<?xml version="1.0" encoding="utf-8" ?>
<!DOCTYPE chapter SYSTEM "chapter.dtd">
<chapter>
<header>
<copyright>
<year>2001</year><year>2016</year>
<holder>Ericsson AB. All Rights Reserved.</holder>
</copyright>
<legalnotice>
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
</legalnotice>
<title>Profiling</title>
<prepared>Ingela Anderton</prepared>
<docno></docno>
<date>2001-11-02</date>
<rev></rev>
<file>profiling.xml</file>
</header>
<section>
<title>Do Not Guess About Performance - Profile</title>
<p>Even experienced software developers often guess wrong about where
the performance bottlenecks are in their programs. Therefore, profile
your program to see where the performance
bottlenecks are and concentrate on optimizing them.</p>
<p>Erlang/OTP contains several tools to help finding bottlenecks:</p>
<list type="bulleted">
<item><p><seealso marker="tools:fprof"><c>fprof</c></seealso> provides
the most detailed information about where the program time is spent,
but it significantly slows down the program it profiles.</p></item>
<item><p><seealso marker="tools:eprof"><c>eprof</c></seealso> provides
time information of each function used in the program. No call graph is
produced, but <c>eprof</c> has considerable less impact on the program it
profiles.</p>
<p>If the program is too large to be profiled by <c>fprof</c> or
<c>eprof</c>, <c>cprof</c> can be used to locate code parts that
are to be more thoroughly profiled using <c>fprof</c> or <c>eprof</c>.</p></item>
<item><p><seealso marker="tools:cprof"><c>cprof</c></seealso> is the
most lightweight tool, but it only provides execution counts on a
function basis (for all processes, not per process).</p></item>
<item><p><seealso marker="runtime_tools:dbg"><c>dbg</c></seealso> is the
generic erlang tracing frontend. By using the <c>timestamp</c> or
<c>cpu_timestamp</c> options it can be used to time how long function
calls in a live system take.</p></item>
<item><p><seealso marker="tools:lcnt"><c>lcnt</c></seealso> is used
to find contention points in the Erlang Run-Time System's internal
locking mechanisms. It is useful when looking for bottlenecks in
interaction between process, port, ets tables and other entities
that can be run in parallel.</p></item>
</list>
<p>The tools are further described in
<seealso marker="#profiling_tools">Tools</seealso>.</p>
<p>There are also several open source tools outside of Erlang/OTP
that can be used to help profiling. Some of them are:</p>
<list type="bulleted">
<item><url href="https://github.com/isacssouza/erlgrind">erlgrind</url>
can be used to visualize fprof data in kcachegrind.</item>
<item><url href="https://github.com/proger/eflame">eflame</url>
is an alternative to fprof that displays the profiling output as a flamegraph.</item>
<item><url href="https://ferd.github.io/recon/index.html">recon</url>
is a collection of Erlang profiling and debugging tools.
This tool comes with an accompanying E-book called
<url href="https://www.erlang-in-anger.com/">Erlang in Anger</url>.</item>
</list>
</section>
<section>
<title>Memory profiling</title>
<pre>eheap_alloc: Cannot allocate 1234567890 bytes of memory (of type "heap").</pre>
<p>The above slogan is one of the more common reasons for Erlang to terminate.
For unknown reasons the Erlang Run-Time System failed to allocate memory to
use. When this happens a crash dump is generated that contains information
about the state of the system as it ran out of mmeory. Use the
<seealso marker="observer:cdv"><c>crashdump_viewer</c></seealso> to get a
view of the memory is being used. Look for processes with large heaps or
many messages, large ets tables, etc.</p>
<p>When looking at memory usage in a running system the most basic function
to get information from is <seealso marker="erts:erlang#memory/0"><c>
erlang:memory()</c></seealso>. It returns the current memory usage
of the system. <seealso marker="tools:instrument"><c>instrument(3)</c></seealso>
can be used to get a more detailed breakdown of where memory is used.</p>
<p>Processes, ports and ets tables can then be inspecting using their
respective info functions, i.e.
<seealso marker="erts:erlang#process_info_memory"><c>erlang:process_info/2
</c></seealso>,
<seealso marker="erts:erlang#port_info_memory"><c>erlang:port_info/2
</c></seealso> and
<seealso marker="stdlib:ets#info/1"><c>ets:info/1</c></seealso>.
</p>
<p>Sometimes the system can enter a state where the reported memory
from <c>erlang:memory(total)</c> is very different from the
memory reported by the OS. This can be because of internal
fragmentation within the Erlang Run-Time System. Data about
how memory is allocated can be retrieved using
<seealso marker="erts:erlang#system_info_allocator">
<c>erlang:system_info(allocator)</c></seealso>.
The data you get from that function is very raw and not very plesant to read.
<url href="http://ferd.github.io/recon/recon_alloc.html">recon_alloc</url>
can be used to extract useful information from system_info
statistics counters.</p>
</section>
<section>
<title>Large Systems</title>
<p>For a large system, it can be interesting to run profiling
on a simulated and limited scenario to start with. But bottlenecks
have a tendency to appear or cause problems only when
many things are going on at the same time, and when
many nodes are involved. Therefore, it is also desirable to run
profiling in a system test plant on a real target system.</p>
<p>For a large system, you do not want to run the profiling
tools on the whole system. Instead you want to concentrate on
central processes and modules, which contribute for a big part
of the execution.</p>
<p>There are also some tools that can be used to get a view of the
whole system with more or less overhead.</p>
<list type="bulleted">
<item><seealso marker="observer:observer"><c>observer</c></seealso>
is a GUI tool that can connect to remote nodes and display a
variety of information about the running system.</item>
<item><seealso marker="observer:etop"><c>etop</c></seealso>
is a command line tool that can connect to remote nodes and
display information similar to what the UNIX tool top shows.</item>
<item><seealso marker="runtime_tools:msacc"><c>msacc</c></seealso>
allows the user to get a view of what the Erlang Run-Time system
is spending its time doing. Has a very low overhead, which makes it
useful to run in heavily loaded systems to get some idea of where
to start doing more granular profiling.</item>
</list>
</section>
<section>
<title>What to Look For</title>
<p>When analyzing the result file from the profiling activity,
look for functions that are called many
times and have a long "own" execution time (time excluding calls
to other functions). Functions that are called a lot of
times can also be interesting, as even small things can add
up to quite a bit if repeated often. Also
ask yourself what you can do to reduce this time. The following
are appropriate types of questions to ask yourself:</p>
<list type="bulleted">
<item>Is it possible to reduce the number of times the function
is called?</item>
<item>Can any test be run less often if the order of tests is
changed?</item>
<item>Can any redundant tests be removed?</item>
<item>Does any calculated expression give the same result
each time?</item>
<item>Are there other ways to do this that are equivalent and
more efficient?</item>
<item>Can another internal data representation be used to make
things more efficient?</item>
</list>
<p>These questions are not always trivial to answer. Some
benchmarks might be needed to back up your theory and to avoid
making things slower if your theory is wrong. For details, see
<seealso marker="#benchmark">Benchmarking</seealso>.</p>
</section>
<section>
<title>Tools</title>
<marker id="profiling_tools"></marker>
<section>
<title>fprof</title>
<p><c>fprof</c> measures the execution time for each function,
both own time, that is, how much time a function has used for its
own execution, and accumulated time, that is, including called
functions. The values are displayed per process. You also get
to know how many times each function has been called.</p>
<p><c>fprof</c> is based on trace to file to minimize runtime
performance impact. Using <c>fprof</c> is just a matter of
calling a few library functions, see the
<seealso marker="tools:fprof">fprof</seealso> manual page in
Tools.</p>
</section>
<section>
<title>eprof</title>
<p><c>eprof</c> is based on the Erlang <c>trace_info</c> BIFs.
<c>eprof</c> shows how much time has been used by each process,
and in which function calls this time has been spent. Time is
shown as percentage of total time and absolute time. For more
information, see the <seealso marker="tools:eprof">eprof</seealso>
manual page in Tools.</p>
</section>
<section>
<title>cprof</title>
<p><c>cprof</c> is something in between <c>fprof</c> and
<c>cover</c> regarding features. It counts how many times each
function is called when the program is run, on a per module
basis. <c>cprof</c> has a low performance degradation effect
(compared with <c>fprof</c>) and does not need to recompile
any modules to profile (compared with <c>cover</c>).
For more information, see the
<seealso marker="tools:cprof">cprof</seealso> manual page in
Tools.</p>
</section>
<section>
<title>Tool Summary</title>
<table>
<row>
<cell><em>Tool</em></cell>
<cell><em>Results</em></cell>
<cell><em>Size of Result</em></cell>
<cell><em>Effects on Program Execution Time</em></cell>
<cell><em>Records Number of Calls</em></cell>
<cell><em>Records Execution Time</em></cell>
<cell><em>Records Called by</em></cell>
<cell><em>Records Garbage Collection</em></cell>
</row>
<row>
<cell><c>fprof</c></cell>
<cell>Per process to screen/file</cell>
<cell>Large</cell>
<cell>Significant slowdown</cell>
<cell>Yes</cell>
<cell>Total and own</cell>
<cell>Yes</cell>
<cell>Yes</cell>
</row>
<row>
<cell><c>eprof</c></cell>
<cell>Per process/function to screen/file</cell>
<cell>Medium</cell>
<cell>Small slowdown</cell>
<cell>Yes</cell>
<cell>Only total</cell>
<cell>No</cell>
<cell>No</cell>
</row>
<row>
<cell><c>cprof</c></cell>
<cell>Per module to caller</cell>
<cell>Small</cell>
<cell>Small slowdown</cell>
<cell>Yes</cell>
<cell>No</cell>
<cell>No</cell>
<cell>No</cell>
</row>
<tcaption>Tool Summary</tcaption>
</table>
</section>
<section>
<title>dbg</title>
<p><c>dbg</c> is a generic Erlang trace tool. By using the
<c>timestamp</c> or <c>cpu_timestamp</c> options it can be used
as a precision instrument to profile how long time a function
call takes for a specific process. This can be very useful when
trying to understand where time is spent in a heavily loaded
system as it is possible to limit the scope of what is profiled
to be very small.
For more information, see the
<seealso marker="runtime_tools:dbg">dbg</seealso> manual page in
Runtime Tools.</p>
</section>
<section>
<title>lcnt</title>
<p><c>lcnt</c> is used to profile interactions inbetween
entities that run in parallel. For example if you have
a process that all other processes in the system needs
to interact with (maybe it has some global configuration),
then <c>lcnt</c> can be used to figure out if the interaction
with that process is a problem.</p>
<p>In the Erlang Run-time System entities are only run in parallel
when there are multiple schedulers. Therefore <c>lcnt</c> will
show more contention points (and thus be more useful) on systems
using many schedulers on many cores.</p>
<p>For more information, see the
<seealso marker="tools:lcnt">lcnt</seealso> manual page in Tools.</p>
</section>
</section>
<section>
<marker id="benchmark"></marker>
<title>Benchmarking</title>
<p>The main purpose of benchmarking is to find out which
implementation of a given algorithm or function is the fastest.
Benchmarking is far from an exact science. Today's operating systems
generally run background tasks that are difficult to turn off.
Caches and multiple CPU cores does not facilitate benchmarking.
It would be best to run UNIX computers in single-user mode when
benchmarking, but that is inconvenient to say the least for casual
testing.</p>
<p>Benchmarks can measure wall-clock time or CPU time.</p>
<list type="bulleted">
<item><seealso marker="stdlib:timer#tc/3">timer:tc/3</seealso> measures
wall-clock time. The advantage with wall-clock time is that I/O,
swapping, and other activities in the operating system kernel are
included in the measurements. The disadvantage is that the
measurements vary a lot. Usually it is best to run the
benchmark several times and note the shortest time, which is to
be the minimum time that is possible to achieve under the best of
circumstances.</item>
<item><seealso marker="erts:erlang#statistics/1">statistics/1</seealso>
with argument <c>runtime</c> measures CPU time spent in the Erlang
virtual machine. The advantage with CPU time is that the results are more
consistent from run to run. The disadvantage is that the time
spent in the operating system kernel (such as swapping and I/O)
is not included. Therefore, measuring CPU time is misleading if
any I/O (file or socket) is involved.</item>
</list>
<p>It is probably a good idea to do both wall-clock measurements and
CPU time measurements.</p>
<p>Some final advice:</p>
<list type="bulleted">
<item>The granularity of both measurement types can be high.
Therefore, ensure that each individual measurement
lasts for at least several seconds.</item>
<item>To make the test fair, each new test run is to run in its own,
newly created Erlang process. Otherwise, if all tests run in the
same process, the later tests start out with larger heap sizes
and therefore probably do fewer garbage collections.
Also consider restarting the Erlang emulator between each test.</item>
<item>Do not assume that the fastest implementation of a given algorithm
on computer architecture X is also the fastest on computer architecture
Y.</item>
</list>
</section>
</chapter>
