Lifecycle Configuration文件无法自动终止Amazon SageMaker Studio实例

我正在配置生命周期配置（LCC）文件，为Amazon SageMaker代码编辑器实现闲置阈值触发的自动终止功能。按照相关方案操作后，能在SageMaker Studio界面看到并选择该LCC文件，但实例闲置1小时后并未关机（测试时设置为60秒也无效）。使用的LCC文件内容如下：

# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved.
# SPDX-License-Identifier: MIT-0

#!/bin/bash
set -eux
ASI_VERSION=0.3.0

# User variables [update as needed]
IDLE_TIME_IN_SECONDS=60  # in seconds, change this to desired idleness time before app shuts down

# System variables [do not change if not needed]
CONDA_HOME=/opt/conda/bin
LOG_FILE=/var/log/apps/app_container.log # Writing to app_container.log delivers logs to CW logs.
SOLUTION_DIR=/var/tmp/auto-stop-idle # Do not use /home/sagemaker-user
PYTHON_PACKAGE=sagemaker_code_editor_auto_shut_down-$ASI_VERSION.tar.gz
PYTHON_SCRIPT_PATH=$SOLUTION_DIR/sagemaker_code_editor_auto_shut_down/auto_stop_idle.py

# Installing cron
sudo apt-get update -y
sudo sh -c 'printf "#!/bin/sh\nexit 0" > /usr/sbin/policy-rc.d'
sudo apt-get install -y cron

# Creating solution directory.
sudo mkdir -p $SOLUTION_DIR

# Downloading autostop idle Python package.
echo "Downloading autostop idle Python package..."
curl -LO --output-dir /var/tmp/ https://github.com/aws-samples/sagemaker-studio-apps-lifecycle-config-examples/releases/download/v$ASI_VERSION/$PYTHON_PACKAGE
sudo $CONDA_HOME/pip install -U -t $SOLUTION_DIR /var/tmp/$PYTHON_PACKAGE

# Touch file to ensure idleness timer is reset to 0
echo "Touching file to reset idleness timer"
touch /opt/amazon/sagemaker/sagemaker-code-editor-server-data/data/User/History/startup_timestamp

# Setting container credential URI variable to /etc/environment to make it available to cron
sudo /bin/bash -c "echo 'AWS_CONTAINER_CREDENTIALS_RELATIVE_URI=$AWS_CONTAINER_CREDENTIALS_RELATIVE_URI' >> /etc/environment"

# Add script to crontab for root.
echo "Adding autostop idle Python script to crontab..."
echo "*/2 * * * * /bin/bash -ic '$CONDA_HOME/python $PYTHON_SCRIPT_PATH --time $IDLE_TIME_IN_SECONDS --region $AWS_DEFAULT_REGION >> $LOG_FILE'" | sudo crontab -

排查与修复步骤 #

• 确认cron服务运行状态：进入代码编辑器终端，执行sudo service cron status检查cron是否正常运行。若未启动，执行sudo service cron start启动服务，并通过sudo systemctl enable cron设置开机自启。
• 查看日志定位错误：执行cat /var/log/apps/app_container.log查看脚本执行日志，排查是否存在依赖缺失、路径错误或权限问题。
• 固定环境变量值：cron环境中可能无法正确获取AWS_DEFAULT_REGION，修改crontab命令，直接替换为具体区域（如us-east-1），避免依赖环境变量。
• 验证脚本路径有效性：执行ls /var/tmp/auto-stop-idle/sagemaker_code_editor_auto_shut_down/auto_stop_idle.py，确认脚本文件是否存在。若路径错误，检查LCC文件中的SOLUTION_DIR和PYTHON_SCRIPT_PATH配置。
• 优化cron命令格式：原命令中的/bin/bash -ic可能导致环境变量加载异常，修改为直接调用conda的Python路径，并添加错误日志输出：

  */2 * * * * $CONDA_HOME/python $PYTHON_SCRIPT_PATH --time $IDLE_TIME_IN_SECONDS --region us-east-1 >> $LOG_FILE 2>&1
  

• 检查IAM权限：确保SageMaker执行角色拥有sagemaker:StopApp权限，否则脚本无法触发应用终止操作。
• 修复计时文件权限：执行sudo chown -R sagemaker-user:sagemaker-user /opt/amazon/sagemaker/sagemaker-code-editor-server-data/data/User/History，确保闲置计时文件可被正确读写更新。

内容的提问来源于stack exchange，提问作者Cyrus Mohammadian