Member Avatar for santakdalai90

I am trying to search for a shell script if it is running and then assigning the count to a variable. Let the script be ScriptXYZ.sh. I wrote:

proc_count=$(/bin/ps -ef | grep ScriptXYZ.sh | grep -v grep | wc -l)

The problem that I am facing is that even when there is no instance of the ScriptXYZ.sh running I get a proc_count = 2. I tried running the command directly in the terminal and it yielded zero. But this variable always fetches value 2. I tried to put the quotes around it.

proc_count="$(/bin/ps -ef | grep ScriptXYZ.sh | grep -v grep | wc -l)"

But that also didn't change the value. Can anyone please explain why is this happening?

  • 3 Contributors
  • 3 Replies
  • 479 Views
  • 2 Days Discussion Span
  • Latest Post Latest Post by santakdalai90

Dani AI

Generated

Likely cause — running the check from inside the script changes what ps sees. When the script itself spawns the pipeline/subshell that does the search, the process table can contain both the running script and the helper process(es) used to evaluate the command substitution, so the pattern matches twice. Quoting the command substitution only affects shell word-splitting, not what ps reports.

Practical fixes and safer checks (examples are different from the ones already posted):

  • Avoid matching the search-process itself by using the "bracket trick" in grep:

    ps -ef | grep "[S]criptXYZ.sh" | wc -l

  • Use pgrep against the full command line (pgrep defaults to argv[0] only; that is why 's simple pgrep may miss the script when it is running under an interpreter). Example:

    pgrep -f "ScriptXYZ.sh" | wc -l

  • If the check runs inside the script and the goal is "how many other instances are running", exclude the current PID ($$). One compact approach:

    pgrep -f "ScriptXYZ.sh" | awk -v me=$$ '$1!=me{c++}END{print c+0}'

Design note: counting with ps/grep/pgrep is brittle (race conditions and substring matches). For reliable single-instance behavior, prefer an explicit lock/pidfile or flock. Example pattern:

exec 200>/var/run/ScriptXYZ.lock
flock -n 200 || exit 1
# now safe to run one instance

Tie-back: was right to run the pipeline interactively to see the difference in context; that diagnostic plus the suggestions above should explain why the same pipeline returns different numbers when run inside ScriptXYZ.sh versus from a terminal.

Recommended Answers

All 3 Replies

Member Avatar for rch1231

Hello,

The first thing to do is to run the command and see what the output is. When I run the following on my system I get the 0 you are looking for:

[rod@hpsrvr ~]$ /bin/ps -ef | grep ScriptXYZ.sh | grep -v grep | wc -l
0

There is no reason that this should not work.
There must be something else in you program that it does not like.
I used your example but checked for httpd instead of a script and th ran fine.

[rod@hpsrvr ~]$ cat test.sh
#!/bin/bash
myvar=$( /bin/ps -ef | grep httpd | grep -v grep | wc -l)
echo $myvar
[rod@hpsrvr ~]$ ./test.sh
12

Post your code and maybe we will see something.

Member Avatar for Watael

hi,

pgrep -c XYZ.sh

that's it! XD

Member Avatar for santakdalai90

Just to add some information that I am runnig it like:

>bash ScriptXYZ.sh [parameters]

and pgrep doesn't show the script as process.

Edited by santakdalai90
Be a part of the DaniWeb community

We're a friendly, industry-focused community of developers, IT pros, digital marketers, and technology enthusiasts meeting, networking, learning, and sharing knowledge.