I am running all on virtualbox on fedora 23. Looks like one of the locator processes stuck.
I killed the process, than start it again and it worked – actually very simple. But since I had an exception when I ran the locator above I didn’t thought that although the exception the process remain stuck – kind of bug if you ask me. That’s it. Dor From: Jens Deppe [mailto:[email protected]] Sent: יום ב 25 ינואר 2016 16:53 To: [email protected] Subject: Re: failed building the latest version downloaded from github That's good news! What did you have to do to get it to work? --Jens On Mon, Jan 25, 2016 at 6:05 AM, Dor Ben Dov <[email protected]<mailto:[email protected]>> wrote: Yes , I succeeded to run both locator service with the default parameters as well as the server. Dor From: Jens Deppe [mailto:[email protected]<mailto:[email protected]>] Sent: יום ב 25 ינואר 2016 15:56 To: [email protected]<mailto:[email protected]> Subject: Re: failed building the latest version downloaded from github Hello Dor, Were you able to get this working correctly? If not can you also send the locator log which is created (even though the locator does not appear to start correctly). I was never able to reproduce this in the same way you had it failing. I tried in both a Fedora 23 Docker container as well as a VM without success. A slightly different question is why are you wanting to run the locator with sudo? There should be no need for that and should be discouraged under almost all circumstances. --Jens On Tue, Jan 19, 2016 at 9:11 AM, Kirk Lund <[email protected]<mailto:[email protected]>> wrote: Hi Dor, Each gemfire process owns its working directory in order to support file-based control which is the alternative to using the Attach API in the tools.jar. What this means is that GFSH will allow you to start only one gemfire process in a directory. By default the name of that directory is the name of your process -- "locator" is the name you used so the directory will also be "locator". I don't know why the 1st "start locator" command didn't return a status, but it did apparently start a locator as evidenced by the failure to start a 2nd locator in that directory. If "start locator" completes normally (by showing a status), then GFSH will auto-connect to that locator. By that I mean that GFSH will connect to it as a JMX RMI client. If this doesn't happen automatically, then you need to use the "connect" command in GFSH to connect to that locator before attempting to "start server". This is why your "start server" command failed. The 2nd "start locator" failed because the 1st one is still running in that sub-directory (default sub-dir name is "locator" because that's the name you specified). You would either have to kill the process via "kill 2997" or issue "stop locator --dir=locator" before starting a new gemfire process in that directory. -Kirk On Mon, Jan 18, 2016 at 10:54 PM, Dor Ben Dov <[email protected]<mailto:[email protected]>> wrote: Kirk Thanks first. Second, Looks locator is not raising up, but it’s a progress look like it is working … never ends although. _________________________ __ / _____/ ______/ ______/ /____/ / / / __/ /___ /_____ / _____ / / /__/ / ____/ _____/ / / / / /______/_/ /______/_/ /_/ v1.0.0-incubating-SNAPSHOT Monitor and Manage GemFire gfsh>start locator --name=locator Starting a GemFire Locator in /home/dor/Downloads/incubator-geode/gemfire-assembly/build/install/apache-geode/bin/locator... ................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................. Broken out of wait... the Locator process will continue to startup in the background. Then tried to raise the server although locator wasn’t started and looks like it is up gfsh>start server --name=server1 Starting a GemFire Server in /home/dor/Downloads/incubator-geode/gemfire-assembly/build/install/apache-geode/bin/server1... .................................. Server in /home/dor/Downloads/incubator-geode/gemfire-assembly/build/install/apache-geode/bin/server1 on 192.168.253.132[40404] as server1 is currently online. Process ID: 5195 Uptime: 18 seconds GemFire Version: 1.0.0-incubating-SNAPSHOT Java Version: 1.8.0_65 Log File: /home/dor/Downloads/incubator-geode/gemfire-assembly/build/install/apache-geode/bin/server1/server1.log JVM Arguments: -Dgemfire.use-cluster-configuration=true -XX:OnOutOfMemoryError=kill -KILL %p -Dgemfire.launcher.registerSignalHandlers=true -Djava.awt.headless=true -Dsun.rmi.dgc.server.gcInterval=9223372036854775806 Class-Path: /home/dor/Downloads/incubator-geode/gemfire-assembly/build/install/apache-geode/lib/gemfire-core-1.0.0-incubating-SNAPSHOT.jar:/home/dor/Downloads/incubator-geode/gemfire-assembly/build/install/apache-geode/lib/gemfire-core-dependencies.jar Then again tried to raise the locator but now it fails gfsh>start locator --name=locator Starting a GemFire Locator in /home/dor/Downloads/incubator-geode/gemfire-assembly/build/install/apache-geode/bin/locator... The Locator process terminated unexpectedly with exit status 1. Please refer to the log file in /home/dor/Downloads/incubator-geode/gemfire-assembly/build/install/apache-geode/bin/locator for full details. Exception in thread "main" java.lang.RuntimeException: A PID file already exists and a Locator may be running in /home/dor/Downloads/incubator-geode/gemfire-assembly/build/install/apache-geode/bin/locator on 192.168.253.132[10334]. at com.gemstone.gemfire.distributed.LocatorLauncher.start(LocatorLauncher.java:628) at com.gemstone.gemfire.distributed.LocatorLauncher.run(LocatorLauncher.java:513) at com.gemstone.gemfire.distributed.LocatorLauncher.main(LocatorLauncher.java:188) Caused by: com.gemstone.gemfire.internal.process.FileAlreadyExistsException: Pid file already exists: /home/dor/Downloads/incubator-geode/gemfire-assembly/build/install/apache-geode/bin/locator/vf.gf.locator.pid for process 2997 at com.gemstone.gemfire.internal.process.LocalProcessLauncher.writePid(LocalProcessLauncher.java:114) at com.gemstone.gemfire.internal.process.LocalProcessLauncher.<init>(LocalProcessLauncher.java:62) at com.gemstone.gemfire.internal.process.ControllableProcess.<init>(ControllableProcess.java:52) at com.gemstone.gemfire.distributed.LocatorLauncher.start(LocatorLauncher.java:594) ... 2 more Dor From: Kirk Lund [mailto:[email protected]<mailto:[email protected]>] Sent: יום ג 19 ינואר 2016 00:33 To: [email protected]<mailto:[email protected]> Subject: Re: failed building the latest version downloaded from github Hi Dor, I committed the fix for GEODE-780 "GFSH should not require Attach API to start Locators or Servers" to develop. You might want to try pulling, rebuilding and see if the start commands behave correctly for you now. Thanks, Kirk On Mon, Jan 18, 2016 at 8:25 AM, Jens Deppe <[email protected]<mailto:[email protected]>> wrote: Hi Dor, I'm unable to reproduce this on Fedora 23. To resolve this you'll have to modify the "gfsh" script to produce some debug output. Please edit the gfsh script and add " -x" to the first line: #!/bin/bash -x Then run the command I had suggested again: sudo sh -c "JAVA_HOME=/usr/java/latest /your/path/to/bin/gfsh" Please send the complete output of running that command. Thanks --Jens On Mon, Jan 18, 2016 at 3:06 AM, Dor Ben Dov <[email protected]<mailto:[email protected]>> wrote: Jen, Any news regard the issue I am having ? Dor From: Dor Ben Dov Sent: יום א 17 ינואר 2016 09:05 To: [email protected]<mailto:[email protected]> Subject: RE: failed building the latest version downloaded from github Tried but it’s not working Dor From: Jens Deppe [mailto:[email protected]] Sent: יום ו 15 ינואר 2016 16:05 To: [email protected]<mailto:[email protected]> Subject: Re: failed building the latest version downloaded from github Dor, Did you try and run the commands I had suggested? It looks like you just ran the original command again? --Jens On Thu, Jan 14, 2016 at 11:50 PM, Dor Ben Dov <[email protected]<mailto:[email protected]>> wrote: Guys Did that Same problem [dor@localhost apache-geode]$ sudo ./bin/gfsh _________________________ __ / _____/ ______/ ______/ /____/ / / / __/ /___ /_____ / _____ / / /__/ / ____/ _____/ / / / / /______/_/ /______/_/ /_/ v1.0.0-incubating-SNAPSHOT Monitor and Manage GemFire gfsh> start locator --name=locator Starting a GemFire Locator in /home/dor/Downloads/incubator-geode/gemfire-assembly/build/install/apache-geode/locator... An error occurred while attempting to start a Locator in /home/dor/Downloads/incubator-geode/gemfire-assembly/build/install/apache-geode/locator on 192.168.253.132[10334]: The Attach API classes could not be found on the classpath. Please include JDK tools.jar on the classpath or add the JDK tools.jar to the jre/lib/ext directory. From: Jens Deppe [mailto:[email protected]<mailto:[email protected]>] Sent: יום ה 14 ינואר 2016 21:57 To: [email protected]<mailto:[email protected]> Subject: Re: failed building the latest version downloaded from github The problem here is the difference between running as sudo vs. running as a regular user. If you run gfsh with shell debugging, you'd see that as a regular user, java is called with a full path but under sudo, it's just being called with 'java' (assuming your PATH variable is correct). In the latter case then, the tools.jar file does not get found. To work around the issue you should be able to do the following (setting paths as appropriate): export JAVA_HOME=/usr/lib/jvm/java-8-oracle sudo sh -c "JAVA_HOME=$JAVA_HOME /bin/path/to/gfsh" --Jens On Wed, Jan 13, 2016 at 2:52 PM, Kirk Lund <[email protected]<mailto:[email protected]>> wrote: This issue is now filed as "GEODE-780: GFSH should not require Attach API to start Locators or Servers" It's caused by GFSH attempting to perform status using the pid of the newly launched process. -Kirk On Wed, Jan 13, 2016 at 1:41 PM, John Blum <[email protected]<mailto:[email protected]>> wrote: > If you're running on Mac, then the contents from tools.jar should already be > in your java environment. This is only true if you are using Apple's JDK (JRE), which ceased to be supported and released after 1.6. In fact, the latest versions of Mac OS X, e.g. Yosemite and I think even Mavericks (perhaps) removed the Apple JDK. Since Geode requires JDK 8 you would need to use an Oracle provided Mac OS X JDK (since 1.7), in which case, all the same rules Kirk mentioned above this line apply. On Wed, Jan 13, 2016 at 1:33 PM, Kirk Lund <[email protected]<mailto:[email protected]>> wrote: In bash, you might check "which java" and then add the tools.jar to the lib/ext dir that's parallel to the bin containing java. If your java is in /somedir/jre/bin then you would copy a tools.jar from a jdk installation and add it to /somedir/jre/lib/ext. If your java is in /somedir/jdk/bin then it should already be finding tools.jar but I suppose you could try adding tools.jar to /somedir/jdk/jre/lib/ext. If you're running on Mac, then the contents from tools.jar should already be in your java environment. -Kirk On Wed, Jan 13, 2016 at 1:05 PM, Dor Ben Dov <[email protected]<mailto:[email protected]>> wrote: This is the part again [dor@localhost apache-geode]$ sudo ./bin/gfsh [sudo] password for dor: _________________________ __ / _____/ ______/ ______/ /____/ / / / __/ /___ /_____ / _____ / / /__/ / ____/ _____/ / / / / /______/_/ /______/_/ /_/ v1.0.0-incubating-SNAPSHOT Monitor and Manage GemFire gfsh>start locator --name=locator Starting a GemFire Locator in /home/dor/Downloads/incubator-geode/gemfire-assembly/build/install/apache-geode/locator... An error occurred while attempting to start a Locator in /home/dor/Downloads/incubator-geode/gemfire-assembly/build/install/apache-geode/locator on 192.168.253.132[10334]: The Attach API classes could not be found on the classpath. Please include JDK tools.jar on the classpath or add the JDK tools.jar to the jre/lib/ext directory. From: Kirk Lund [mailto:[email protected]<mailto:[email protected]>] Sent: יום ד 13 ינואר 2016 22:52 To: [email protected]<mailto:[email protected]> Subject: Re: failed building the latest version downloaded from github GFSH is a custom CLI tool for GemFire. It doesn't support linux commands such as sudo. You can add the tools.jar from the JDK to your classpath to fix the last problem you hit, however the code no longer requires the Attach API, so I'm surprised that you're seeing an error message about it. I'll try to fit in some time to do some experimenting with this later this week or next week. -Kirk On Wed, Jan 13, 2016 at 12:45 PM, Dor Ben Dov <[email protected]<mailto:[email protected]>> wrote: Anthony and Dan, Cloning from git and then building the develop worked But when I am running the locator I am getting this : gfsh>sudo start locator --name=locator Command "sudo start locator --name=locator" is not valid. gfsh> gfsh>exit Exiting... [dor@localhost apache-geode]$ sudo ./bin/gfsh [sudo] password for dor: _________________________ __ / _____/ ______/ ______/ /____/ / / / __/ /___ /_____ / _____ / / /__/ / ____/ _____/ / / / / /______/_/ /______/_/ /_/ v1.0.0-incubating-SNAPSHOT Monitor and Manage GemFire gfsh>start locator --name=locator Starting a GemFire Locator in /home/dor/Downloads/incubator-geode/gemfire-assembly/build/install/apache-geode/locator... An error occurred while attempting to start a Locator in /home/dor/Downloads/incubator-geode/gemfire-assembly/build/install/apache-geode/locator on 192.168.253.132[10334]: The Attach API classes could not be found on the classpath. Please include JDK tools.jar on the classpath or add the JDK tools.jar to the jre/lib/ext directory. From: Anthony Baker [mailto:[email protected]<mailto:[email protected]>] Sent: יום ד 13 ינואר 2016 17:02 To: [email protected]<mailto:[email protected]> Subject: Re: failed building the latest version downloaded from github This should be fixed now: https://github.com/apache/incubator-geode/commit/0a890a56232e144d1dd76df05576bcdbb7601c87 Anthony On Jan 12, 2016, at 10:45 PM, Dor Ben Dov <[email protected]<mailto:[email protected]>> wrote: Dan I will try cloning from git repository and then I will try again and update you guys, but if you ask me, it’s weird that the downloaded zip doesn’t compile with gradle. Dor From: Dan Smith [mailto:[email protected]] Sent: יום ד 13 ינואר 2016 03:04 To: [email protected]<mailto:[email protected]> Subject: Re: failed building the latest version downloaded from github Can you try cloning the git repository instead of downloading the source zip? It looks like it's trying to read metadata from a git checkout. -Dan On Tue, Jan 12, 2016 at 1:04 PM, Dor Ben Dov <[email protected]<mailto:[email protected]>> wrote: Hi, I am using fedora 23 with gradle and java 8 Downloaded the latest bundle from github https://github.com/apache/incubator-geode/archive/develop.zip Then tried to gradle build according to the guidelines you wrote on the same page, But it keep failing with this error: dor@localhost incubator-geode-develop]$ sudo ./gradlew build installDist Starting a new Gradle Daemon for this build (subsequent builds will be faster). FAILURE: Build failed with an exception. * Where: Build file '/home/dor/Downloads/incubator-geode-develop/build.gradle' line: 57 * What went wrong: A problem occurred evaluating root project 'gemfire'. > Failed to apply plugin [id 'org.ajoberstar.grgit'] > One of setGitDir or setWorkTree must be called. * Try: Run with --stacktrace option to get the stack trace. Run with --info or --debug option to get more log output. BUILD FAILED Total time: 10.757 secs [dor@localhost incubator-geode-develop]$ Any ideas? What am I doing wrong ? Regards, Dor Ben Dovd This message and the information contained herein is proprietary and confidential and subject to the Amdocs policy statement, you may review at http://www.amdocs.com/email_disclaimer.asp -- -John 503-504-8657<tel:503-504-8657> john.blum10101 (skype)
