Displaying #maven-dev/2017-03-04.log:

Sat Mar 4 00:47:16 2017  hboutemy:Joined the channel
Sat Mar 4 01:38:50 2017  theRealGent:Joined the channel
Sat Mar 4 04:27:26 2017  mbenson__:Joined the channel
Sat Mar 4 04:30:26 2017  mbenson:Joined the channel
Sat Mar 4 04:30:26 2017  mbenson:Joined the channel
Sat Mar 4 07:19:21 2017  tibor_:Joined the channel
Sat Mar 4 08:38:02 2017  hboutemy:Joined the channel
Sat Mar 4 09:39:23 2017  Michael-O:Joined the channel
Sat Mar 4 10:02:25 2017  tibor_:Joined the channel
Sat Mar 4 10:31:39 2017  Michael-O:Joined the channel
Sat Mar 4 10:38:38 2017  Michael-O:Joined the channel
Sat Mar 4 10:40:46 2017  rfscholte:Joined the channel
Sat Mar 4 10:43:24 2017  rfscholte:hboutemy: stephenc: aheritier: I'm trying to release modello, but I'm getting an Access Denied. Could one of you verify the staging profiles under https://oss.sonatype.org/#stagingProfiles and see if there is a separate one for modello?
Sat Mar 4 12:01:21 2017  rfscholte:Joined the channel
Sat Mar 4 12:46:59 2017  Michael-O:Joined the channel
Sat Mar 4 13:30:07 2017  rfscholte:Left the channel
Sat Mar 4 13:30:27 2017  rfscholte:Joined the channel
Sat Mar 4 15:15:33 2017  theRealGent:Joined the channel
Sat Mar 4 15:38:59 2017  hboutemy:rfscholte: I see a org.codehaus.modello staging profile that is different from Plexus
Sat Mar 4 15:51:15 2017  rfscholte:right, as expected
Sat Mar 4 15:51:42 2017  rfscholte:I've asked for permission on modello at sonatype
Sat Mar 4 15:52:13 2017  rfscholte:normally should have it right after the weekend
Sat Mar 4 17:38:28 2017  tibor_:Joined the channel
Sat Mar 4 20:08:51 2017  Michael-O:Joined the channel
Sat Mar 4 21:16:55 2017  tibor_:Michael-O: I am just extending the small test code. The 2.19.2-experimental has already the logger and I commented out the main().
Sat Mar 4 21:17:24 2017  Michael-O:Alright
Sat Mar 4 21:22:50 2017  Michael-O:I will run the tests with your updates
Sat Mar 4 21:26:44 2017  Michael-O:Joined the channel
Sat Mar 4 21:36:48 2017  Michael-O:Joined the channel
Sat Mar 4 21:42:43 2017  Michael-O:tibor_: tests done
Sat Mar 4 21:42:47 2017  Michael-O:Did you expect more errors?
Sat Mar 4 21:43:00 2017  tibor_:yes
Sat Mar 4 21:43:14 2017  tibor_:because the main is empty of course
Sat Mar 4 21:43:33 2017  tibor_:but more important is what the logs tell me
Sat Mar 4 21:44:18 2017  Michael-O:updated tarball
Sat Mar 4 21:44:35 2017  tibor_:I am just preparing sup for tomorrow
Sat Mar 4 21:44:48 2017  tibor_:I will download it..
Sat Mar 4 21:51:03 2017  Michael-O:Joined the channel
Sat Mar 4 21:57:57 2017  Michael-O:Joined the channel
Sat Mar 4 21:58:23 2017  Michael-O:tibor_, do you have it?
Sat Mar 4 21:58:33 2017  tibor_:yes
Sat Mar 4 21:58:37 2017  tibor_:looks good
Sat Mar 4 21:58:57 2017  tibor_:ProcessBuilder.start() took only 3 millis
Sat Mar 4 22:00:05 2017  Michael-O:looks fast to me
Sat Mar 4 22:00:19 2017  Michael-O:I think that we need the ACK at the start of the VM too
Sat Mar 4 22:01:13 2017  tibor_:It exists
Sat Mar 4 22:01:31 2017  tibor_:the jvm sends event next_test
Sat Mar 4 22:01:47 2017  tibor_:initially jvm sends all sys props.
Sat Mar 4 22:02:21 2017  Michael-O:but does the Forker acknowledges that?
Sat Mar 4 22:02:45 2017  tibor_:now I have pushed a change, where timers and the reader Thread run.
Sat Mar 4 22:08:45 2017  Michael-O:running
Sat Mar 4 22:19:50 2017  Michael-O:Joined the channel
Sat Mar 4 22:21:06 2017  Michael-O:tibor_, tarball updated
Sat Mar 4 22:22:08 2017  tibor_:ok
Sat Mar 4 22:27:38 2017  tibor_:Something is here wrong.
Sat Mar 4 22:27:58 2017  tibor_:I will compare the same test with previous log.
Sat Mar 4 22:30:48 2017  Michael-O:Different behavior?
Sat Mar 4 22:38:28 2017  tibor_:no
Sat Mar 4 22:38:36 2017  tibor_:no different behavior
Sat Mar 4 22:38:41 2017  tibor_:I will show you
Sat Mar 4 22:41:11 2017  Michael-O:ok...
Sat Mar 4 22:41:27 2017  tibor_:1488663649962 OutputStreamFlushableCommandline#2092055 before starting forked JVM 1488663653154
Sat Mar 4 22:41:28 2017  tibor_:OutputStreamFlushableCommandline#2092055 after started forked JVM
Sat Mar 4 22:41:41 2017  tibor_:sorry
Sat Mar 4 22:41:50 2017  tibor_:1488663653154 OutputStreamFlushableCommandline#2092055 after started forked JVM
Sat Mar 4 22:42:09 2017  tibor_:this is exactly 3 sec delay we had only in main()
Sat Mar 4 22:42:29 2017  tibor_:So it looks like the JVM is implemented like this.
Sat Mar 4 22:42:52 2017  Michael-O:What does this mean?
Sat Mar 4 22:42:56 2017  tibor_:start() method does take as long time as our console appl runs
Sat Mar 4 22:43:22 2017  tibor_:we can confirm with our sample code tomorrow
Sat Mar 4 22:44:13 2017  tibor_:Our sample code without Utils have different experience
Sat Mar 4 22:44:22 2017  tibor_:the difference is the command
Sat Mar 4 22:44:35 2017  Michael-O:Is this something we didn't expect in our code, incomplete?
Sat Mar 4 22:45:18 2017  tibor_:yes
Sat Mar 4 22:45:20 2017  tibor_:look
Sat Mar 4 22:45:33 2017  tibor_:this is command in Surefire when started fork
Sat Mar 4 22:45:41 2017  tibor_:wait
Sat Mar 4 22:45:57 2017  tibor_: /bin/sh, -c, cd /usr/home/osipovmi/Projekte/maven-surefire/surefire-integration-tests/target/ForkModeIT_testForkCountTwoReuse && /usr/local/openjdk7/jre/bin/java -jar /usr/home/osipovmi/Projekte/maven-surefire/surefire-integration-tests/target/ForkModeIT_testForkCountTwoReuse/target/surefire/surefirebooter591651615184671210.jar /usr/home/osipovmi/Projekte/maven-surefire/surefire-integration-tests/target/ForkModeIT_testForkCountT
Sat Mar 4 22:46:06 2017  tibor_:and sample code has:
Sat Mar 4 22:46:42 2017  tibor_:String[] cmd = {"/bin/sh", "-c", "java -jar fork.jar"};
Sat Mar 4 22:47:11 2017  Michael-O:is the cd an issue?
Sat Mar 4 22:47:20 2017  Michael-O:ProcessBuilder allows us to set the directory directory
Sat Mar 4 22:47:34 2017  tibor_:I dont know because we have not check it out yet in sample code
Sat Mar 4 22:47:45 2017  tibor_:possible
Sat Mar 4 22:47:57 2017  tibor_:We will try
Sat Mar 4 22:48:41 2017  Michael-O:The JVM may never start because the command is invalid
Sat Mar 4 22:48:51 2017  tibor_:invalid?
Sat Mar 4 22:48:57 2017  tibor_:what you mean
Sat Mar 4 22:49:39 2017  tibor_:this is maybe better string
Sat Mar 4 22:49:41 2017  tibor_: /bin/sh -c cd /usr/home/osipovmi/Projekte/maven-surefire/surefire-integration-tests/target/ForkModeIT_testForkCountTwoReuse && /usr/local/openjdk7/jre/bin/java -jar /usr/home/osipovmi/Projekte/maven-surefire/surefire-integration-tests/target/ForkModeIT_testForkCountTwoReuse/target/surefire/surefirebooter591651615184671210.jar
Sat Mar 4 22:50:01 2017  Michael-O:See http://stackoverflow.com/q/17216049/696632
Sat Mar 4 22:50:09 2017  Michael-O:this is a similar case
Sat Mar 4 22:50:57 2017  Michael-O:http://stackoverflow.com/q/12268550/696632
Sat Mar 4 22:51:34 2017  tibor_:I see
Sat Mar 4 22:52:03 2017  tibor_:but the ProcessBuilder has a method for working directory
Sat Mar 4 22:52:15 2017  tibor_:this is added maybe by Utils
Sat Mar 4 22:52:35 2017  tibor_:the cd.. is maybe not necessary
Sat Mar 4 22:52:41 2017  tibor_:WDYT?
Sat Mar 4 22:53:07 2017  tibor_:but cd should be fast
Sat Mar 4 22:53:41 2017  Michael-O:Yes, this is my assumption. It is not the speed, but the exit code or the way the second (java) command treats &&
Sat Mar 4 22:53:48 2017  Michael-O:same is written in the fist SO questin
Sat Mar 4 22:53:52 2017  Michael-O:question*
Sat Mar 4 22:54:02 2017  tibor_:we can try now
Sat Mar 4 22:55:15 2017  Michael-O:https://www.google.de/search?q=java+processbuilder++multiple+commands&ie=utf-8&oe=utf-8&client=firefox-b&gfe_rd=cr&ei=VUS7WMSjKYTe8gfvhYrgBw
Sat Mar 4 22:55:18 2017  Michael-O:full of questions
Sat Mar 4 22:56:17 2017  tibor_:so ";"?
Sat Mar 4 22:56:21 2017  tibor_:String[] cmd = {"/bin/sh", "-c", "java -jar fork.jar"};
Sat Mar 4 22:56:30 2017  tibor_:would we ass cd here?
Sat Mar 4 22:56:34 2017  tibor_:add
Sat Mar 4 22:57:07 2017  Michael-O:No, use https://docs.oracle.com/javase/7/docs/api/java/lang/ProcessBuilder.html#directory(java.io.File)
Sat Mar 4 22:57:13 2017  tibor_:or you want to modify sample code?
Sat Mar 4 22:57:33 2017  tibor_:it is there, it was
Sat Mar 4 22:57:46 2017  tibor_:but Utils has it too
Sat Mar 4 22:58:15 2017  Michael-O:if so, why is there still a cd?
Sat Mar 4 22:58:36 2017  tibor_:but if we add cd .. && in sample code we would have chance to confirm bug in Utils
Sat Mar 4 22:58:57 2017  tibor_:I think cd is useless
Sat Mar 4 22:59:00 2017  tibor_:in Utils
Sat Mar 4 23:00:34 2017  tibor_:The maven-shared-utils printed the command I showed you from toString() method
Sat Mar 4 23:00:40 2017  tibor_:return StringUtils.join( getShellCommandline( true ), " " );
Sat Mar 4 23:00:48 2017  tibor_:not by Java ProcessBuilder
Sat Mar 4 23:01:51 2017  Michael-O:let me try something here
Sat Mar 4 23:02:43 2017  tibor_:BourneShell.java:
Sat Mar 4 23:02:45 2017  tibor_: sb.append( "cd " ); sb.append( unifyQuotes( dir ) ); sb.append( " && " );
Sat Mar 4 23:03:18 2017  tibor_:in Utils
Sat Mar 4 23:03:25 2017  Michael-O:Yes, this needs to be replaced with the process builder
Sat Mar 4 23:03:32 2017  tibor_:yes
Sat Mar 4 23:03:57 2017  Michael-O:Just tried
Sat Mar 4 23:04:00 2017  Michael-O:the cd works in freebsd-issue
Sat Mar 4 23:04:52 2017  tibor_:what else is different
Sat Mar 4 23:04:58 2017  tibor_:path to java?
Sat Mar 4 23:05:14 2017  tibor_:or arguments for main class?
Sat Mar 4 23:05:48 2017  tibor_:The length of command > 200 characters?
Sat Mar 4 23:06:00 2017  Michael-O:I added cd with userDir and appended target
Sat Mar 4 23:06:17 2017  Michael-O:the command runs successfully without changing to target manually
Sat Mar 4 23:08:31 2017  Michael-O:I do not think that the lengths is an issue
Sat Mar 4 23:08:35 2017  tibor_: /target/fork.jar is what you changed?
Sat Mar 4 23:08:59 2017  Michael-O:userDir.getCanonicalFile(= + "/target/fork.jar"
Sat Mar 4 23:09:13 2017  tibor_:Hey
Sat Mar 4 23:09:16 2017  tibor_:good idea
Sat Mar 4 23:09:34 2017  tibor_:lets copy /target/surefire
Sat Mar 4 23:10:02 2017  tibor_:and run same command, but the number hash will be different
Sat Mar 4 23:10:25 2017  tibor_:we can run it from sample code
Sat Mar 4 23:11:01 2017  tibor_:if you prolong the delay to 10 sec, you have chance to copy it.
Sat Mar 4 23:11:37 2017  Michael-O:Just checked ForkModeIT_testForkModeOnce
Sat Mar 4 23:12:06 2017  Michael-O:it appears that forked VM ran everything successfully
Sat Mar 4 23:12:31 2017  Michael-O:I really think that the stdio communication is doomed
Sat Mar 4 23:12:53 2017  tibor_:what you mean "it appears that forked VM ran everything successfully"
Sat Mar 4 23:13:48 2017  Michael-O:2017-03-04T11-15-09_961.dumpstream says:
Sat Mar 4 23:13:57 2017  Michael-O:# Created on 2017-03-04T11:15:09.962
Sat Mar 4 23:13:57 2017  Michael-O:Mrz 04, 2017 11:15:09 PM org.apache.maven.surefire.booter.ForkedBooter main
Sat Mar 4 23:13:57 2017  Michael-O:# Created on 2017-03-04T11:15:09.962
Sat Mar 4 23:13:57 2017  Michael-O:INFORMATION: ForkedBooter.main() :: Forked JVM started.
Sat Mar 4 23:13:57 2017  Michael-O:# Created on 2017-03-04T11:15:13.013
Sat Mar 4 23:13:57 2017  Michael-O:Mrz 04, 2017 11:15:13 PM org.apache.maven.surefire.booter.ForkedBooter main
Sat Mar 4 23:13:59 2017  Michael-O:# Created on 2017-03-04T11:15:13.014
Sat Mar 4 23:14:01 2017  Michael-O:INFORMATION: ForkedBooter.main() :: Forked JVM finished.
Sat Mar 4 23:15:07 2017  tibor_:It should not be here
Sat Mar 4 23:15:21 2017  tibor_:it should be handled by Java Logger API
Sat Mar 4 23:15:30 2017  tibor_:in <number>.log
Sat Mar 4 23:16:33 2017  Michael-O:I have this one too: 5569901946719381.log
Sat Mar 4 23:16:40 2017  Michael-O:Mrz 04, 2017 11:15:09 PM org.apache.maven.surefire.booter.ForkedBooter main
Sat Mar 4 23:16:40 2017  Michael-O:INFORMATION: ForkedBooter.main() :: Forked JVM started.
Sat Mar 4 23:16:40 2017  Michael-O:Mrz 04, 2017 11:15:13 PM org.apache.maven.surefire.booter.ForkedBooter main
Sat Mar 4 23:16:40 2017  Michael-O:INFORMATION: ForkedBooter.main() :: Forked JVM finished.
Sat Mar 4 23:17:04 2017  tibor_:that's right
Sat Mar 4 23:17:54 2017  Michael-O:So the forked VM is absolutely fine and did its job?
Sat Mar 4 23:18:18 2017  tibor_:it always runs
Sat Mar 4 23:18:29 2017  tibor_:but the problem is different
Sat Mar 4 23:19:08 2017  Michael-O:1488665709962 StreamPumper 2394337 line=INFORMATION: ForkedBooter.main() :: Forked JVM started.
Sat Mar 4 23:19:19 2017  Michael-O:I do not see the counterpart in log.txt
Sat Mar 4 23:21:58 2017  tibor_:I will add Utils in the test tomorrow
Sat Mar 4 23:23:09 2017  tibor_:Concentrate on these lines in the log:
Sat Mar 4 23:23:11 2017  tibor_:before starting forked JVM
Sat Mar 4 23:23:19 2017  tibor_:after started forked JVM
Sat Mar 4 23:23:34 2017  tibor_:You will see it in Utils
Sat Mar 4 23:24:08 2017  tibor_:in Commandline.java
Sat Mar 4 23:24:30 2017  Michael-O:Can you add who is forker and who is forked?
Sat Mar 4 23:24:36 2017  Michael-O:Just a prefix to the sysout?
Sat Mar 4 23:24:52 2017  tibor_:Since I committed delay=3sec in main() the difference between these two lines must not be 3sec.
Sat Mar 4 23:25:19 2017  tibor_:ok I will
Sat Mar 4 23:27:36 2017  tibor_:but log.txt now has only "forker"
Sat Mar 4 23:27:54 2017  tibor_:the <number>.log has forked
Sat Mar 4 23:28:47 2017  Michael-O:yes, but not log.txt
Sat Mar 4 23:28:58 2017  Michael-O:It is hard for me to read who sent what
Sat Mar 4 23:29:46 2017  Michael-O:ForkModeOnce log.txt looks good to me. But the forker declares the VM to be dead even the exit code is 0. So some message does not arrive properly.
Sat Mar 4 23:30:57 2017  tibor_:we commented out ForkedBooter.main() and it is "forked"
Sat Mar 4 23:36:50 2017  tibor_:Michael please see what the method is doing ProcessBuilder.start() : Process, and then when and how streams are observed.
Sat Mar 4 23:36:57 2017  tibor_:This is key point.
Sat Mar 4 23:39:32 2017  Michael-O:You mean tomorrow after your change?
Sat Mar 4 23:41:10 2017  tibor_:tomorrow
Sat Mar 4 23:42:01 2017  Michael-O:Alright
Sat Mar 4 23:52:06 2017  tibor_:I am trying to understand why CLI hashCode() is printed twice
Sat Mar 4 23:52:07 2017  tibor_:1488663649962 OutputStreamFlushableCommandline#2092055 before starting forked JVM
Sat Mar 4 23:52:15 2017  tibor_:the same
Sat Mar 4 23:54:49 2017  Michael-O:Which test?

Comments