Question

所以我有这些功能，是更大的bash脚本的一部分。问题是我的wait步骤抛出错误

pid 1不是该shell的子代

function getJobId(){
  local result=$(echo "$1" | sed s/^[^0-9]*// | sed s/[^0-9].*$//)
  echo $result
}

function deploy(){
  echo "Building now"
  nohup docker-compose build $1 &> logs/$2.txt &
  lastJob=$(getJobId $(jobs %%))
  echo "lastjob is --- $lastJob"
  jobsList="$jobsList $lastJob"
}

for job in $jobsList
do
  wait $job
  exit_code=$?
  if [[ "$exit_code" != 0 ]]; then
    handleError
  fi
done

我检查了lastJob，它包含一个有效值。 bash脚本从GCP ubuntu计算机的终端以root用户身份执行。

Answer 1

wait期望pid，而不是job id。您最简单（最短？）的更改代码（未经测试）似乎是：

function deploy(){
  echo "Building now"
  nohup docker-compose build $1 &> logs/$2.txt &
  # lastJob=$(getJobId $(jobs %%))
  # as you are exactly creating a new process here, you
  # already have the new spawned pid in the $! variable:
  lastJob=$! 
  echo "lastjob is --- $lastJob"
  jobsList="$jobsList $lastJob"
}

一个有趣的参考可能是this堆栈溢出问题。

重击：为后台进程的pid引发错误

1 个答案: