diff options
Diffstat (limited to 'erts/doc/src/erlang.xml')
| -rw-r--r-- | erts/doc/src/erlang.xml | 108 |
1 files changed, 82 insertions, 26 deletions
diff --git a/erts/doc/src/erlang.xml b/erts/doc/src/erlang.xml index fbe7b36163..8c438b0bd7 100644 --- a/erts/doc/src/erlang.xml +++ b/erts/doc/src/erlang.xml @@ -4881,6 +4881,7 @@ true</pre> <v>Type, Res -- see below</v> </type> <desc> + <p>All times are in milliseconds unless otherwise specified.</p> <p>Returns information about the system as specified by <c>Type</c>:</p> <taglist> @@ -4894,15 +4895,20 @@ true</pre> <item> <p>Returns <c>{Total_Exact_Reductions, Exact_Reductions_Since_Last_Call}</c>.</p> - <p><em>NOTE:</em><c>statistics(exact_reductions)</c> is - a more expensive operation than - <seealso marker="#statistics_reductions">statistics(reductions)</seealso> - especially on an Erlang machine with SMP support.</p> + <note><p><c>statistics(exact_reductions)</c> is + a more expensive operation than + <seealso marker="#statistics_reductions">statistics(reductions)</seealso> + especially on an Erlang machine with SMP support.</p> + </note> </item> <tag><c>garbage_collection</c></tag> <item> <p>Returns <c>{Number_of_GCs, Words_Reclaimed, 0}</c>. This information may not be valid for all implementations.</p> + <pre> +> <input>statistics(garbage_collection).</input> +{85,23961,0} +</pre> </item> <tag><c>io</c></tag> <item> @@ -4914,12 +4920,18 @@ true</pre> <tag><marker id="statistics_reductions"><c>reductions</c></marker></tag> <item> <p>Returns - <c>{Total_Reductions, Reductions_Since_Last_Call}</c>.</p> - <p><em>NOTE:</em> From erts version 5.5 (OTP release R11B) - this value does not include reductions performed in current - time slices of currently scheduled processes. If an - exact value is wanted, use - <seealso marker="#statistics_exact_reductions">statistics(exact_reductions)</seealso>.</p> + <c>{Total_Reductions, Reductions_Since_Last_Call}</c>.</p> + <note> + <p>From erts version 5.5 (OTP release R11B) + this value does not include reductions performed in current + time slices of currently scheduled processes. If an + exact value is wanted, use + <seealso marker="#statistics_exact_reductions">statistics(exact_reductions)</seealso>.</p> + </note> + <pre> +> <input>statistics(reductions).</input> +{2046,11} +</pre> </item> <tag><c>run_queue</c></tag> <item> @@ -4932,20 +4944,72 @@ true</pre> Note that the run-time is the sum of the run-time for all threads in the Erlang run-time system and may therefore be greater than the wall-clock time.</p> + <pre> +> <input>statistics(runtime).</input> +{1690,1620} +</pre> </item> <tag><marker id="statistics_scheduler_wall_time"><c>scheduler_wall_time</c></marker></tag> <item> - <p>Returns - <c>[{Scheduler_Id, Scheduler_Worked_Time, Scheduler_Total_Time}]</c>, time lapses are since the - the system flag <seealso marker="#system_flag_scheduler_wall_time">scheduler_wall_time</seealso> - was set to true. + <p>Returns a list of tuples with + <c>{SchedulerId, ActiveTime, TotalTime}</c>, where <c>SchedulerId</c> is an integer id of the scheduler, <c>ActiveTime</c> is + the duration the scheduler has been busy, <c>TotalTime</c> is the total time duration since + <seealso marker="#system_flag_scheduler_wall_time">scheduler_wall_time</seealso> + activation. The time unit is not defined and may be subject to change + between releases, operating systems and system restarts. + <c>scheduler_wall_time</c> should only be used to calculate relative + values for scheduler-utilization. <c>ActiveTime</c> can never exceed <c>TotalTime</c>. + </p> + + <p>The definition of a busy scheduler is when it is not idle or not + scheduling (selecting) a process or port, meaning; executing process + code, executing linked-in-driver or NIF code, executing + built-in-functions or any other runtime handling, garbage collecting + or handling any other memory management. Note, a scheduler may also be + busy even if the operating system has scheduled out the scheduler + thread. + </p> + + <p> Returns <c>undefined</c> if the system flag <seealso marker="#system_flag_scheduler_wall_time"> - scheduler_wall_time</seealso> is set to false. + scheduler_wall_time</seealso> is turned off. </p> - <p>The list of scheduler information is unsorted and may come in different order - between calls. The time unit is undefined and may be changed and should only be used - to calculate relative utilization. + + <p>The list of scheduler information is unsorted and may appear in different order + between calls. </p> + <p>Using <c>scheduler_wall_time</c> to calculate scheduler utilization.</p> +<pre> +> <input>erlang:system_flag(scheduler_wall_time, true).</input> +false +> <input>Ts0 = lists:sort(erlang:statistics(scheduler_wall_time)), ok.</input> +ok +</pre> + <p>Some time later we will take another snapshot and calculate scheduler-utilization per scheduler.</p> +<pre> +> <input>Ts1 = lists:sort(erlang:statistics(scheduler_wall_time)), ok.</input> +ok +> <input>lists:map(fun({{I, A0, T0}, {I, A1, T1}}) -> + {I, (A1 - A0)/(T1 - T0)} end, lists:zip(Ts0,Ts1)).</input> +[{1,0.9743474730177548}, + {2,0.9744843782751444}, + {3,0.9995902361669045}, + {4,0.9738012596572161}, + {5,0.9717956667018103}, + {6,0.9739235846420741}, + {7,0.973237033077876}, + {8,0.9741297293248656}] +</pre> + <p>Using the same snapshots to calculate a total scheduler-utilization.</p> +<pre> +> <input>{A, T} = lists:foldl(fun({{_, A0, T0}, {_, A1, T1}}, {Ai,Ti}) -> + {Ai + (A1 - A0), Ti + (T1 - T0)} end, {0, 0}, lists:zip(Ts0,Ts1)), A/T.</input> +0.9769136803764825 +</pre> + + <note> + <p><c>scheduler_wall_time</c> is by default disabled. Use <c>erlang:system_flag(scheduler_wall_time, true)</c> to enable it. </p> + </note> </item> <tag><c>wall_clock</c></tag> @@ -4957,14 +5021,6 @@ true</pre> opposed to runtime or CPU time.</p> </item> </taglist> - <p>All times are in milliseconds.</p> - <pre> -> <input>statistics(runtime).</input> -{1690,1620} -> <input>statistics(reductions).</input> -{2046,11} -> <input>statistics(garbage_collection).</input> -{85,23961,0}</pre> </desc> </func> <func> |