Memtrace is a streaming client for OCaml's Memprof, which generates compact traces of a program's memory use. At Jane Street it has been used to find memory leaks in real systems, and it's one of the core tools we teach in our Performance analysis teach-in.
The Memtrace viewer analyzes the events produced by Memtrace and presents graphical views. Filters allow you to interactively narrow the view until the source of the memory problem becomes clear.
Our "Finding memory leaks with Memtrace" blog post walks through an example session.
Memtrace knows more about allocations than their creation: it can tell you, for a particular sample, when---if ever---it is collected, and filter on combinations of these data. This can be very helpful for finding leaks.
There is no special compiler or runtime configuration needed to collect a
trace of your program. You need only link in the memtrace library and add
a line or two of code. The library is available in OPAM, so you can install
it by running
$ opam install memtrace
and link it into your application by adding it to your dune file:
(executable
(name my_program)
(libraries base foo bar memtrace))
For most applications, it will suffice to add a call to Memtrace.trace_if_requested
wherever you want to start tracing. Typically this will be right at the beginning of your
startup code:
let _ =
Memtrace.trace_if_requested ~context:Version_util.version ();
...If the MEMTRACE environment variable is present, tracing begins to
the filename it specifies. (If it's absent, nothing happens.)
The ~context parameter is optional, and can be set to any string that
helps to identify the trace file.
If the program daemonises, the call to trace_if_requested should
occur after the program forks, to ensure the right process is
traced.