Add support for reading cpuacct usage metric

README.md

@@ -2,7 +2,7 @@ cg_mon
 ======
 
 Simple application to extend osmon functionality when using cgroups.
-Now it only supports reading memory cgroup metrics and can't automatically detect where cgroups mounted.
+Now it only supports reading memory and cpuacct cgroup metrics and can't automatically detect where cgroups mounted.
 
 Usage example.
 --------------
@@ -19,10 +19,12 @@ ok
 1941913600
 6> cg_mem_sup:usage().
 4880506880
+7> cg_cpu_sup:usage().
+8526844361646
 ```
 
 
 To be done:
-1. Add support for another cgroups such as cpuacct, blkio and so on.
+1. Add support for another cgroups such as blkio.
 2. Add event notifications (for example, about excessing some limit of memory usage).
 3. Add documentation.

src/cg_cpu_sup.erl

@@ -0,0 +1,33 @@
+
+-module(cg_cpu_sup).
+-author("Viacheslav V. Kovalev").
+
+%% API
+-export([
+    provided_resource/0,
+    key_value_sources/0,
+    single_value_sources/0
+]).
+
+-export([
+    usage/0
+]).
+
+
+-define(PROVIDED_RESOURCE, cpuacct).
+-define(KEY_VALUE_SOURCES, []).
+-define(SINGLE_VALUE_SOURCES, [
+    {usage, "cpuacct.usage"}
+]).
+
+provided_resource() ->
+    ?PROVIDED_RESOURCE.
+
+key_value_sources() ->
+    ?KEY_VALUE_SOURCES.
+
+single_value_sources() ->
+    ?SINGLE_VALUE_SOURCES.
+
+usage() ->
+    cg_mon_reader:read_meter(?PROVIDED_RESOURCE, usage).

src/cg_mon_app.erl

@@ -130,7 +130,7 @@ cgroup_discovery_file() ->
     end.
 
 handler_modules() ->
-    [cg_mem_sup].
+    [cg_mem_sup, cg_cpu_sup].
 
 
 %%%===================================================================