I am checking via a bash script whether a process is running based on the PID stored in a file. I get the error shown above when the file does not exist as the command is trying to access the file and check whether the PID is really there, so it's normal, but I would like to avoid it if possible.
Is there such a way?
The command I am using can be found below.
if kill -0 $(cat "$pid_file")
Thank you in advance.
[[ -f "$pid_file ]] && (cat "$pid_file" | head -n 1 | tr -d "\n")
This command will return a non-zero status. If you want an "always return zero" (sacrificing legibility):
[[ ! -f "$pid_file ]] || (cat "$pid_file" | head -n 1 | tr -d "\n")
You need a condition to check whether the file exists, and if you don't want to see any errors when you attempt to kill a PID that no longer exists, redirect its stderr to /dev/null. You'll end up with something like:
PID_EXISTS=$([[ -f "$pid_file" ]] && kill -0 $(<"$pid_file") 2> /dev/null && echo 1 || echo 0) This will store 1 for an existing pid, and 0 for a non-existing one. You may use the variable in a comparison, such as:
if [[ $PID_EXISTS -eq 1 ]]; then echo FOUND; else echo NOTFOUND; fi Or just print its result:
echo $PID_EXISTS No cat. Meow!
sed... Even if low there is a probability that this PID before or after will be used again... Or, you can just use pgrep instead:
if pgrep -F "$pid_file" &>/dev/null; then echo yep else echo nope fi 
\nat the end of that file.cat "$pid_file" | head -n 1 | tr -d "\n"pgrep? check the-Fswitch.