Docker:来自守护程序的错误响应:rpc 错误:代码 = 2 desc = "oci runtime error: exec format error"

Question

我写了以下 docker 文件

FROM cloudera/quickstart

MAINTAINER abhishek "http://www.foobar.com"

ADD ./SparkIntegrationTestsAssembly.jar /
ADD ./entrypoint.sh /
ADD ./twitter.avro /

EXPOSE 8020 50070 50010 50020 50075 8030 8031 8032 8033 8088 8040 8042 10020 19888 11000 8888 18080 7077

RUN chmod +x /entrypoint.sh
ENTRYPOINT ["/entrypoint.sh"]

我使用命令构建了我的图像

docker build --tag foobar:auto .

这个命令的输出是

Sending build context to Docker daemon  93.1 MB
Step 1 : FROM cloudera/quickstart
 ---> 4239cd2958c6
Step 2 : MAINTAINER abhishek "http://www.foobar.com"
 ---> Running in 3ad11fe4aa77
 ---> 22a2f2840475
Removing intermediate container 3ad11fe4aa77
Step 3 : ADD ./SparkIntegrationTestsAssembly.jar /
 ---> 1ebae604e632
Removing intermediate container 0f047ec885a8
Step 4 : ADD ./entrypoint.sh /
 ---> 880cf4ff22aa
Removing intermediate container 0808ba44c97a
Step 5 : ADD ./twitter.avro /
 ---> 6978f2adf422
Removing intermediate container 43d812aaa3ae
Step 6 : EXPOSE 8020 50070 50010 50020 50075 8030 8031 8032 8033 8088 8040 8042 10020 19888 11000 8888 18080 7077
 ---> Running in af90e145f295
 ---> 6fcfb5ad934c
Removing intermediate container af90e145f295
Step 7 : RUN chmod +x /entrypoint.sh
 ---> Running in 4696faa2d330
 ---> 843ee5165937
Removing intermediate container 4696faa2d330
Step 8 : ENTRYPOINT /entrypoint.sh
 ---> Running in 4caf6e225007
 ---> 81cca7ee3198
Removing intermediate container 4caf6e225007
Successfully built 81cca7ee3198

但是当我尝试使用运行我的容器时

docker run --hostname=quickstart.cloudera --rm --privileged=true -t -i  -p "8020:8020" -p "50070:50070" -p "50010:50010" -p "50020:50020" -p "50075:50075" -p "8030:8030" -p "8031:8031" -p "8032:8032" -p "8033:8033" -p "8088:8088" -p "8040:8040" -p "8042:8042" -p "10020:10020" -p "19888:19888" -p "11000:11000" -p "8888:8888" -p "18080:18080" -p "7077:7077" foobar:auto

我收到一个错误

docker: Error response from daemon: rpc error: code = 2 desc = "oci runtime error: exec format error".

我的 entrypoint.sh 文件看起来像

/usr/bin/docker-quickstart
service hadoop-hdfs-namenode restart
hdfs dfs -mkdir -p input
hdfs dfs -put /twitter.avro /input/twitter.avro
spark-submit --class com.abhi.HelloWorld --master local[1] SparkIntegrationTestsAssembly.jar /input/twitter.avro /output

Answer 1

您是否发布了完整的 entrypoint.sh？内核尝试通过查看可执行文件的第一个字节来识别文件类型。对于脚本，您需要添加一个所谓的 shebang线。您可能需要在 entrypoint.sh 的最顶部添加一个 shebang 行，例如:

#!/bin/sh
/usr/bin/docker-quickstart
service hadoop-hdfs-namenode restart
hdfs dfs -mkdir -p input
hdfs dfs -put /twitter.avro /input/twitter.avro
spark-submit --class com.abhi.HelloWorld --master local[1] SparkIntegrationTestsAssembly.jar /input/twitter.avro /output