June 2009 - logback-dev - mailman3.qos.ch

[JIRA] Updated: (LBCLASSIC-36) Synchronization of SimpleDateFormat object in DateConverter
by Ralph Goers (JIRA) 15 Jun '09

15 Jun '09

[ http://jira.qos.ch/browse/LBCLASSIC-36?page=com.atlassian.jira.plugin.syste… ] Ralph Goers updated LBCLASSIC-36: --------------------------------- Attachment: DateFormatPerfTest.java Updated version of test. > Synchronization of SimpleDateFormat object in DateConverter > ----------------------------------------------------------- > > Key: LBCLASSIC-36 > URL: http://jira.qos.ch/browse/LBCLASSIC-36 > Project: logback-classic > Issue Type: Bug > Components: Other > Affects Versions: unspecified > Environment: Operating System: All > Platform: All > Reporter: Anders Hammar > Assignee: Ceki Gulcu > Attachments: DateConverter_patch.java, DateFormatPerfTest.java, dateformattest.zip > > > According to the API documentation of SimpleDateFormat, date formats are not synchronized and it is recommended to create separate format instances for each thread. > However, a possibly simpler (mer efficient) solution would be to synchronize on the SimpleDateFormat object in the convert() method to prevent two threads to access it concurrently. -- This message is automatically generated by JIRA. - If you think it was sent incorrectly contact one of the administrators: http://jira.qos.ch/secure/Administrators.jspa - For more information on JIRA, see: http://www.atlassian.com/software/jira

1 0

[JIRA] Commented: (LBCLASSIC-36) Synchronization of SimpleDateFormat object in DateConverter
by Ralph Goers (JIRA) 15 Jun '09

15 Jun '09

[ http://jira.qos.ch/browse/LBCLASSIC-36?page=com.atlassian.jira.plugin.syste… ] Ralph Goers commented on LBCLASSIC-36: -------------------------------------- There are several problems with your updated tests. 1. There is no reason to synchronize in the Joda test. DateTimeFormatter should be declared external to the method and the synchronization should be removed since DateTimeFormatter is thread safe. 2. You are not testing thread contention. I have modified the test to correct this and am adding it here. I am not sure that it adequately emulates logging though, since there is no other work occurring besides formatting the time. > Synchronization of SimpleDateFormat object in DateConverter > ----------------------------------------------------------- > > Key: LBCLASSIC-36 > URL: http://jira.qos.ch/browse/LBCLASSIC-36 > Project: logback-classic > Issue Type: Bug > Components: Other > Affects Versions: unspecified > Environment: Operating System: All > Platform: All > Reporter: Anders Hammar > Assignee: Ceki Gulcu > Attachments: DateConverter_patch.java, dateformattest.zip > > > According to the API documentation of SimpleDateFormat, date formats are not synchronized and it is recommended to create separate format instances for each thread. > However, a possibly simpler (mer efficient) solution would be to synchronize on the SimpleDateFormat object in the convert() method to prevent two threads to access it concurrently. -- This message is automatically generated by JIRA. - If you think it was sent incorrectly contact one of the administrators: http://jira.qos.ch/secure/Administrators.jspa - For more information on JIRA, see: http://www.atlassian.com/software/jira

1 0

[JIRA] Issue Comment Edited: (LBCLASSIC-36) Synchronization of SimpleDateFormat object in DateConverter
by Ceki Gulcu (JIRA) 14 Jun '09

14 Jun '09

[ http://jira.qos.ch/browse/LBCLASSIC-36?page=com.atlassian.jira.plugin.syste… ] Ceki Gulcu edited comment on LBCLASSIC-36 at 6/14/09 9:38 PM: -------------------------------------------------------------- Thank you for this contribution. I have ran a modified (and simplified) version of your tests. It looks like SimpleDateFormat and Yoda time have roughly the same performance, with the former having the edge. Your version of the test had Yoda time transform Date objects. However, in logback, a "long" as obtained by System.currentMillis(), needs to be transformed. Transforming Date objects is pretty much beside the point, invalidating any argument made by the test you contributed. Just as importantly, logback will invoke Yoda time or SimpleDateformat, at most once per millisecond. So the performance of SimpleDateformat as well as Yoda time are as satisfactory as they are unimportant. Only the performance of System.currentMillis really matters. was (Author: noreply.ceki(a)qos.ch) Thank you for this contribution. I have ran a modified (and simplified) version of your tests. It looks like SimpleDateFormat and Yoda time have roughly the same performance. with the former having the edge. Your version of the test had Yoda time transform Date objects. However, in logback, a "long" as obtained by System.currentMillis(), needs to be transformed. Transforming Date objects is pretty much beside the point. > Synchronization of SimpleDateFormat object in DateConverter > ----------------------------------------------------------- > > Key: LBCLASSIC-36 > URL: http://jira.qos.ch/browse/LBCLASSIC-36 > Project: logback-classic > Issue Type: Bug > Components: Other > Affects Versions: unspecified > Environment: Operating System: All > Platform: All > Reporter: Anders Hammar > Assignee: Ceki Gulcu > Attachments: DateConverter_patch.java, dateformattest.zip > > > According to the API documentation of SimpleDateFormat, date formats are not synchronized and it is recommended to create separate format instances for each thread. > However, a possibly simpler (mer efficient) solution would be to synchronize on the SimpleDateFormat object in the convert() method to prevent two threads to access it concurrently. -- This message is automatically generated by JIRA. - If you think it was sent incorrectly contact one of the administrators: http://jira.qos.ch/secure/Administrators.jspa - For more information on JIRA, see: http://www.atlassian.com/software/jira

3 4

[JIRA] Commented: (LBCLASSIC-36) Synchronization of SimpleDateFormat object in DateConverter
by Ceki Gulcu (JIRA) 14 Jun '09

14 Jun '09

[ http://jira.qos.ch/browse/LBCLASSIC-36?page=com.atlassian.jira.plugin.syste… ] Ceki Gulcu commented on LBCLASSIC-36: ------------------------------------- I modified DateFormatPerftest in ch.qos.logback.classic.issue.lbclassic36 so as to take into account the kind of millisecond cache that we would keep. This is reflected in the "guarded" versions of the various methods. Here is a sample result. Raw Joda: average tick 2792.437593 nanoseconds Raw SDF: average tick 2147.457408 nanoseconds Guarded Joda: average tick 138.560775 nanoseconds Guarded SDF: average tick 73.691388 nanoseconds The guarded SDF version is very quick (less than 100 nanos) and is probably the code that will be used in the future. Please let me know if you see any faults or downsides to this proposal. > Synchronization of SimpleDateFormat object in DateConverter > ----------------------------------------------------------- > > Key: LBCLASSIC-36 > URL: http://jira.qos.ch/browse/LBCLASSIC-36 > Project: logback-classic > Issue Type: Bug > Components: Other > Affects Versions: unspecified > Environment: Operating System: All > Platform: All > Reporter: Anders Hammar > Assignee: Ceki Gulcu > Attachments: DateConverter_patch.java, dateformattest.zip > > > According to the API documentation of SimpleDateFormat, date formats are not synchronized and it is recommended to create separate format instances for each thread. > However, a possibly simpler (mer efficient) solution would be to synchronize on the SimpleDateFormat object in the convert() method to prevent two threads to access it concurrently. -- This message is automatically generated by JIRA. - If you think it was sent incorrectly contact one of the administrators: http://jira.qos.ch/secure/Administrators.jspa - For more information on JIRA, see: http://www.atlassian.com/software/jira

1 0

svn commit: r2292 - logback/trunk/logback-classic/src/test/java/ch/qos/logback/classic/issue/lbclassic36
by noreply.ceki＠qos.ch 14 Jun '09

14 Jun '09

Author: ceki Date: Sun Jun 14 21:56:16 2009 New Revision: 2292 Added: logback/trunk/logback-classic/src/test/java/ch/qos/logback/classic/issue/lbclassic36/DateFormatPerfTest.java - copied, changed from r2291, /logback/trunk/logback-classic/src/test/java/ch/qos/logback/classic/issue/lbclassic36/TestDateFormat.java logback/trunk/logback-classic/src/test/java/ch/qos/logback/classic/issue/lbclassic36/TestDateFormatOriginal.java Removed: logback/trunk/logback-classic/src/test/java/ch/qos/logback/classic/issue/lbclassic36/TestDateFormat.java Log: LBCLASSIC-36 Rename application Copied: logback/trunk/logback-classic/src/test/java/ch/qos/logback/classic/issue/lbclassic36/DateFormatPerfTest.java (from r2291, /logback/trunk/logback-classic/src/test/java/ch/qos/logback/classic/issue/lbclassic36/TestDateFormat.java) ============================================================================== --- /logback/trunk/logback-classic/src/test/java/ch/qos/logback/classic/issue/lbclassic36/TestDateFormat.java (original) +++ logback/trunk/logback-classic/src/test/java/ch/qos/logback/classic/issue/lbclassic36/DateFormatPerfTest.java Sun Jun 14 21:56:16 2009 @@ -6,7 +6,7 @@ import org.joda.time.format.DateTimeFormat; import org.joda.time.format.DateTimeFormatter; -public class TestDateFormat { +public class DateFormatPerfTest { public static final String ISO8601_PATTERN = "yyyy-MM-dd HH:mm:ss,SSS"; static final long NANOS_IN_ONE_SEC = 1000 * 1000 * 1000L; Added: logback/trunk/logback-classic/src/test/java/ch/qos/logback/classic/issue/lbclassic36/TestDateFormatOriginal.java ============================================================================== --- (empty file) +++ logback/trunk/logback-classic/src/test/java/ch/qos/logback/classic/issue/lbclassic36/TestDateFormatOriginal.java Sun Jun 14 21:56:16 2009 @@ -0,0 +1,239 @@ +package ch.qos.logback.classic.issue.lbclassic36; + +import junit.framework.TestCase; +import junit.framework.Test; +import junit.framework.TestSuite; + +import java.text.SimpleDateFormat; +import java.util.Date; + +import org.joda.time.format.DateTimeFormatter; +import org.joda.time.format.DateTimeFormat; +import org.joda.time.DateTime; + +public class TestDateFormatOriginal extends TestCase { + public static final String ISO8601_PATTERN = "yyyy-MM-dd HH:mm:ss,SSS"; + static final long NANOS_IN_ONE_SEC = 1000 * 1000 * 1000L; + + /** + * Create the test case + * + * @param testName + * name of the test case + */ + public TestDateFormatOriginal(String testName) { + super(testName); + } + + /** + * @return the suite of tests being tested + */ + public static Test suite() { + return new TestSuite(TestDateFormatOriginal.class); + } + + public static void main(String[] args) { + junit.textui.TestRunner.run(suite()); + } + + public void setUp() throws Exception { + super.setUp(); + } + + public void tearDown() throws Exception { + super.tearDown(); + } + + public void testRaw() throws Exception { + SimpleDateFormat simpleFormat = new SimpleDateFormat(ISO8601_PATTERN); + DateTimeFormatter jodaFormat = DateTimeFormat.forPattern(ISO8601_PATTERN); + + Date date = new Date(); + DateTime dateTime = new DateTime(date); + + long start = System.nanoTime(); + for (int i = 0; i < 100000; ++i) { + jodaFormat.print(dateTime); + } + long jodaAvg = (System.nanoTime() - start) / 100000; + + + start = System.nanoTime(); + for (int i = 0; i < 100000; ++i) { + simpleFormat.format(date); + } + long simpleAvg = (System.nanoTime() - start) / 100000; + + float diff = (((float) (simpleAvg - jodaAvg)) / simpleAvg) * 100; + System.out.println("Raw - JDK: " + simpleAvg + " ns Joda: " + jodaAvg + + " ns - Difference: " + diff + "%"); + } + + public void testSynchronized() throws Exception { + SynchronizedDateFormatter formatter = new SynchronizedDateFormatter(); + int threads = 10; + int iterations = 10000; + Thread[] formatThreads = new Thread[threads]; + Date date = new Date(); + + for (int i = 0; i < threads; i++) { + formatThreads[i] = new DateFormatThread(formatter, date, iterations); + } + long start = System.nanoTime(); + for (Thread thread : formatThreads) { + thread.start(); + } + for (Thread thread : formatThreads) { + thread.join(); + } + long end = System.nanoTime(); + double actual = ((double) (end - start)) / NANOS_IN_ONE_SEC; + System.out.printf("Synchronized DateFormat: %,.4f seconds\n", actual); + + } + + public void testUnSynchronized() throws Exception { + UnsynchronizedDateFormatter formatter = new UnsynchronizedDateFormatter(); + int threads = 10; + int iterations = 10000; + Thread[] formatThreads = new Thread[threads]; + Date date = new Date(); + + for (int i = 0; i < threads; i++) { + formatThreads[i] = new DateFormatThread(formatter, date, iterations); + } + long start = System.nanoTime(); + for (Thread thread : formatThreads) { + thread.start(); + } + for (Thread thread : formatThreads) { + thread.join(); + } + long end = System.nanoTime(); + double actual = ((double) (end - start)) / NANOS_IN_ONE_SEC; + System.out.printf("Unsynchronized DateFormat: %,.4f seconds\n", actual); + + } + + public void testThreadLocal() throws Exception { + ThreadLocalDateFormatter formatter = new ThreadLocalDateFormatter(); + int threads = 10; + int iterations = 10000; + Thread[] formatThreads = new Thread[threads]; + Date date = new Date(); + + for (int i = 0; i < threads; i++) { + formatThreads[i] = new DateFormatThread(formatter, date, iterations); + } + long start = System.nanoTime(); + for (Thread thread : formatThreads) { + thread.start(); + } + for (Thread thread : formatThreads) { + thread.join(); + } + long end = System.nanoTime(); + double actual = ((double) (end - start)) / NANOS_IN_ONE_SEC; + System.out.printf("ThreadLocal DateFormat: %,.4f seconds\n", actual); + + } + + public void testDateTimeFormatter() throws Exception { + int threads = 10; + int iterations = 10000; + Thread[] formatThreads = new DateTimeFormatThread[threads]; + JodaFormatter formatter = new JodaFormatter(); + Date date = new Date(); + DateTime dateTime = new DateTime(date); + + for (int i = 0; i < threads; i++) { + formatThreads[i] = new DateTimeFormatThread(formatter, dateTime, + iterations); + } + long start = System.nanoTime(); + for (Thread thread : formatThreads) { + thread.start(); + } + for (Thread thread : formatThreads) { + thread.join(); + } + long end = System.nanoTime(); + double actual = ((double) (end - start)) / NANOS_IN_ONE_SEC; + System.out.printf("Joda DateTimeFormatter: %,.4f seconds\n", actual); + + } + + public static interface Formatter { + String format(Date date); + } + + public static class SynchronizedDateFormatter implements Formatter { + SimpleDateFormat simpleFormat = new SimpleDateFormat(ISO8601_PATTERN); + + public synchronized String format(Date date) { + return simpleFormat.format(date); + } + } + + public static class UnsynchronizedDateFormatter implements Formatter { + public synchronized String format(Date date) { + return new SimpleDateFormat(ISO8601_PATTERN).format(date); + } + } + + public static class ThreadLocalDateFormatter implements Formatter { + ThreadLocal<SimpleDateFormat> formatter = new ThreadLocal<SimpleDateFormat>() { + protected synchronized SimpleDateFormat initialValue() { + return new SimpleDateFormat(ISO8601_PATTERN); + } + }; + + public String format(Date date) { + return formatter.get().format(date); + } + } + + public static class JodaFormatter { + DateTimeFormatter formatter = DateTimeFormat.forPattern(ISO8601_PATTERN); + + public String format(DateTime date) { + return formatter.print(date); + } + } + + public static class DateFormatThread extends Thread { + Formatter formatter; + Date date; + long iterCount; + + public DateFormatThread(Formatter f, Date date, long iterations) { + this.formatter = f; + this.date = date; + this.iterCount = iterations; + } + + public void run() { + for (int i = 0; i < iterCount; i++) { + formatter.format(this.date); + } + } + } + + public static class DateTimeFormatThread extends Thread { + JodaFormatter formatter; + DateTime date; + long iterCount; + + public DateTimeFormatThread(JodaFormatter f, DateTime date, long iterations) { + this.formatter = f; + this.date = date; + this.iterCount = iterations; + } + + public void run() { + for (int i = 0; i < iterCount; i++) { + formatter.format(this.date); + } + } + } +}

1 0

svn commit: r2291 - logback/trunk/logback-classic/src/test/java/ch/qos/logback/classic/issue/lbclassic36
by noreply.ceki＠qos.ch 14 Jun '09

14 Jun '09

Author: ceki Date: Sun Jun 14 21:48:08 2009 New Revision: 2291 Modified: logback/trunk/logback-classic/src/test/java/ch/qos/logback/classic/issue/lbclassic36/TestDateFormat.java Log: Guarded versions of the tests so as to resemble the code found the logback. Modified: logback/trunk/logback-classic/src/test/java/ch/qos/logback/classic/issue/lbclassic36/TestDateFormat.java ============================================================================== --- logback/trunk/logback-classic/src/test/java/ch/qos/logback/classic/issue/lbclassic36/TestDateFormat.java (original) +++ logback/trunk/logback-classic/src/test/java/ch/qos/logback/classic/issue/lbclassic36/TestDateFormat.java Sun Jun 14 21:48:08 2009 @@ -10,18 +10,21 @@ public static final String ISO8601_PATTERN = "yyyy-MM-dd HH:mm:ss,SSS"; static final long NANOS_IN_ONE_SEC = 1000 * 1000 * 1000L; - static long RUN_LENGTH = 1000*1000; + static long RUN_LENGTH = 1000 * 1000; public static void main(String[] args) { for (int i = 0; i < 5; i++) { doRawJoda(); - doRawSimpleDateFormat(); - doSyncronizedSimpleDateFormat(); + doRawSDF(); + doGuardedJoda(); + doGuardedSDF(); } - print("Raw Joda: ", doRawJoda()); - print("Sync SimpleDateFormat: ", doSyncronizedSimpleDateFormat()); - print("Raw SimpleDateFormat: ", doRawSimpleDateFormat()); + print("Raw Joda: ", doRawJoda()); + print("Raw SDF: ", doRawSDF()); + print("Guarded Joda: ", doGuardedJoda()); + print("Guarded SDF: ", doGuardedSDF()); + } static void print(String msg, double avg) { @@ -38,7 +41,23 @@ return (System.nanoTime() - start) * 1.0d / RUN_LENGTH; } - static double doRawSimpleDateFormat() { + static double doGuardedJoda() { + DateTimeFormatter jodaFormat = DateTimeFormat.forPattern(ISO8601_PATTERN); + long start = System.nanoTime(); + long cache = 0; + for (int i = 0; i < RUN_LENGTH; ++i) { + synchronized (jodaFormat) { + long now = System.currentTimeMillis(); + if (cache != now) { + cache = now; + jodaFormat.print(now); + } + } + } + return (System.nanoTime() - start) * 1.0d / RUN_LENGTH; + } + + static double doRawSDF() { SimpleDateFormat simpleFormat = new SimpleDateFormat(ISO8601_PATTERN); long timeInMillis = new Date().getTime(); long start = System.nanoTime(); @@ -48,16 +67,19 @@ return (System.nanoTime() - start) * 1.0d / RUN_LENGTH; } - static double doSyncronizedSimpleDateFormat() { + static double doGuardedSDF() { SimpleDateFormat simpleFormat = new SimpleDateFormat(ISO8601_PATTERN); - long timeInMillis = new Date().getTime(); long start = System.nanoTime(); + long cache = 0; for (int i = 0; i < RUN_LENGTH; ++i) { synchronized (simpleFormat) { - simpleFormat.format(timeInMillis); + long now = System.currentTimeMillis(); + if (cache != now) { + cache = now; + simpleFormat.format(now); + } } } return (System.nanoTime() - start) * 1.0d / RUN_LENGTH; } - }

1 0

[JIRA] Issue Comment Edited: (LBCLASSIC-36) Synchronization of SimpleDateFormat object in DateConverter
by Ceki Gulcu (JIRA) 14 Jun '09

14 Jun '09

[ http://jira.qos.ch/browse/LBCLASSIC-36?page=com.atlassian.jira.plugin.syste… ] Ceki Gulcu edited comment on LBCLASSIC-36 at 6/14/09 9:28 PM: -------------------------------------------------------------- Thank you for this contribution. I have ran a modified (and simplified) version of your tests. It looks like SimpleDateFormat and Yoda time have roughly the same performance. with the former having the edge. Your version of the test had Yoda time transform Date objects. However, in logback, a "long" as obtained by System.currentMillis(), needs to be transformed. Transforming Date objects is pretty much beside the point. was (Author: noreply.ceki(a)qos.ch) Thank you for this patch. I have ran a modified (and simplified) version of your tests. It looks like SimpleDateFormat and Yoda time have roughly the same performance. with the former having the edge. > Synchronization of SimpleDateFormat object in DateConverter > ----------------------------------------------------------- > > Key: LBCLASSIC-36 > URL: http://jira.qos.ch/browse/LBCLASSIC-36 > Project: logback-classic > Issue Type: Bug > Components: Other > Affects Versions: unspecified > Environment: Operating System: All > Platform: All > Reporter: Anders Hammar > Assignee: Ceki Gulcu > Attachments: DateConverter_patch.java, dateformattest.zip > > > According to the API documentation of SimpleDateFormat, date formats are not synchronized and it is recommended to create separate format instances for each thread. > However, a possibly simpler (mer efficient) solution would be to synchronize on the SimpleDateFormat object in the convert() method to prevent two threads to access it concurrently. -- This message is automatically generated by JIRA. - If you think it was sent incorrectly contact one of the administrators: http://jira.qos.ch/secure/Administrators.jspa - For more information on JIRA, see: http://www.atlassian.com/software/jira

1 0

[JIRA] Commented: (LBCLASSIC-36) Synchronization of SimpleDateFormat object in DateConverter
by Ceki Gulcu (JIRA) 14 Jun '09

14 Jun '09

[ http://jira.qos.ch/browse/LBCLASSIC-36?page=com.atlassian.jira.plugin.syste… ] Ceki Gulcu commented on LBCLASSIC-36: ------------------------------------- Thank you for this patch. I have ran a modified (and simplified) version of your tests. It looks like SimpleDateFormat and Yoda time have roughly the same performance. with the former having the edge. > Synchronization of SimpleDateFormat object in DateConverter > ----------------------------------------------------------- > > Key: LBCLASSIC-36 > URL: http://jira.qos.ch/browse/LBCLASSIC-36 > Project: logback-classic > Issue Type: Bug > Components: Other > Affects Versions: unspecified > Environment: Operating System: All > Platform: All > Reporter: Anders Hammar > Assignee: Ceki Gulcu > Attachments: DateConverter_patch.java, dateformattest.zip > > > According to the API documentation of SimpleDateFormat, date formats are not synchronized and it is recommended to create separate format instances for each thread. > However, a possibly simpler (mer efficient) solution would be to synchronize on the SimpleDateFormat object in the convert() method to prevent two threads to access it concurrently. -- This message is automatically generated by JIRA. - If you think it was sent incorrectly contact one of the administrators: http://jira.qos.ch/secure/Administrators.jspa - For more information on JIRA, see: http://www.atlassian.com/software/jira

1 0

svn commit: r2290 - in logback/trunk/logback-classic: . src/test/java/ch/qos/logback/classic/issue/lbclassic36
by noreply.ceki＠qos.ch 14 Jun '09

14 Jun '09

Author: ceki Date: Sun Jun 14 21:22:00 2009 New Revision: 2290 Added: logback/trunk/logback-classic/src/test/java/ch/qos/logback/classic/issue/lbclassic36/ logback/trunk/logback-classic/src/test/java/ch/qos/logback/classic/issue/lbclassic36/TestDateFormat.java Modified: logback/trunk/logback-classic/pom.xml Log: LBCLASSIC-36 Test performance of Yoda time compared to SimpleDateFormat. Here are sample results on my machine: Raw Joda: average tick 2714.052123 nanoseconds Sync SimpleDateFormat: average tick 2129.857639 nanoseconds Raw SimpleDateFormat: average tick 2138.306659 nanoseconds Conclusion: Yoda time and SimpleDataFormat have roughly equivalent performace, with the latter having the edge. Modified: logback/trunk/logback-classic/pom.xml ============================================================================== --- logback/trunk/logback-classic/pom.xml (original) +++ logback/trunk/logback-classic/pom.xml Sun Jun 14 21:22:00 2009 @@ -51,7 +51,7 @@ <version>${slf4j.version}</version> <scope>test</scope> </dependency> - + <dependency> <groupId>org.slf4j</groupId> <artifactId>slf4j-api</artifactId> @@ -134,6 +134,13 @@ </exclusion> </exclusions> </dependency> + + <dependency> + <groupId>joda-time</groupId> + <artifactId>joda-time</artifactId> + <version>1.6</version> + <scope>test</scope> + </dependency> </dependencies> <build> @@ -151,10 +158,12 @@ <plugin> <groupId>org.apache.maven.plugins</groupId> <artifactId>maven-jar-plugin</artifactId> - <configuration> - <archive> - <manifestFile>${project.build.outputDirectory}/META-INF/MANIFEST.MF</manifestFile> - </archive> + <configuration> + <archive> + <manifestFile> + ${project.build.outputDirectory}/META-INF/MANIFEST.MF + </manifestFile> + </archive> </configuration> <executions> <execution> @@ -191,21 +200,23 @@ <execution> <id>bundle-manifest</id> <phase>process-classes</phase> - <goals> + <goals> <goal>manifest</goal> - </goals> + </goals> </execution> </executions> <configuration> <instructions> <Export-Package>ch.qos.logback.classic.*</Export-Package> <Export-Package>org.slf4j.impl;version=1.5.6</Export-Package> - <Import-Package>sun.reflect;resolution:=optional, - javax.jms;resolution:=optional, - * + <Import-Package> + sun.reflect;resolution:=optional, javax.jms;resolution:=optional, + * </Import-Package> - <Bundle-RequiredExecutionEnvironment>J2SE-1.5</Bundle-RequiredExecutionEnvironment> + <Bundle-RequiredExecutionEnvironment> + J2SE-1.5 + </Bundle-RequiredExecutionEnvironment> </instructions> </configuration> </plugin> Added: logback/trunk/logback-classic/src/test/java/ch/qos/logback/classic/issue/lbclassic36/TestDateFormat.java ============================================================================== --- (empty file) +++ logback/trunk/logback-classic/src/test/java/ch/qos/logback/classic/issue/lbclassic36/TestDateFormat.java Sun Jun 14 21:22:00 2009 @@ -0,0 +1,63 @@ +package ch.qos.logback.classic.issue.lbclassic36; + +import java.text.SimpleDateFormat; +import java.util.Date; + +import org.joda.time.format.DateTimeFormat; +import org.joda.time.format.DateTimeFormatter; + +public class TestDateFormat { + public static final String ISO8601_PATTERN = "yyyy-MM-dd HH:mm:ss,SSS"; + static final long NANOS_IN_ONE_SEC = 1000 * 1000 * 1000L; + + static long RUN_LENGTH = 1000*1000; + + public static void main(String[] args) { + for (int i = 0; i < 5; i++) { + doRawJoda(); + doRawSimpleDateFormat(); + doSyncronizedSimpleDateFormat(); + } + + print("Raw Joda: ", doRawJoda()); + print("Sync SimpleDateFormat: ", doSyncronizedSimpleDateFormat()); + print("Raw SimpleDateFormat: ", doRawSimpleDateFormat()); + } + + static void print(String msg, double avg) { + System.out.println(msg + " average tick " + avg + " nanoseconds"); + } + + static double doRawJoda() { + DateTimeFormatter jodaFormat = DateTimeFormat.forPattern(ISO8601_PATTERN); + long timeInMillis = new Date().getTime(); + long start = System.nanoTime(); + for (int i = 0; i < RUN_LENGTH; ++i) { + jodaFormat.print(timeInMillis); + } + return (System.nanoTime() - start) * 1.0d / RUN_LENGTH; + } + + static double doRawSimpleDateFormat() { + SimpleDateFormat simpleFormat = new SimpleDateFormat(ISO8601_PATTERN); + long timeInMillis = new Date().getTime(); + long start = System.nanoTime(); + for (int i = 0; i < RUN_LENGTH; ++i) { + simpleFormat.format(timeInMillis); + } + return (System.nanoTime() - start) * 1.0d / RUN_LENGTH; + } + + static double doSyncronizedSimpleDateFormat() { + SimpleDateFormat simpleFormat = new SimpleDateFormat(ISO8601_PATTERN); + long timeInMillis = new Date().getTime(); + long start = System.nanoTime(); + for (int i = 0; i < RUN_LENGTH; ++i) { + synchronized (simpleFormat) { + simpleFormat.format(timeInMillis); + } + } + return (System.nanoTime() - start) * 1.0d / RUN_LENGTH; + } + +}

1 0

[JIRA] Commented: (LBCLASSIC-36) Synchronization of SimpleDateFormat object in DateConverter
by Ralph Goers (JIRA) 14 Jun '09

14 Jun '09

[ http://jira.qos.ch/browse/LBCLASSIC-36?page=com.atlassian.jira.plugin.syste… ] Ralph Goers commented on LBCLASSIC-36: -------------------------------------- Since you asked... I created a test to compare Joda Time to Simple Date Format. The results I get on my Mac are: Running TestDateFormat Raw - JDK: 3673 ns Joda: 2587 ns - Difference: 29.56711% Synchronized DateFormat: 0.4640 seconds Unsynchronized DateFormat: 2.2500 seconds ThreadLocal DateFormat: 0.2224 seconds Joda DateTimeFormatter: 0.1631 seconds Tests run: 5, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 3.871 sec The raw test just performs a date format 100,000 times in a loop on a single thread. The synchronized test behaves like Logback currently does, synchronizing calls to format the date. The unsynchronized uses a new SimpleDateTime for each invocation. The threadlocal test creates a SimpleDateFormat for each thread. The joda test uses Joda Time's DateTimeFormatter. > Synchronization of SimpleDateFormat object in DateConverter > ----------------------------------------------------------- > > Key: LBCLASSIC-36 > URL: http://jira.qos.ch/browse/LBCLASSIC-36 > Project: logback-classic > Issue Type: Bug > Components: Other > Affects Versions: unspecified > Environment: Operating System: All > Platform: All > Reporter: Anders Hammar > Assignee: Ceki Gulcu > Attachments: DateConverter_patch.java, dateformattest.zip > > > According to the API documentation of SimpleDateFormat, date formats are not synchronized and it is recommended to create separate format instances for each thread. > However, a possibly simpler (mer efficient) solution would be to synchronize on the SimpleDateFormat object in the convert() method to prevent two threads to access it concurrently. -- This message is automatically generated by JIRA. - If you think it was sent incorrectly contact one of the administrators: http://jira.qos.ch/secure/Administrators.jspa - For more information on JIRA, see: http://www.atlassian.com/software/jira

1 0