: Your request was too big.
Expected behavior
It works, I don't get such an error, or at least cli can recover and advise me on how to workaround the issue.
How to reproduce
Steps to reproduce the current behavior:
gh repo clone nextstrain/rsv
Your environment: if running Nextstrain locally
nextstrain.cli 6.2.1
Full logs