Timer utility class for things like timing how long a task takes, etc? Most of the searches on Google return results for timers that schedule threads and tasks, which is not what I want.
This question is related to
java
timing
execution-time
Performance measurements on my machine
System.nanoTime() : 750nsSystem.currentTimeMillis() : 18nsAs mentioned, System.nanoTime() is thought to measure elapsed time. Just be aware of the cost if used inside a loop or the like.
Use a profiler (JProfiler, Netbeans Profiler, Visual VM, Eclipse Profiler, etc). You'll get the most accurate results and is the least intrusive. They use the built-in JVM mechanism for profiling which can also give you extra information like stack traces, execution paths, and more comprehensive results if necessary.
When using a fully integrated profiler, it's faily trivial to profile a method. Right click, Profiler -> Add to Root Methods. Then run the profiler just like you were doing a test run or debugger.
A strategy that works to me in java ee was:
Create a class with a method annotated with @AroundInvoke;
@Singleton
public class TimedInterceptor implements Serializable {
@AroundInvoke
public Object logMethod(InvocationContext ic) throws Exception {
Date start = new Date();
Object result = ic.proceed();
Date end = new Date();
System.out.println("time: " + (end.getTime - start.getTime()));
return result;
}
}
Annotate the method that you want to monitoring:
@Interceptors(TimedInterceptor.class)
public void onMessage(final Message message) { ...
I hope this can help.
Just a small twist, if you don't use tooling and want to time methods with low execution time: execute it many times, each time doubling the number of times it is executed until you reach a second, or so. Thus, the time of the Call to System.nanoTime and so forth, nor the accuracy of System.nanoTime does affect the result much.
int runs = 0, runsPerRound = 10;
long begin = System.nanoTime(), end;
do {
for (int i=0; i<runsPerRound; ++i) timedMethod();
end = System.nanoTime();
runs += runsPerRound;
runsPerRound *= 2;
} while (runs < Integer.MAX_VALUE / 2 && 1000000000L > end - begin);
System.out.println("Time for timedMethod() is " +
0.000000001 * (end-begin) / runs + " seconds");
Of course, the caveats about using the wall clock apply: influences of JIT-compilation, multiple threads / processes etc. Thus, you need to first execute the method a lot of times first, such that the JIT compiler does its work, and then repeat this test multiple times and take the lowest execution time.
This probably isn't what you wanted me to say, but this is a good use of AOP. Whip an proxy interceptor around your method, and do the timing in there.
The what, why and how of AOP is rather beyond the scope of this answer, sadly, but that's how I'd likely do it.
Edit: Here's a link to Spring AOP to get you started, if you're keen. This is the most accessible implementation of AOP that Iive come across for java.
Also, given everyone else's very simple suggestions, I should add that AOP is for when you don't want stuff like timing to invade your code. But in many cases, that sort of simple and easy approach is fine.
I go with the simple answer. Works for me.
long startTime = System.currentTimeMillis();
doReallyLongThing();
long endTime = System.currentTimeMillis();
System.out.println("That took " + (endTime - startTime) + " milliseconds");
It works quite well. The resolution is obviously only to the millisecond, you can do better with System.nanoTime(). There are some limitations to both (operating system schedule slices, etc.) but this works pretty well.
Average across a couple of runs (the more the better) and you'll get a decent idea.
FYI, JEP 230: Microbenchmark Suite is an OpenJDK project to:
Add a basic suite of microbenchmarks to the JDK source code, and make it easy for developers to run existing microbenchmarks and create new ones.
This feature arrived in Java 12.
For earlier versions of Java, take a look at the Java Microbenchmark Harness (JMH) project on which JEP 230 is based.
long startTime = System.currentTimeMillis();
// code goes here
long finishTime = System.currentTimeMillis();
long elapsedTime = finishTime - startTime; // elapsed time in milliseconds
I basically do variations of this, but considering how hotspot compilation works, if you want to get accurate results you need to throw out the first few measurements and make sure you are using the method in a real world (read application specific) application.
If the JIT decides to compile it your numbers will vary heavily. so just be aware
This probably isn't what you wanted me to say, but this is a good use of AOP. Whip an proxy interceptor around your method, and do the timing in there.
The what, why and how of AOP is rather beyond the scope of this answer, sadly, but that's how I'd likely do it.
Edit: Here's a link to Spring AOP to get you started, if you're keen. This is the most accessible implementation of AOP that Iive come across for java.
Also, given everyone else's very simple suggestions, I should add that AOP is for when you don't want stuff like timing to invade your code. But in many cases, that sort of simple and easy approach is fine.
I implemented a simple timer, And I think it's really useful:
public class Timer{
private static long start_time;
public static double tic(){
return start_time = System.nanoTime();
}
public static double toc(){
return (System.nanoTime()-start_time)/1000000000.0;
}
}
That way you can time one or more actions:
Timer.tic();
// Code 1
System.out.println("Code 1 runtime: "+Timer.toc()+" seconds.");
// Code 2
System.out.println("(Code 1 + Code 2) runtime: "+Timer.toc()+"seconds");
Timer.tic();
// Code 3
System.out.println("Code 3 runtime: "+Timer.toc()+" seconds.");
You can use Metrics library which provides various measuring instruments. Add dependency:
<dependencies>
<dependency>
<groupId>io.dropwizard.metrics</groupId>
<artifactId>metrics-core</artifactId>
<version>${metrics.version}</version>
</dependency>
</dependencies>
And configure it for your environment.
Methods can be annotated with @Timed:
@Timed
public void exampleMethod(){
// some code
}
or piece of code wrapped with Timer:
final Timer timer = metricsRegistry.timer("some_name");
final Timer.Context context = timer.time();
// timed code
context.stop();
Aggregated metrics can exported to console, JMX, CSV or other.
@Timed metrics output example:
com.example.ExampleService.exampleMethod
count = 2
mean rate = 3.11 calls/minute
1-minute rate = 0.96 calls/minute
5-minute rate = 0.20 calls/minute
15-minute rate = 0.07 calls/minute
min = 17.01 milliseconds
max = 1006.68 milliseconds
mean = 511.84 milliseconds
stddev = 699.80 milliseconds
median = 511.84 milliseconds
75% <= 1006.68 milliseconds
95% <= 1006.68 milliseconds
98% <= 1006.68 milliseconds
99% <= 1006.68 milliseconds
99.9% <= 1006.68 milliseconds
System.currentTimeMillis(); IS NOT a good approach for measuring the performance of your algorithms. It measures the total time you experience as a user watching the computer screen. It includes also time consumed by everything else running on your computer in the background. This could make a huge difference in case you have a lot of programs running on your workstation.
Proper approach is using java.lang.management package.
From http://nadeausoftware.com/articles/2008/03/java_tip_how_get_cpu_and_user_time_benchmarking website (archive link):
getCpuTime() method gives you sum of those:
import java.lang.management.ManagementFactory;
import java.lang.management.ThreadMXBean;
public class CPUUtils {
/** Get CPU time in nanoseconds. */
public static long getCpuTime( ) {
ThreadMXBean bean = ManagementFactory.getThreadMXBean( );
return bean.isCurrentThreadCpuTimeSupported( ) ?
bean.getCurrentThreadCpuTime( ) : 0L;
}
/** Get user time in nanoseconds. */
public static long getUserTime( ) {
ThreadMXBean bean = ManagementFactory.getThreadMXBean( );
return bean.isCurrentThreadCpuTimeSupported( ) ?
bean.getCurrentThreadUserTime( ) : 0L;
}
/** Get system time in nanoseconds. */
public static long getSystemTime( ) {
ThreadMXBean bean = ManagementFactory.getThreadMXBean( );
return bean.isCurrentThreadCpuTimeSupported( ) ?
(bean.getCurrentThreadCpuTime( ) - bean.getCurrentThreadUserTime( )) : 0L;
}
}
Here is pretty printed string ready formated seconds elapsed similar to google search time took to search:
long startTime = System.nanoTime();
// ... methodToTime();
long endTime = System.nanoTime();
long duration = (endTime - startTime);
long seconds = (duration / 1000) % 60;
// formatedSeconds = (0.xy seconds)
String formatedSeconds = String.format("(0.%d seconds)", seconds);
System.out.println("formatedSeconds = "+ formatedSeconds);
// i.e actual formatedSeconds = (0.52 seconds)
If you want wall-clock time
long start_time = System.currentTimeMillis();
object.method();
long end_time = System.currentTimeMillis();
long execution_time = end_time - start_time;
I go with the simple answer. Works for me.
long startTime = System.currentTimeMillis();
doReallyLongThing();
long endTime = System.currentTimeMillis();
System.out.println("That took " + (endTime - startTime) + " milliseconds");
It works quite well. The resolution is obviously only to the millisecond, you can do better with System.nanoTime(). There are some limitations to both (operating system schedule slices, etc.) but this works pretty well.
Average across a couple of runs (the more the better) and you'll get a decent idea.
I basically do variations of this, but considering how hotspot compilation works, if you want to get accurate results you need to throw out the first few measurements and make sure you are using the method in a real world (read application specific) application.
If the JIT decides to compile it your numbers will vary heavily. so just be aware
Just a small twist, if you don't use tooling and want to time methods with low execution time: execute it many times, each time doubling the number of times it is executed until you reach a second, or so. Thus, the time of the Call to System.nanoTime and so forth, nor the accuracy of System.nanoTime does affect the result much.
int runs = 0, runsPerRound = 10;
long begin = System.nanoTime(), end;
do {
for (int i=0; i<runsPerRound; ++i) timedMethod();
end = System.nanoTime();
runs += runsPerRound;
runsPerRound *= 2;
} while (runs < Integer.MAX_VALUE / 2 && 1000000000L > end - begin);
System.out.println("Time for timedMethod() is " +
0.000000001 * (end-begin) / runs + " seconds");
Of course, the caveats about using the wall clock apply: influences of JIT-compilation, multiple threads / processes etc. Thus, you need to first execute the method a lot of times first, such that the JIT compiler does its work, and then repeat this test multiple times and take the lowest execution time.
I basically do variations of this, but considering how hotspot compilation works, if you want to get accurate results you need to throw out the first few measurements and make sure you are using the method in a real world (read application specific) application.
If the JIT decides to compile it your numbers will vary heavily. so just be aware
Spring provides a utility class org.springframework.util.StopWatch, as per JavaDoc:
Simple stop watch, allowing for timing of a number of tasks, exposing total running time and running time for each named task.
Usage:
StopWatch stopWatch = new StopWatch("Performance Test Result");
stopWatch.start("Method 1");
doSomething1();//method to test
stopWatch.stop();
stopWatch.start("Method 2");
doSomething2();//method to test
stopWatch.stop();
System.out.println(stopWatch.prettyPrint());
Output:
StopWatch 'Performance Test Result': running time (millis) = 12829
-----------------------------------------
ms % Task name
-----------------------------------------
11907 036% Method 1
00922 064% Method 2
With Aspects:
@Around("execution(* my.package..*.*(..))")
public Object logTime(ProceedingJoinPoint joinPoint) throws Throwable {
StopWatch stopWatch = new StopWatch();
stopWatch.start();
Object retVal = joinPoint.proceed();
stopWatch.stop();
log.info(" execution time: " + stopWatch.getTotalTimeMillis() + " ms");
return retVal;
}
I basically do variations of this, but considering how hotspot compilation works, if you want to get accurate results you need to throw out the first few measurements and make sure you are using the method in a real world (read application specific) application.
If the JIT decides to compile it your numbers will vary heavily. so just be aware
System.nanoTime() is a pretty precise system utility to measure execution time. But be careful, if you're running on pre-emptive scheduler mode (default), this utility actually measures wall-clock time and not CPU time. Therefore, you may notice different execution time values from run to run, depending on system loads. If you look for CPU time, I think that running your program in real-time mode will do the trick. You have to use RT linux. link: Real-time programming with Linux
Use a profiler (JProfiler, Netbeans Profiler, Visual VM, Eclipse Profiler, etc). You'll get the most accurate results and is the least intrusive. They use the built-in JVM mechanism for profiling which can also give you extra information like stack traces, execution paths, and more comprehensive results if necessary.
When using a fully integrated profiler, it's faily trivial to profile a method. Right click, Profiler -> Add to Root Methods. Then run the profiler just like you were doing a test run or debugger.
As "skaffman" said, use AOP OR you can use run time bytecode weaving, just like unit test method coverage tools use to transparently add timing info to methods invoked.
You can look at code used by open source tools tools like Emma (http://downloads.sourceforge.net/emma/emma-2.0.5312-src.zip?modtime=1118607545&big_mirror=0). The other opensource coverage tool is http://prdownloads.sourceforge.net/cobertura/cobertura-1.9-src.zip?download.
If you eventually manage to do what you set out for, pls. share it back with the community here with your ant task/jars.
Gathered all possible ways together into one place.
Date startDate = Calendar.getInstance().getTime();
long d_StartTime = new Date().getTime();
Thread.sleep(1000 * 4);
Date endDate = Calendar.getInstance().getTime();
long d_endTime = new Date().getTime();
System.out.format("StartDate : %s, EndDate : %s \n", startDate, endDate);
System.out.format("Milli = %s, ( D_Start : %s, D_End : %s ) \n", (d_endTime - d_StartTime),d_StartTime, d_endTime);
System.currentTimeMillis()
long startTime = System.currentTimeMillis();
Thread.sleep(1000 * 4);
long endTime = System.currentTimeMillis();
long duration = (endTime - startTime);
System.out.format("Milli = %s, ( S_Start : %s, S_End : %s ) \n", duration, startTime, endTime );
System.out.println("Human-Readable format : "+millisToShortDHMS( duration ) );
Human Readable Format
public static String millisToShortDHMS(long duration) {
String res = ""; // java.util.concurrent.TimeUnit;
long days = TimeUnit.MILLISECONDS.toDays(duration);
long hours = TimeUnit.MILLISECONDS.toHours(duration) -
TimeUnit.DAYS.toHours(TimeUnit.MILLISECONDS.toDays(duration));
long minutes = TimeUnit.MILLISECONDS.toMinutes(duration) -
TimeUnit.HOURS.toMinutes(TimeUnit.MILLISECONDS.toHours(duration));
long seconds = TimeUnit.MILLISECONDS.toSeconds(duration) -
TimeUnit.MINUTES.toSeconds(TimeUnit.MILLISECONDS.toMinutes(duration));
long millis = TimeUnit.MILLISECONDS.toMillis(duration) -
TimeUnit.SECONDS.toMillis(TimeUnit.MILLISECONDS.toSeconds(duration));
if (days == 0) res = String.format("%02d:%02d:%02d.%04d", hours, minutes, seconds, millis);
else res = String.format("%dd %02d:%02d:%02d.%04d", days, hours, minutes, seconds, millis);
return res;
}
Guava: Google StopwatchJAR « An object of Stopwatch is to measures elapsed time in nanoseconds.
com.google.common.base.Stopwatch g_SW = Stopwatch.createUnstarted();
g_SW.start();
Thread.sleep(1000 * 4);
g_SW.stop();
System.out.println("Google StopWatch : "+g_SW);
Apache Commons LangJAR « StopWatch provides a convenient API for timings.
org.apache.commons.lang3.time.StopWatch sw = new StopWatch();
sw.start();
Thread.sleep(1000 * 4);
sw.stop();
System.out.println("Apache StopWatch : "+ millisToShortDHMS(sw.getTime()) );
JODA-TIME
public static void jodaTime() throws InterruptedException, ParseException{
java.text.SimpleDateFormat ms_SDF = new SimpleDateFormat("yyyy/MM/dd HH:mm:ss.SSS");
String start = ms_SDF.format( new Date() ); // java.util.Date
Thread.sleep(10000);
String end = ms_SDF.format( new Date() );
System.out.println("Start:"+start+"\t Stop:"+end);
Date date_1 = ms_SDF.parse(start);
Date date_2 = ms_SDF.parse(end);
Interval interval = new org.joda.time.Interval( date_1.getTime(), date_2.getTime() );
Period period = interval.toPeriod(); //org.joda.time.Period
System.out.format("%dY/%dM/%dD, %02d:%02d:%02d.%04d \n",
period.getYears(), period.getMonths(), period.getDays(),
period.getHours(), period.getMinutes(), period.getSeconds(), period.getMillis());
}
Java date time API from Java 8 « A Duration object represents a period of time between two Instant objects.
Instant start = java.time.Instant.now();
Thread.sleep(1000);
Instant end = java.time.Instant.now();
Duration between = java.time.Duration.between(start, end);
System.out.println( between ); // PT1.001S
System.out.format("%dD, %02d:%02d:%02d.%04d \n", between.toDays(),
between.toHours(), between.toMinutes(), between.getSeconds(), between.toMillis()); // 0D, 00:00:01.1001
Spring Framework provides StopWatch utility class to measure elapsed time in Java.
StopWatch sw = new org.springframework.util.StopWatch();
sw.start("Method-1"); // Start a named task
Thread.sleep(500);
sw.stop();
sw.start("Method-2");
Thread.sleep(300);
sw.stop();
sw.start("Method-3");
Thread.sleep(200);
sw.stop();
System.out.println("Total time in milliseconds for all tasks :\n"+sw.getTotalTimeMillis());
System.out.println("Table describing all tasks performed :\n"+sw.prettyPrint());
System.out.format("Time taken by the last task : [%s]:[%d]",
sw.getLastTaskName(),sw.getLastTaskTimeMillis());
System.out.println("\n Array of the data for tasks performed « Task Name: Time Taken");
TaskInfo[] listofTasks = sw.getTaskInfo();
for (TaskInfo task : listofTasks) {
System.out.format("[%s]:[%d]\n",
task.getTaskName(), task.getTimeMillis());
}
OutPut:
Total time in milliseconds for all tasks :
999
Table describing all tasks performed :
StopWatch '': running time (millis) = 999
-----------------------------------------
ms % Task name
-----------------------------------------
00500 050% Method-1
00299 030% Method-2
00200 020% Method-3
Time taken by the last task : [Method-3]:[200]
Array of the data for tasks performed « Task Name: Time Taken
[Method-1]:[500]
[Method-2]:[299]
[Method-3]:[200]
You can use javaagent to modify the java class bytes ,add the monitor codes dynamically.there is some open source tools on the github that can do this for you.
If you want to do it by yourself, just implements the javaagent,use javassist to modify the methods you want to monitor,and the monitor code before your method return.it's clean and you can monitor systems that you even don't have source code.
There are a couple of ways to do that. I normally fall back to just using something like this:
long start = System.currentTimeMillis();
// ... do something ...
long end = System.currentTimeMillis();
or the same thing with System.nanoTime();
For something more on the benchmarking side of things there seems also to be this one: http://jetm.void.fm/ Never tried it though.
Come on guys! Nobody mentioned the Guava way to do that (which is arguably awesome):
import com.google.common.base.Stopwatch;
Stopwatch timer = Stopwatch.createStarted();
//method invocation
LOG.info("Method took: " + timer.stop());
The nice thing is that Stopwatch.toString() does a good job of selecting time units for the measurement. I.e. if the value is small, it'll output 38 ns, if it's long, it'll show 5m 3s
Even nicer:
Stopwatch timer = Stopwatch.createUnstarted();
for (...) {
timer.start();
methodToTrackTimeFor();
timer.stop();
methodNotToTrackTimeFor();
}
LOG.info("Method took: " + timer);
Note: Google Guava requires Java 1.6+
If you want wall-clock time
long start_time = System.currentTimeMillis();
object.method();
long end_time = System.currentTimeMillis();
long execution_time = end_time - start_time;
With Java 8 you can do also something like this with every normal methods:
Object returnValue = TimeIt.printTime(() -> methodeWithReturnValue());
//do stuff with your returnValue
with TimeIt like:
public class TimeIt {
public static <T> T printTime(Callable<T> task) {
T call = null;
try {
long startTime = System.currentTimeMillis();
call = task.call();
System.out.print((System.currentTimeMillis() - startTime) / 1000d + "s");
} catch (Exception e) {
//...
}
return call;
}
}
With this methode you can make easy time measurement anywhere in your code without breaking it. In this simple example i just print the time. May you add a Switch for TimeIt, e.g. to only print the time in DebugMode or something.
If you are working with Function you can do somthing like this:
Function<Integer, Integer> yourFunction= (n) -> {
return IntStream.range(0, n).reduce(0, (a, b) -> a + b);
};
Integer returnValue = TimeIt.printTime2(yourFunction).apply(10000);
//do stuff with your returnValue
public static <T, R> Function<T, R> printTime2(Function<T, R> task) {
return (t) -> {
long startTime = System.currentTimeMillis();
R apply = task.apply(t);
System.out.print((System.currentTimeMillis() - startTime) / 1000d
+ "s");
return apply;
};
}
Just a small twist, if you don't use tooling and want to time methods with low execution time: execute it many times, each time doubling the number of times it is executed until you reach a second, or so. Thus, the time of the Call to System.nanoTime and so forth, nor the accuracy of System.nanoTime does affect the result much.
int runs = 0, runsPerRound = 10;
long begin = System.nanoTime(), end;
do {
for (int i=0; i<runsPerRound; ++i) timedMethod();
end = System.nanoTime();
runs += runsPerRound;
runsPerRound *= 2;
} while (runs < Integer.MAX_VALUE / 2 && 1000000000L > end - begin);
System.out.println("Time for timedMethod() is " +
0.000000001 * (end-begin) / runs + " seconds");
Of course, the caveats about using the wall clock apply: influences of JIT-compilation, multiple threads / processes etc. Thus, you need to first execute the method a lot of times first, such that the JIT compiler does its work, and then repeat this test multiple times and take the lowest execution time.
Also We can use StopWatch class of Apache commons for measuring the time.
Sample code
org.apache.commons.lang.time.StopWatch sw = new org.apache.commons.lang.time.StopWatch();
System.out.println("getEventFilterTreeData :: Start Time : " + sw.getTime());
sw.start();
// Method execution code
sw.stop();
System.out.println("getEventFilterTreeData :: End Time : " + sw.getTime());
You can try this way if just want know the time.
long startTime = System.currentTimeMillis();
//@ Method call
System.out.println("Total time [ms]: " + (System.currentTimeMillis() - startTime));
Ok, this is a simple class to be used for simple simple timing of your functions. There is an example below it.
public class Stopwatch {
static long startTime;
static long splitTime;
static long endTime;
public Stopwatch() {
start();
}
public void start() {
startTime = System.currentTimeMillis();
splitTime = System.currentTimeMillis();
endTime = System.currentTimeMillis();
}
public void split() {
split("");
}
public void split(String tag) {
endTime = System.currentTimeMillis();
System.out.println("Split time for [" + tag + "]: " + (endTime - splitTime) + " ms");
splitTime = endTime;
}
public void end() {
end("");
}
public void end(String tag) {
endTime = System.currentTimeMillis();
System.out.println("Final time for [" + tag + "]: " + (endTime - startTime) + " ms");
}
}
Sample of use:
public static Schedule getSchedule(Activity activity_context) {
String scheduleJson = null;
Schedule schedule = null;
/*->*/ Stopwatch stopwatch = new Stopwatch();
InputStream scheduleJsonInputStream = activity_context.getResources().openRawResource(R.raw.skating_times);
/*->*/ stopwatch.split("open raw resource");
scheduleJson = FileToString.convertStreamToString(scheduleJsonInputStream);
/*->*/ stopwatch.split("file to string");
schedule = new Gson().fromJson(scheduleJson, Schedule.class);
/*->*/ stopwatch.split("parse Json");
/*->*/ stopwatch.end("Method getSchedule");
return schedule;
}
Sample of console output:
Split time for [file to string]: 672 ms
Split time for [parse Json]: 893 ms
Final time for [get Schedule]: 1565 ms
I have written a method to print the method execution time in a much readable form. For example, to calculate the factorial of 1 Million, it takes approximately 9 minutes. So the execution time get printed as:
Execution Time: 9 Minutes, 36 Seconds, 237 MicroSeconds, 806193 NanoSeconds
The code is here:
public class series
{
public static void main(String[] args)
{
long startTime = System.nanoTime();
long n = 10_00_000;
printFactorial(n);
long endTime = System.nanoTime();
printExecutionTime(startTime, endTime);
}
public static void printExecutionTime(long startTime, long endTime)
{
long time_ns = endTime - startTime;
long time_ms = TimeUnit.NANOSECONDS.toMillis(time_ns);
long time_sec = TimeUnit.NANOSECONDS.toSeconds(time_ns);
long time_min = TimeUnit.NANOSECONDS.toMinutes(time_ns);
long time_hour = TimeUnit.NANOSECONDS.toHours(time_ns);
System.out.print("\nExecution Time: ");
if(time_hour > 0)
System.out.print(time_hour + " Hours, ");
if(time_min > 0)
System.out.print(time_min % 60 + " Minutes, ");
if(time_sec > 0)
System.out.print(time_sec % 60 + " Seconds, ");
if(time_ms > 0)
System.out.print(time_ms % 1E+3 + " MicroSeconds, ");
if(time_ns > 0)
System.out.print(time_ns % 1E+6 + " NanoSeconds");
}
}
We are using AspectJ and Java annotations for this purpose. If we need to know to execution time for a method, we simple annotate it. A more advanced version could use an own log level that can enabled and disabled at runtime.
public @interface Trace {
boolean showParameters();
}
@Aspect
public class TraceAspect {
[...]
@Around("tracePointcut() && @annotation(trace) && !within(TraceAspect)")
public Object traceAdvice ( ProceedingJintPoint jP, Trace trace ) {
Object result;
// initilize timer
try {
result = jp.procced();
} finally {
// calculate execution time
}
return result;
}
[...]
}
Spring provides a utility class org.springframework.util.StopWatch, as per JavaDoc:
Simple stop watch, allowing for timing of a number of tasks, exposing total running time and running time for each named task.
Usage:
StopWatch stopWatch = new StopWatch("Performance Test Result");
stopWatch.start("Method 1");
doSomething1();//method to test
stopWatch.stop();
stopWatch.start("Method 2");
doSomething2();//method to test
stopWatch.stop();
System.out.println(stopWatch.prettyPrint());
Output:
StopWatch 'Performance Test Result': running time (millis) = 12829
-----------------------------------------
ms % Task name
-----------------------------------------
11907 036% Method 1
00922 064% Method 2
With Aspects:
@Around("execution(* my.package..*.*(..))")
public Object logTime(ProceedingJoinPoint joinPoint) throws Throwable {
StopWatch stopWatch = new StopWatch();
stopWatch.start();
Object retVal = joinPoint.proceed();
stopWatch.stop();
log.info(" execution time: " + stopWatch.getTotalTimeMillis() + " ms");
return retVal;
}
Just a small twist, if you don't use tooling and want to time methods with low execution time: execute it many times, each time doubling the number of times it is executed until you reach a second, or so. Thus, the time of the Call to System.nanoTime and so forth, nor the accuracy of System.nanoTime does affect the result much.
int runs = 0, runsPerRound = 10;
long begin = System.nanoTime(), end;
do {
for (int i=0; i<runsPerRound; ++i) timedMethod();
end = System.nanoTime();
runs += runsPerRound;
runsPerRound *= 2;
} while (runs < Integer.MAX_VALUE / 2 && 1000000000L > end - begin);
System.out.println("Time for timedMethod() is " +
0.000000001 * (end-begin) / runs + " seconds");
Of course, the caveats about using the wall clock apply: influences of JIT-compilation, multiple threads / processes etc. Thus, you need to first execute the method a lot of times first, such that the JIT compiler does its work, and then repeat this test multiple times and take the lowest execution time.
You can use Perf4j. Very cool utility. Usage is simple
String watchTag = "target.SomeMethod";
StopWatch stopWatch = new LoggingStopWatch(watchTag);
Result result = null; // Result is a type of a return value of a method
try {
result = target.SomeMethod();
stopWatch.stop(watchTag + ".success");
} catch (Exception e) {
stopWatch.stop(watchTag + ".fail", "Exception was " + e);
throw e;
}
More information can be found in Developer Guide
Edit: Project seems dead
In Java 8 a new class named Instant is introduced. As per doc:
Instant represents the start of a nanosecond on the time line. This class is useful for generating a time stamp to represent machine time. The range of an instant requires the storage of a number larger than a long. To achieve this, the class stores a long representing epoch-seconds and an int representing nanosecond-of-second, which will always be between 0 and 999,999,999. The epoch-seconds are measured from the standard Java epoch of 1970-01-01T00:00:00Z where instants after the epoch have positive values, and earlier instants have negative values. For both the epoch-second and nanosecond parts, a larger value is always later on the time-line than a smaller value.
This can be used as:
Instant start = Instant.now();
try {
Thread.sleep(7000);
} catch (InterruptedException e) {
e.printStackTrace();
}
Instant end = Instant.now();
System.out.println(Duration.between(start, end));
It prints PT7.001S.
long startTime = System.currentTimeMillis();
// code goes here
long finishTime = System.currentTimeMillis();
long elapsedTime = finishTime - startTime; // elapsed time in milliseconds
In Java 8 a new class named Instant is introduced. As per doc:
Instant represents the start of a nanosecond on the time line. This class is useful for generating a time stamp to represent machine time. The range of an instant requires the storage of a number larger than a long. To achieve this, the class stores a long representing epoch-seconds and an int representing nanosecond-of-second, which will always be between 0 and 999,999,999. The epoch-seconds are measured from the standard Java epoch of 1970-01-01T00:00:00Z where instants after the epoch have positive values, and earlier instants have negative values. For both the epoch-second and nanosecond parts, a larger value is always later on the time-line than a smaller value.
This can be used as:
Instant start = Instant.now();
try {
Thread.sleep(7000);
} catch (InterruptedException e) {
e.printStackTrace();
}
Instant end = Instant.now();
System.out.println(Duration.between(start, end));
It prints PT7.001S.
If you want wall-clock time
long start_time = System.currentTimeMillis();
object.method();
long end_time = System.currentTimeMillis();
long execution_time = end_time - start_time;
Gathered all possible ways together into one place.
Date startDate = Calendar.getInstance().getTime();
long d_StartTime = new Date().getTime();
Thread.sleep(1000 * 4);
Date endDate = Calendar.getInstance().getTime();
long d_endTime = new Date().getTime();
System.out.format("StartDate : %s, EndDate : %s \n", startDate, endDate);
System.out.format("Milli = %s, ( D_Start : %s, D_End : %s ) \n", (d_endTime - d_StartTime),d_StartTime, d_endTime);
System.currentTimeMillis()
long startTime = System.currentTimeMillis();
Thread.sleep(1000 * 4);
long endTime = System.currentTimeMillis();
long duration = (endTime - startTime);
System.out.format("Milli = %s, ( S_Start : %s, S_End : %s ) \n", duration, startTime, endTime );
System.out.println("Human-Readable format : "+millisToShortDHMS( duration ) );
Human Readable Format
public static String millisToShortDHMS(long duration) {
String res = ""; // java.util.concurrent.TimeUnit;
long days = TimeUnit.MILLISECONDS.toDays(duration);
long hours = TimeUnit.MILLISECONDS.toHours(duration) -
TimeUnit.DAYS.toHours(TimeUnit.MILLISECONDS.toDays(duration));
long minutes = TimeUnit.MILLISECONDS.toMinutes(duration) -
TimeUnit.HOURS.toMinutes(TimeUnit.MILLISECONDS.toHours(duration));
long seconds = TimeUnit.MILLISECONDS.toSeconds(duration) -
TimeUnit.MINUTES.toSeconds(TimeUnit.MILLISECONDS.toMinutes(duration));
long millis = TimeUnit.MILLISECONDS.toMillis(duration) -
TimeUnit.SECONDS.toMillis(TimeUnit.MILLISECONDS.toSeconds(duration));
if (days == 0) res = String.format("%02d:%02d:%02d.%04d", hours, minutes, seconds, millis);
else res = String.format("%dd %02d:%02d:%02d.%04d", days, hours, minutes, seconds, millis);
return res;
}
Guava: Google StopwatchJAR « An object of Stopwatch is to measures elapsed time in nanoseconds.
com.google.common.base.Stopwatch g_SW = Stopwatch.createUnstarted();
g_SW.start();
Thread.sleep(1000 * 4);
g_SW.stop();
System.out.println("Google StopWatch : "+g_SW);
Apache Commons LangJAR « StopWatch provides a convenient API for timings.
org.apache.commons.lang3.time.StopWatch sw = new StopWatch();
sw.start();
Thread.sleep(1000 * 4);
sw.stop();
System.out.println("Apache StopWatch : "+ millisToShortDHMS(sw.getTime()) );
JODA-TIME
public static void jodaTime() throws InterruptedException, ParseException{
java.text.SimpleDateFormat ms_SDF = new SimpleDateFormat("yyyy/MM/dd HH:mm:ss.SSS");
String start = ms_SDF.format( new Date() ); // java.util.Date
Thread.sleep(10000);
String end = ms_SDF.format( new Date() );
System.out.println("Start:"+start+"\t Stop:"+end);
Date date_1 = ms_SDF.parse(start);
Date date_2 = ms_SDF.parse(end);
Interval interval = new org.joda.time.Interval( date_1.getTime(), date_2.getTime() );
Period period = interval.toPeriod(); //org.joda.time.Period
System.out.format("%dY/%dM/%dD, %02d:%02d:%02d.%04d \n",
period.getYears(), period.getMonths(), period.getDays(),
period.getHours(), period.getMinutes(), period.getSeconds(), period.getMillis());
}
Java date time API from Java 8 « A Duration object represents a period of time between two Instant objects.
Instant start = java.time.Instant.now();
Thread.sleep(1000);
Instant end = java.time.Instant.now();
Duration between = java.time.Duration.between(start, end);
System.out.println( between ); // PT1.001S
System.out.format("%dD, %02d:%02d:%02d.%04d \n", between.toDays(),
between.toHours(), between.toMinutes(), between.getSeconds(), between.toMillis()); // 0D, 00:00:01.1001
Spring Framework provides StopWatch utility class to measure elapsed time in Java.
StopWatch sw = new org.springframework.util.StopWatch();
sw.start("Method-1"); // Start a named task
Thread.sleep(500);
sw.stop();
sw.start("Method-2");
Thread.sleep(300);
sw.stop();
sw.start("Method-3");
Thread.sleep(200);
sw.stop();
System.out.println("Total time in milliseconds for all tasks :\n"+sw.getTotalTimeMillis());
System.out.println("Table describing all tasks performed :\n"+sw.prettyPrint());
System.out.format("Time taken by the last task : [%s]:[%d]",
sw.getLastTaskName(),sw.getLastTaskTimeMillis());
System.out.println("\n Array of the data for tasks performed « Task Name: Time Taken");
TaskInfo[] listofTasks = sw.getTaskInfo();
for (TaskInfo task : listofTasks) {
System.out.format("[%s]:[%d]\n",
task.getTaskName(), task.getTimeMillis());
}
OutPut:
Total time in milliseconds for all tasks :
999
Table describing all tasks performed :
StopWatch '': running time (millis) = 999
-----------------------------------------
ms % Task name
-----------------------------------------
00500 050% Method-1
00299 030% Method-2
00200 020% Method-3
Time taken by the last task : [Method-3]:[200]
Array of the data for tasks performed « Task Name: Time Taken
[Method-1]:[500]
[Method-2]:[299]
[Method-3]:[200]
If you want wall-clock time
long start_time = System.currentTimeMillis();
object.method();
long end_time = System.currentTimeMillis();
long execution_time = end_time - start_time;
As "skaffman" said, use AOP OR you can use run time bytecode weaving, just like unit test method coverage tools use to transparently add timing info to methods invoked.
You can look at code used by open source tools tools like Emma (http://downloads.sourceforge.net/emma/emma-2.0.5312-src.zip?modtime=1118607545&big_mirror=0). The other opensource coverage tool is http://prdownloads.sourceforge.net/cobertura/cobertura-1.9-src.zip?download.
If you eventually manage to do what you set out for, pls. share it back with the community here with your ant task/jars.
You can use stopwatch class from spring core project:
Code:
StopWatch stopWatch = new StopWatch()
stopWatch.start(); //start stopwatch
// write your function or line of code.
stopWatch.stop(); //stop stopwatch
stopWatch.getTotalTimeMillis() ; ///get total time
Documentation for Stopwatch: Simple stop watch, allowing for timing of a number of tasks, exposing total running time and running time for each named task. Conceals use of System.currentTimeMillis(), improving the readability of application code and reducing the likelihood of calculation errors. Note that this object is not designed to be thread-safe and does not use synchronization. This class is normally used to verify performance during proof-of-concepts and in development, rather than as part of production applications.
A strategy that works to me in java ee was:
Create a class with a method annotated with @AroundInvoke;
@Singleton
public class TimedInterceptor implements Serializable {
@AroundInvoke
public Object logMethod(InvocationContext ic) throws Exception {
Date start = new Date();
Object result = ic.proceed();
Date end = new Date();
System.out.println("time: " + (end.getTime - start.getTime()));
return result;
}
}
Annotate the method that you want to monitoring:
@Interceptors(TimedInterceptor.class)
public void onMessage(final Message message) { ...
I hope this can help.
I have written a method to print the method execution time in a much readable form. For example, to calculate the factorial of 1 Million, it takes approximately 9 minutes. So the execution time get printed as:
Execution Time: 9 Minutes, 36 Seconds, 237 MicroSeconds, 806193 NanoSeconds
The code is here:
public class series
{
public static void main(String[] args)
{
long startTime = System.nanoTime();
long n = 10_00_000;
printFactorial(n);
long endTime = System.nanoTime();
printExecutionTime(startTime, endTime);
}
public static void printExecutionTime(long startTime, long endTime)
{
long time_ns = endTime - startTime;
long time_ms = TimeUnit.NANOSECONDS.toMillis(time_ns);
long time_sec = TimeUnit.NANOSECONDS.toSeconds(time_ns);
long time_min = TimeUnit.NANOSECONDS.toMinutes(time_ns);
long time_hour = TimeUnit.NANOSECONDS.toHours(time_ns);
System.out.print("\nExecution Time: ");
if(time_hour > 0)
System.out.print(time_hour + " Hours, ");
if(time_min > 0)
System.out.print(time_min % 60 + " Minutes, ");
if(time_sec > 0)
System.out.print(time_sec % 60 + " Seconds, ");
if(time_ms > 0)
System.out.print(time_ms % 1E+3 + " MicroSeconds, ");
if(time_ns > 0)
System.out.print(time_ns % 1E+6 + " NanoSeconds");
}
}
Ok, this is a simple class to be used for simple simple timing of your functions. There is an example below it.
public class Stopwatch {
static long startTime;
static long splitTime;
static long endTime;
public Stopwatch() {
start();
}
public void start() {
startTime = System.currentTimeMillis();
splitTime = System.currentTimeMillis();
endTime = System.currentTimeMillis();
}
public void split() {
split("");
}
public void split(String tag) {
endTime = System.currentTimeMillis();
System.out.println("Split time for [" + tag + "]: " + (endTime - splitTime) + " ms");
splitTime = endTime;
}
public void end() {
end("");
}
public void end(String tag) {
endTime = System.currentTimeMillis();
System.out.println("Final time for [" + tag + "]: " + (endTime - startTime) + " ms");
}
}
Sample of use:
public static Schedule getSchedule(Activity activity_context) {
String scheduleJson = null;
Schedule schedule = null;
/*->*/ Stopwatch stopwatch = new Stopwatch();
InputStream scheduleJsonInputStream = activity_context.getResources().openRawResource(R.raw.skating_times);
/*->*/ stopwatch.split("open raw resource");
scheduleJson = FileToString.convertStreamToString(scheduleJsonInputStream);
/*->*/ stopwatch.split("file to string");
schedule = new Gson().fromJson(scheduleJson, Schedule.class);
/*->*/ stopwatch.split("parse Json");
/*->*/ stopwatch.end("Method getSchedule");
return schedule;
}
Sample of console output:
Split time for [file to string]: 672 ms
Split time for [parse Json]: 893 ms
Final time for [get Schedule]: 1565 ms
Using AOP/AspectJ and @Loggable annotation from jcabi-aspects you can do it easy and compact:
@Loggable(Loggable.DEBUG)
public String getSomeResult() {
// return some value
}
Every call to this method will be sent to SLF4J logging facility with DEBUG logging level. And every log message will include execution time.
I implemented a simple timer, And I think it's really useful:
public class Timer{
private static long start_time;
public static double tic(){
return start_time = System.nanoTime();
}
public static double toc(){
return (System.nanoTime()-start_time)/1000000000.0;
}
}
That way you can time one or more actions:
Timer.tic();
// Code 1
System.out.println("Code 1 runtime: "+Timer.toc()+" seconds.");
// Code 2
System.out.println("(Code 1 + Code 2) runtime: "+Timer.toc()+"seconds");
Timer.tic();
// Code 3
System.out.println("Code 3 runtime: "+Timer.toc()+" seconds.");
FYI, JEP 230: Microbenchmark Suite is an OpenJDK project to:
Add a basic suite of microbenchmarks to the JDK source code, and make it easy for developers to run existing microbenchmarks and create new ones.
This feature arrived in Java 12.
For earlier versions of Java, take a look at the Java Microbenchmark Harness (JMH) project on which JEP 230 is based.
Performance measurements on my machine
System.nanoTime() : 750nsSystem.currentTimeMillis() : 18nsAs mentioned, System.nanoTime() is thought to measure elapsed time. Just be aware of the cost if used inside a loop or the like.
new Timer(""){{
// code to time
}}.timeMe();
public class Timer {
private final String timerName;
private long started;
public Timer(String timerName) {
this.timerName = timerName;
this.started = System.currentTimeMillis();
}
public void timeMe() {
System.out.println(
String.format("Execution of '%s' takes %dms.",
timerName,
started-System.currentTimeMillis()));
}
}
I go with the simple answer. Works for me.
long startTime = System.currentTimeMillis();
doReallyLongThing();
long endTime = System.currentTimeMillis();
System.out.println("That took " + (endTime - startTime) + " milliseconds");
It works quite well. The resolution is obviously only to the millisecond, you can do better with System.nanoTime(). There are some limitations to both (operating system schedule slices, etc.) but this works pretty well.
Average across a couple of runs (the more the better) and you'll get a decent idea.
I modified the code from correct answer to get result in seconds:
long startTime = System.nanoTime();
methodCode ...
long endTime = System.nanoTime();
double duration = (double)(endTime - startTime) / (Math.pow(10, 9));
Log.v(TAG, "MethodName time (s) = " + duration);
There are a couple of ways to do that. I normally fall back to just using something like this:
long start = System.currentTimeMillis();
// ... do something ...
long end = System.currentTimeMillis();
or the same thing with System.nanoTime();
For something more on the benchmarking side of things there seems also to be this one: http://jetm.void.fm/ Never tried it though.
long startTime = System.currentTimeMillis();
// code goes here
long finishTime = System.currentTimeMillis();
long elapsedTime = finishTime - startTime; // elapsed time in milliseconds
Use a profiler (JProfiler, Netbeans Profiler, Visual VM, Eclipse Profiler, etc). You'll get the most accurate results and is the least intrusive. They use the built-in JVM mechanism for profiling which can also give you extra information like stack traces, execution paths, and more comprehensive results if necessary.
When using a fully integrated profiler, it's faily trivial to profile a method. Right click, Profiler -> Add to Root Methods. Then run the profiler just like you were doing a test run or debugger.
In Spring framework we have a call called StopWatch (org.springframework.util.StopWatch)
//measuring elapsed time using Spring StopWatch
StopWatch watch = new StopWatch();
watch.start();
for(int i=0; i< 1000; i++){
Object obj = new Object();
}
watch.stop();
System.out.println("Total execution time to create 1000 objects in Java using StopWatch in millis: "
+ watch.getTotalTimeMillis());
Also We can use StopWatch class of Apache commons for measuring the time.
Sample code
org.apache.commons.lang.time.StopWatch sw = new org.apache.commons.lang.time.StopWatch();
System.out.println("getEventFilterTreeData :: Start Time : " + sw.getTime());
sw.start();
// Method execution code
sw.stop();
System.out.println("getEventFilterTreeData :: End Time : " + sw.getTime());
There are a couple of ways to do that. I normally fall back to just using something like this:
long start = System.currentTimeMillis();
// ... do something ...
long end = System.currentTimeMillis();
or the same thing with System.nanoTime();
For something more on the benchmarking side of things there seems also to be this one: http://jetm.void.fm/ Never tried it though.
I go with the simple answer. Works for me.
long startTime = System.currentTimeMillis();
doReallyLongThing();
long endTime = System.currentTimeMillis();
System.out.println("That took " + (endTime - startTime) + " milliseconds");
It works quite well. The resolution is obviously only to the millisecond, you can do better with System.nanoTime(). There are some limitations to both (operating system schedule slices, etc.) but this works pretty well.
Average across a couple of runs (the more the better) and you'll get a decent idea.
new Timer(""){{
// code to time
}}.timeMe();
public class Timer {
private final String timerName;
private long started;
public Timer(String timerName) {
this.timerName = timerName;
this.started = System.currentTimeMillis();
}
public void timeMe() {
System.out.println(
String.format("Execution of '%s' takes %dms.",
timerName,
started-System.currentTimeMillis()));
}
}
You can try this way if just want know the time.
long startTime = System.currentTimeMillis();
//@ Method call
System.out.println("Total time [ms]: " + (System.currentTimeMillis() - startTime));
There are a couple of ways to do that. I normally fall back to just using something like this:
long start = System.currentTimeMillis();
// ... do something ...
long end = System.currentTimeMillis();
or the same thing with System.nanoTime();
For something more on the benchmarking side of things there seems also to be this one: http://jetm.void.fm/ Never tried it though.
We are using AspectJ and Java annotations for this purpose. If we need to know to execution time for a method, we simple annotate it. A more advanced version could use an own log level that can enabled and disabled at runtime.
public @interface Trace {
boolean showParameters();
}
@Aspect
public class TraceAspect {
[...]
@Around("tracePointcut() && @annotation(trace) && !within(TraceAspect)")
public Object traceAdvice ( ProceedingJintPoint jP, Trace trace ) {
Object result;
// initilize timer
try {
result = jp.procced();
} finally {
// calculate execution time
}
return result;
}
[...]
}
System.nanoTime() is a pretty precise system utility to measure execution time. But be careful, if you're running on pre-emptive scheduler mode (default), this utility actually measures wall-clock time and not CPU time. Therefore, you may notice different execution time values from run to run, depending on system loads. If you look for CPU time, I think that running your program in real-time mode will do the trick. You have to use RT linux. link: Real-time programming with Linux
It would be nice if java had a better functional support, so that the action, that needs to be measured, could be wrapped into a block:
measure {
// your operation here
}
In java this could be done by anonymous functions, that look too verbose
public interface Timer {
void wrap();
}
public class Logger {
public static void logTime(Timer timer) {
long start = System.currentTimeMillis();
timer.wrap();
System.out.println("" + (System.currentTimeMillis() - start) + "ms");
}
public static void main(String a[]) {
Logger.logTime(new Timer() {
public void wrap() {
// Your method here
timeConsumingOperation();
}
});
}
public static void timeConsumingOperation() {
for (int i = 0; i<=10000; i++) {
System.out.println("i=" +i);
}
}
}
This probably isn't what you wanted me to say, but this is a good use of AOP. Whip an proxy interceptor around your method, and do the timing in there.
The what, why and how of AOP is rather beyond the scope of this answer, sadly, but that's how I'd likely do it.
Edit: Here's a link to Spring AOP to get you started, if you're keen. This is the most accessible implementation of AOP that Iive come across for java.
Also, given everyone else's very simple suggestions, I should add that AOP is for when you don't want stuff like timing to invade your code. But in many cases, that sort of simple and easy approach is fine.
Pure Java SE code, no need for adding dependency, using TimeTracedExecuter :
public static void main(String[] args) {
Integer square = new TimeTracedExecutor<>(Main::calculateSquare)
.executeWithInput("calculate square of num",5,logger);
}
public static int calculateSquare(int num){
return num*num;
}
Will produce result like this:
INFO: It took 3 milliseconds to calculate square of num
Custom reusable class: TimeTracedExecutor
import java.text.NumberFormat;
import java.time.Duration;
import java.time.Instant;
import java.util.function.Function;
import java.util.logging.Logger;
public class TimeTracedExecutor<T,R> {
Function<T,R> methodToExecute;
public TimeTracedExecutor(Function<T, R> methodToExecute) {
this.methodToExecute = methodToExecute;
}
public R executeWithInput(String taskDescription, T t, Logger logger){
Instant start = Instant.now();
R r= methodToExecute.apply(t);
Instant finish = Instant.now();
String format = "It took %s milliseconds to "+taskDescription;
String elapsedTime = NumberFormat.getNumberInstance().format(Duration.between(start, finish).toMillis());
logger.info(String.format(format, elapsedTime));
return r;
}
}
Really good code.
http://www.rgagnon.com/javadetails/java-0585.html
import java.util.concurrent.TimeUnit;
long startTime = System.currentTimeMillis();
........
........
........
long finishTime = System.currentTimeMillis();
String diff = millisToShortDHMS(finishTime - startTime);
/**
* converts time (in milliseconds) to human-readable format
* "<dd:>hh:mm:ss"
*/
public static String millisToShortDHMS(long duration) {
String res = "";
long days = TimeUnit.MILLISECONDS.toDays(duration);
long hours = TimeUnit.MILLISECONDS.toHours(duration)
- TimeUnit.DAYS.toHours(TimeUnit.MILLISECONDS.toDays(duration));
long minutes = TimeUnit.MILLISECONDS.toMinutes(duration)
- TimeUnit.HOURS.toMinutes(TimeUnit.MILLISECONDS.toHours(duration));
long seconds = TimeUnit.MILLISECONDS.toSeconds(duration)
- TimeUnit.MINUTES.toSeconds(TimeUnit.MILLISECONDS.toMinutes(duration));
if (days == 0) {
res = String.format("%02d:%02d:%02d", hours, minutes, seconds);
}
else {
res = String.format("%dd%02d:%02d:%02d", days, hours, minutes, seconds);
}
return res;
}
For java 8+, another possible solution (more general, func-style and without aspects) could be to create some utility method accepting code as a parameter
public static <T> T timed (String description, Consumer<String> out, Supplier<T> code) {
final LocalDateTime start = LocalDateTime.now ();
T res = code.get ();
final long execTime = Duration.between (start, LocalDateTime.now ()).toMillis ();
out.accept (String.format ("%s: %d ms", description, execTime));
return res;
}
And the calling code could be smth like that:
public static void main (String[] args) throws InterruptedException {
timed ("Simple example", System.out::println, Timing::myCode);
}
public static Object myCode () {
try {
Thread.sleep (1500);
} catch (InterruptedException e) {
e.printStackTrace ();
}
return null;
}
Use a profiler (JProfiler, Netbeans Profiler, Visual VM, Eclipse Profiler, etc). You'll get the most accurate results and is the least intrusive. They use the built-in JVM mechanism for profiling which can also give you extra information like stack traces, execution paths, and more comprehensive results if necessary.
When using a fully integrated profiler, it's faily trivial to profile a method. Right click, Profiler -> Add to Root Methods. Then run the profiler just like you were doing a test run or debugger.
You can use stopwatch class from spring core project:
Code:
StopWatch stopWatch = new StopWatch()
stopWatch.start(); //start stopwatch
// write your function or line of code.
stopWatch.stop(); //stop stopwatch
stopWatch.getTotalTimeMillis() ; ///get total time
Documentation for Stopwatch: Simple stop watch, allowing for timing of a number of tasks, exposing total running time and running time for each named task. Conceals use of System.currentTimeMillis(), improving the readability of application code and reducing the likelihood of calculation errors. Note that this object is not designed to be thread-safe and does not use synchronization. This class is normally used to verify performance during proof-of-concepts and in development, rather than as part of production applications.
With Java 8 you can do also something like this with every normal methods:
Object returnValue = TimeIt.printTime(() -> methodeWithReturnValue());
//do stuff with your returnValue
with TimeIt like:
public class TimeIt {
public static <T> T printTime(Callable<T> task) {
T call = null;
try {
long startTime = System.currentTimeMillis();
call = task.call();
System.out.print((System.currentTimeMillis() - startTime) / 1000d + "s");
} catch (Exception e) {
//...
}
return call;
}
}
With this methode you can make easy time measurement anywhere in your code without breaking it. In this simple example i just print the time. May you add a Switch for TimeIt, e.g. to only print the time in DebugMode or something.
If you are working with Function you can do somthing like this:
Function<Integer, Integer> yourFunction= (n) -> {
return IntStream.range(0, n).reduce(0, (a, b) -> a + b);
};
Integer returnValue = TimeIt.printTime2(yourFunction).apply(10000);
//do stuff with your returnValue
public static <T, R> Function<T, R> printTime2(Function<T, R> task) {
return (t) -> {
long startTime = System.currentTimeMillis();
R apply = task.apply(t);
System.out.print((System.currentTimeMillis() - startTime) / 1000d
+ "s");
return apply;
};
}
Come on guys! Nobody mentioned the Guava way to do that (which is arguably awesome):
import com.google.common.base.Stopwatch;
Stopwatch timer = Stopwatch.createStarted();
//method invocation
LOG.info("Method took: " + timer.stop());
The nice thing is that Stopwatch.toString() does a good job of selecting time units for the measurement. I.e. if the value is small, it'll output 38 ns, if it's long, it'll show 5m 3s
Even nicer:
Stopwatch timer = Stopwatch.createUnstarted();
for (...) {
timer.start();
methodToTrackTimeFor();
timer.stop();
methodNotToTrackTimeFor();
}
LOG.info("Method took: " + timer);
Note: Google Guava requires Java 1.6+
For java 8+, another possible solution (more general, func-style and without aspects) could be to create some utility method accepting code as a parameter
public static <T> T timed (String description, Consumer<String> out, Supplier<T> code) {
final LocalDateTime start = LocalDateTime.now ();
T res = code.get ();
final long execTime = Duration.between (start, LocalDateTime.now ()).toMillis ();
out.accept (String.format ("%s: %d ms", description, execTime));
return res;
}
And the calling code could be smth like that:
public static void main (String[] args) throws InterruptedException {
timed ("Simple example", System.out::println, Timing::myCode);
}
public static Object myCode () {
try {
Thread.sleep (1500);
} catch (InterruptedException e) {
e.printStackTrace ();
}
return null;
}
In Spring framework we have a call called StopWatch (org.springframework.util.StopWatch)
//measuring elapsed time using Spring StopWatch
StopWatch watch = new StopWatch();
watch.start();
for(int i=0; i< 1000; i++){
Object obj = new Object();
}
watch.stop();
System.out.println("Total execution time to create 1000 objects in Java using StopWatch in millis: "
+ watch.getTotalTimeMillis());
This probably isn't what you wanted me to say, but this is a good use of AOP. Whip an proxy interceptor around your method, and do the timing in there.
The what, why and how of AOP is rather beyond the scope of this answer, sadly, but that's how I'd likely do it.
Edit: Here's a link to Spring AOP to get you started, if you're keen. This is the most accessible implementation of AOP that Iive come across for java.
Also, given everyone else's very simple suggestions, I should add that AOP is for when you don't want stuff like timing to invade your code. But in many cases, that sort of simple and easy approach is fine.
As "skaffman" said, use AOP OR you can use run time bytecode weaving, just like unit test method coverage tools use to transparently add timing info to methods invoked.
You can look at code used by open source tools tools like Emma (http://downloads.sourceforge.net/emma/emma-2.0.5312-src.zip?modtime=1118607545&big_mirror=0). The other opensource coverage tool is http://prdownloads.sourceforge.net/cobertura/cobertura-1.9-src.zip?download.
If you eventually manage to do what you set out for, pls. share it back with the community here with your ant task/jars.
We are using AspectJ and Java annotations for this purpose. If we need to know to execution time for a method, we simple annotate it. A more advanced version could use an own log level that can enabled and disabled at runtime.
public @interface Trace {
boolean showParameters();
}
@Aspect
public class TraceAspect {
[...]
@Around("tracePointcut() && @annotation(trace) && !within(TraceAspect)")
public Object traceAdvice ( ProceedingJintPoint jP, Trace trace ) {
Object result;
// initilize timer
try {
result = jp.procced();
} finally {
// calculate execution time
}
return result;
}
[...]
}
You can use Metrics library which provides various measuring instruments. Add dependency:
<dependencies>
<dependency>
<groupId>io.dropwizard.metrics</groupId>
<artifactId>metrics-core</artifactId>
<version>${metrics.version}</version>
</dependency>
</dependencies>
And configure it for your environment.
Methods can be annotated with @Timed:
@Timed
public void exampleMethod(){
// some code
}
or piece of code wrapped with Timer:
final Timer timer = metricsRegistry.timer("some_name");
final Timer.Context context = timer.time();
// timed code
context.stop();
Aggregated metrics can exported to console, JMX, CSV or other.
@Timed metrics output example:
com.example.ExampleService.exampleMethod
count = 2
mean rate = 3.11 calls/minute
1-minute rate = 0.96 calls/minute
5-minute rate = 0.20 calls/minute
15-minute rate = 0.07 calls/minute
min = 17.01 milliseconds
max = 1006.68 milliseconds
mean = 511.84 milliseconds
stddev = 699.80 milliseconds
median = 511.84 milliseconds
75% <= 1006.68 milliseconds
95% <= 1006.68 milliseconds
98% <= 1006.68 milliseconds
99% <= 1006.68 milliseconds
99.9% <= 1006.68 milliseconds
As "skaffman" said, use AOP OR you can use run time bytecode weaving, just like unit test method coverage tools use to transparently add timing info to methods invoked.
You can look at code used by open source tools tools like Emma (http://downloads.sourceforge.net/emma/emma-2.0.5312-src.zip?modtime=1118607545&big_mirror=0). The other opensource coverage tool is http://prdownloads.sourceforge.net/cobertura/cobertura-1.9-src.zip?download.
If you eventually manage to do what you set out for, pls. share it back with the community here with your ant task/jars.
long startTime = System.currentTimeMillis();
// code goes here
long finishTime = System.currentTimeMillis();
long elapsedTime = finishTime - startTime; // elapsed time in milliseconds
Really good code.
http://www.rgagnon.com/javadetails/java-0585.html
import java.util.concurrent.TimeUnit;
long startTime = System.currentTimeMillis();
........
........
........
long finishTime = System.currentTimeMillis();
String diff = millisToShortDHMS(finishTime - startTime);
/**
* converts time (in milliseconds) to human-readable format
* "<dd:>hh:mm:ss"
*/
public static String millisToShortDHMS(long duration) {
String res = "";
long days = TimeUnit.MILLISECONDS.toDays(duration);
long hours = TimeUnit.MILLISECONDS.toHours(duration)
- TimeUnit.DAYS.toHours(TimeUnit.MILLISECONDS.toDays(duration));
long minutes = TimeUnit.MILLISECONDS.toMinutes(duration)
- TimeUnit.HOURS.toMinutes(TimeUnit.MILLISECONDS.toHours(duration));
long seconds = TimeUnit.MILLISECONDS.toSeconds(duration)
- TimeUnit.MINUTES.toSeconds(TimeUnit.MILLISECONDS.toMinutes(duration));
if (days == 0) {
res = String.format("%02d:%02d:%02d", hours, minutes, seconds);
}
else {
res = String.format("%dd%02d:%02d:%02d", days, hours, minutes, seconds);
}
return res;
}
It would be nice if java had a better functional support, so that the action, that needs to be measured, could be wrapped into a block:
measure {
// your operation here
}
In java this could be done by anonymous functions, that look too verbose
public interface Timer {
void wrap();
}
public class Logger {
public static void logTime(Timer timer) {
long start = System.currentTimeMillis();
timer.wrap();
System.out.println("" + (System.currentTimeMillis() - start) + "ms");
}
public static void main(String a[]) {
Logger.logTime(new Timer() {
public void wrap() {
// Your method here
timeConsumingOperation();
}
});
}
public static void timeConsumingOperation() {
for (int i = 0; i<=10000; i++) {
System.out.println("i=" +i);
}
}
}
Pure Java SE code, no need for adding dependency, using TimeTracedExecuter :
public static void main(String[] args) {
Integer square = new TimeTracedExecutor<>(Main::calculateSquare)
.executeWithInput("calculate square of num",5,logger);
}
public static int calculateSquare(int num){
return num*num;
}
Will produce result like this:
INFO: It took 3 milliseconds to calculate square of num
Custom reusable class: TimeTracedExecutor
import java.text.NumberFormat;
import java.time.Duration;
import java.time.Instant;
import java.util.function.Function;
import java.util.logging.Logger;
public class TimeTracedExecutor<T,R> {
Function<T,R> methodToExecute;
public TimeTracedExecutor(Function<T, R> methodToExecute) {
this.methodToExecute = methodToExecute;
}
public R executeWithInput(String taskDescription, T t, Logger logger){
Instant start = Instant.now();
R r= methodToExecute.apply(t);
Instant finish = Instant.now();
String format = "It took %s milliseconds to "+taskDescription;
String elapsedTime = NumberFormat.getNumberInstance().format(Duration.between(start, finish).toMillis());
logger.info(String.format(format, elapsedTime));
return r;
}
}
I modified the code from correct answer to get result in seconds:
long startTime = System.nanoTime();
methodCode ...
long endTime = System.nanoTime();
double duration = (double)(endTime - startTime) / (Math.pow(10, 9));
Log.v(TAG, "MethodName time (s) = " + duration);
You can use javaagent to modify the java class bytes ,add the monitor codes dynamically.there is some open source tools on the github that can do this for you.
If you want to do it by yourself, just implements the javaagent,use javassist to modify the methods you want to monitor,and the monitor code before your method return.it's clean and you can monitor systems that you even don't have source code.
Using AOP/AspectJ and @Loggable annotation from jcabi-aspects you can do it easy and compact:
@Loggable(Loggable.DEBUG)
public String getSomeResult() {
// return some value
}
Every call to this method will be sent to SLF4J logging facility with DEBUG logging level. And every log message will include execution time.
You can use Perf4j. Very cool utility. Usage is simple
String watchTag = "target.SomeMethod";
StopWatch stopWatch = new LoggingStopWatch(watchTag);
Result result = null; // Result is a type of a return value of a method
try {
result = target.SomeMethod();
stopWatch.stop(watchTag + ".success");
} catch (Exception e) {
stopWatch.stop(watchTag + ".fail", "Exception was " + e);
throw e;
}
More information can be found in Developer Guide
Edit: Project seems dead
Here is pretty printed string ready formated seconds elapsed similar to google search time took to search:
long startTime = System.nanoTime();
// ... methodToTime();
long endTime = System.nanoTime();
long duration = (endTime - startTime);
long seconds = (duration / 1000) % 60;
// formatedSeconds = (0.xy seconds)
String formatedSeconds = String.format("(0.%d seconds)", seconds);
System.out.println("formatedSeconds = "+ formatedSeconds);
// i.e actual formatedSeconds = (0.52 seconds)
We are using AspectJ and Java annotations for this purpose. If we need to know to execution time for a method, we simple annotate it. A more advanced version could use an own log level that can enabled and disabled at runtime.
public @interface Trace {
boolean showParameters();
}
@Aspect
public class TraceAspect {
[...]
@Around("tracePointcut() && @annotation(trace) && !within(TraceAspect)")
public Object traceAdvice ( ProceedingJintPoint jP, Trace trace ) {
Object result;
// initilize timer
try {
result = jp.procced();
} finally {
// calculate execution time
}
return result;
}
[...]
}
Source: Stackoverflow.com