How to check if a daemon successfully started as part of a script?

Question

Let's say I have this script:

start daemon1 &
start daemon2 &
echo "Running..."

daemon2 can only be started if daemon1 was started successfully. if daemon1 did not start successfully, then the script most be aborted

"Running..." should be displayed only if daemon2 started successfully. if daemon2 did not start successfully, then the script most be aborted

How can I make this with a shell script ?

If `start daemonX` exits with a non zero value in failure, you can just do `start daemon1 && start daemon2 && echo Running` — Teshan Shanuka J, Mar 22 '20 at 11:17
Does the `start daemonX` command hang there after executing? Is that why you're using `&`? — Teshan Shanuka J, Mar 22 '20 at 11:24
I just tried this, it is not working. In fact, each daemonX blocks the terminal, that is why I run them in the background, and when I append &&, they are not run in the background — acmoune, Mar 22 '20 at 11:24
May be the `wait` command can help, but it looks like it returns when the daemon is shutdown, not when the daemon has completed its startup process. — acmoune, Mar 22 '20 at 11:31
Possible duplicate, https://stackoverflow.com/questions/1570262/get-exit-code-of-a-background-process — Jetchisel, Mar 22 '20 at 11:36

score 2 · Answer 1 · answered Mar 22 '20 at 11:38

You can check the PID of the started process to see if it is running

start daemon1 &
P=$!
if kill -0 $P > /dev/null 2>&1 ; then 
    start daemon2 &
    P=$!
    if kill -0 $P > /dev/null 2>&1 ; then
        echo "Running..."
    fi
fi

Untested code. Comment if something is not right

score 1 · Accepted Answer · answered Mar 22 '20 at 11:48

I propose you capture the daemon's pid (process id) and then determine if the pid exists (after some delay in case daemon1 takes a while to process to start and crash). So here is a way of achieving that (in Linux, I'm ignoring the 'start' in your commands since I'm not familiar with the windows cmdline environment ):

start daemon1 &
pid1=$!
sleep 3 # give daemon1 some time to get going 
if  
  kill -0 $pid1 2>/dev/null
then
  start daemon2 &
  pid2=$!
  sleep 3 # give daemon2 some time to get going 
  if
     kill -0 $pid2 2>/dev/null
  then
     echo "Running..."
  fi
fi

The necessary ingredients for this recipe are:

$! returns the child's pid (of the last background process run)
kill -0 <pid> is a way of determining if a pid is valid (in the process table)

I will go for that one, it makes the job. Thanks – acmoune Mar 22 '20 at 11:58 — acmoune, Mar 22 '20 at 11:58

How to check if a daemon successfully started as part of a script?

2 Answers2