conda env create error

[w4n9H]

SQL语句为

load csv.`wine-quality.csv`  options header="true"  and inferSchema="true"  as t1;

train t1 as PythonAlg.`mltest` 
where pythonScriptPath="MLtest" 
and keepVersion="true" 
and dataLocalFormat="csv"
and `systemParam.envs`='''{"MLFLOW_CONDA_HOME": "/opt/miniconda3"}'''
as res;

save overwrite res as json.`new_pyml_result`;

使用spark-submit通过yarn向集群提交任务，出现报错 java.lang.RuntimeException: Nonzero exit value: 1，错误出现在Project.scala中创建conda env的那一步
打印详细的create语句，出现了多条
/opt/miniconda3/bin/conda env create -n mlflow-a6e270601633eae621911dbfd740bdebe63e458a --file /tmp/27af026b-f309-499f-a536-df14ceb42a20.yaml
还有一条为
/opt/miniconda3/bin/conda env create -n mlflow-a6e270601633eae621911dbfd740bdebe63e458a --file /tmp/4bf24767-853f-41ac-9575-79925eb03c85.yaml
看这个报错应该是 --file 后面的文件不存在导致的
使用local模式在本地跑是正常的

[allwefantasy]

Have you configured your cluster by this tutorial?

[allwefantasy]

Also there are some bugs about python in v1.1.7 . more detail please check here, and we will fix it in bug-release version v.1.1.7.1 and maybe this weekend.

[allwefantasy]

We have released v1.1.7.1 to fix this issue.
The doc is here- how to manage you python environment

We will close this issue. If you find any problem again ,please reopen it or create a new one .