Peak memory usage of a linux/unix process

Question

[Edit: Works on Ubuntu 14.04: /usr/bin/time -v command Make sure to use the full path.]

Looks like /usr/bin/time does give you that info, if you pass -v (this is on Ubuntu 8.10). See, e.g., Maximum resident set size below:

$ /usr/bin/time -v ls /
....
        Command being timed: "ls /"
        User time (seconds): 0.00
        System time (seconds): 0.01
        Percent of CPU this job got: 250%
        Elapsed (wall clock) time (h:mm:ss or m:ss): 0:00.00
        Average shared text size (kbytes): 0
        Average unshared data size (kbytes): 0
        Average stack size (kbytes): 0
        Average total size (kbytes): 0
        Maximum resident set size (kbytes): 0
        Average resident set size (kbytes): 0
        Major (requiring I/O) page faults: 0
        Minor (reclaiming a frame) page faults: 315
        Voluntary context switches: 2
        Involuntary context switches: 0
        Swaps: 0
        File system inputs: 0
        File system outputs: 0
        Socket messages sent: 0
        Socket messages received: 0
        Signals delivered: 0
        Page size (bytes): 4096
        Exit status: 0

(This is an already answered, old question.. but just for the record :)

I was inspired by Yang's script, and came up with this small tool, named memusg. I simply increased the sampling rate to 0.1 to handle much short living processes. Instead of monitoring a single process, I made it measure rss sum of the process group. (Yeah, I write lots of separate programs that work together) It currently works on Mac OS X and Linux. The usage had to be similar to that of time:

memusg ls -alR / >/dev/null

It only shows the peak for the moment, but I'm interested in slight extensions for recording other (rough) statistics.

It's good to have such simple tool for just taking a look before we start any serious profiling.

Valgrind one-liner:

valgrind --tool=massif --pages-as-heap=yes --massif-out-file=massif.out ./test.sh; grep mem_heap_B massif.out | sed -e 's/mem_heap_B=\(.*\)/\1/' | sort -g | tail -n 1

Note use of --pages-as-heap to measure all memory in a process. More info here: http://valgrind.org/docs/manual/ms-manual.html

This will slow down your command significantly.

On Linux:

Use /usr/bin/time -v <program> <args> and look for "Maximum resident set size".

(Not to be confused with the Bash time built-in command! So use the full path, /usr/bin/time)

For example:

> /usr/bin/time -v ./myapp
        User time (seconds): 0.00
        . . .
        Maximum resident set size (kbytes): 2792
        . . .

On BSD, MacOS:

Use /usr/bin/time -l <program> <args>, looking for "maximum resident set size":

>/usr/bin/time -l ./myapp
        0.01 real         0.00 user         0.00 sys
      1440  maximum resident set size
      . . .

Here's a one-liner that doesn't require any external scripts or utilities and doesn't require you to start the process via another program like Valgrind or time, so you can use it for any process that's already running:

grep ^VmPeak /proc/$PID/status

(replace $PID with the PID of the process you're interested in)

Perhaps (gnu) time(1) already does what you want. For instance:

$ /usr/bin/time -f "%P %M" command
43% 821248

But other profiling tools may give more accurate results depending on what you are looking for.

/usr/bin/time maybe does what you want, actually. Something like.

 /usr/bin/time --format='(%Xtext+%Ddata %Mmax)'

See time(1) for details...