how to run TachyonWorker in tachyon0.8.0&hadoop1.0.4?

[renzhix...@gmail.com]

Hi everyone!

     I want to establish the tachyon based on hadoop in one machine.My enviroment is using tachyon0.8.0 & hadoop1.0.4.When I run the comand "bin/tachyon-start.sh all Mount",I can

see the result as follows:

[root@de24 tachyon-0.8.0]# bin/tachyon-start.sh all Mount

Killed 0 processes on de24

Killed 0 processes on de24

Connecting to de24 as root...

Killed 0 processes on de24

Connection to de24 closed.

Starting master @ de24

Connecting to de24 as root...

Formatting RamFS: /mnt/ramdisk (1gb)

Starting worker @ de24

Connection to de24 closed.

[root@de24 tachyon-0.8.0]# jps

14213 TaskTracker

13794 DataNode

13951 SecondaryNameNode

19256 Jps

14058 JobTracker

19067 TachyonMaster

The result is not ture,because there is no TachyonWorker running. So I catch the file /home/xiaorz/tachyon-0.8.0/logs/worker.log,which content is listed below,

2015-10-26 17:54:55,930 INFO  internal.PlatformDependent (Slf4JLogger.java:info) - Your platform does not provide complete low-level API for accessing direct buffers reliably. Unless explicitly requested, heap buffer will always be preferred to avoid potential system unstability.

2015-10-26 17:54:56,031 INFO   (FileUtils.java:createStorageDirPath) - Folder /mnt/ramdisk/tachyonworker was created!

2015-10-26 17:54:56,050 ERROR WORKER_LOGGER (MetricsConfig.java:loadConfigFile) - Error loading metrics configuration file.

2015-10-26 17:54:56,097 ERROR WORKER_LOGGER (TachyonWorker.java:main) - Failed to initialize the block worker, exiting.

java.lang.RuntimeException: java.lang.reflect.InvocationTargetException

at com.google.common.base.Throwables.propagate(Throwables.java:160)

at tachyon.worker.DataServer$Factory.createDataServer(DataServer.java:42)

at tachyon.worker.block.BlockWorker.<init>(BlockWorker.java:182)

at tachyon.worker.TachyonWorker.main(TachyonWorker.java:42)

Caused by: java.lang.reflect.InvocationTargetException

at sun.reflect.NativeConstructorAccessorImpl.newInstance0(Native Method)

at sun.reflect.NativeConstructorAccessorImpl.newInstance(NativeConstructorAccessorImpl.java:57)

at sun.reflect.DelegatingConstructorAccessorImpl.newInstance(DelegatingConstructorAccessorImpl.java:45)

at java.lang.reflect.Constructor.newInstance(Constructor.java:532)

at tachyon.util.CommonUtils.createNewClassInstance(CommonUtils.java:138)

at tachyon.worker.DataServer$Factory.createDataServer(DataServer.java:37)

... 2 more

Caused by: java.lang.IllegalStateException: failed to create a child event loop

at io.netty.util.concurrent.MultithreadEventExecutorGroup.<init>(MultithreadEventExecutorGroup.java:68)

at io.netty.channel.MultithreadEventLoopGroup.<init>(MultithreadEventLoopGroup.java:49)

at io.netty.channel.epoll.EpollEventLoopGroup.<init>(EpollEventLoopGroup.java:61)

at io.netty.channel.epoll.EpollEventLoopGroup.<init>(EpollEventLoopGroup.java:49)

at tachyon.util.network.NettyUtils.createEventLoop(NettyUtils.java:55)

at tachyon.worker.netty.NettyDataServer.createBootstrapOfType(NettyDataServer.java:146)

at tachyon.worker.netty.NettyDataServer.createBootstrap(NettyDataServer.java:73)

at tachyon.worker.netty.NettyDataServer.<init>(NettyDataServer.java:54)

... 8 more

Caused by: java.lang.NullPointerException

at io.netty.util.internal.PlatformDependent0.allocateMemory(PlatformDependent0.java:373)

at io.netty.util.internal.PlatformDependent.allocateMemory(PlatformDependent.java:197)

at io.netty.channel.epoll.EpollEventArray.allocate(EpollEventArray.java:56)

at io.netty.channel.epoll.EpollEventArray.<init>(EpollEventArray.java:52)

at io.netty.channel.epoll.EpollEventLoop.<init>(EpollEventLoop.java:65)

at io.netty.channel.epoll.EpollEventLoopGroup.newChild(EpollEventLoopGroup.java:76)

at io.netty.util.concurrent.MultithreadEventExecutorGroup.<init>(MultithreadEventExecutorGroup.java:64)

... 15 more

So what can I do to solve this problem? Thank you very much !

[Calvin Jia]

Hi,

Which tarball did you download?

Thanks,

Calvin

[renzhix...@gmail.com]

Hi Calvin！

    Certainly！ I have download the tarball software. In the version of hadoop1.0.4& tachyon0.7.1, I can make the test successfully when

I use the comand"./bin/tachyon  runTests".Are you encounter this problem?

 

Thanks,

Renzhi Xiao

在 2015年10月27日星期二 UTC+8上午12:09:19，Calvin Jia写道：

[renzhix...@gmail.com]

Hi guys！

    Additional information， my Linux version is REHL6.2 & jdk 1.7. If you can find the

root of this problem，please contact me，thanks！

Yours sincerely，

Thomas XIAO

在 2015年10月26日星期一 UTC+8下午10:41:55，renzhix...@gmail.com写道：

[Bin Fan]

Hi Xiao,

It seems the worker failed to initialize the dataserver.

did you change any setting regarding data server 

(e.g., by default, we have tachyon.worker.data.server.class=tachyon.worker.netty.NettyDataServer), or you simply run the test from a clean tarball?

Thanks

- Bin

--
You received this message because you are subscribed to the Google Groups "Tachyon Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to tachyon-user...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[renzhix...@gmail.com]

Hi Bin，

    Maybe，I can passed the command “./bin/tachyon runTests” when I use hadoop1.0.4 & tachyon0.7.1.

But when I test the tachyon0.8.0,I have encountered this problem. I don't know the real reason. I thank that I

should read the official documents more cafefully next in order to avoid some important messages.

Thanks!

Renzhi Xiao

在 2015年10月27日星期二 UTC+8下午10:57:06，Bin Fan写道：

[Bin Fan]

Hi Xiao,

Sorry for the confusion. I was asking the parameter because if it is set to something wrong by mistake, that could cause the error you encounter.

So I was trying to understand if you are running Tachyon with default configuration. In theory, Tachyon should work with default configuration setting with your environment.

After digging into the code, I think the real cause is dataserver failed to initialize as I mentioned.

But the reason is more likely to have something to do the network address binding.

Data server failed to bind to a network address. Is there anything tricky in your network?

- Bin

[renzhix...@gmail.com]

Hi Bin，

    Maybe，I can successfully run tachyon with the local filesystem '/tmp'. I will try more next in running tachyon on Virtual Box ! It is well be 

appreciated if you can send more information about how to use the gluster based on hadoop ,my email is xiaoren...@163.com.Thanks again.

Best wishes

Renzhi Xiao

在 2015年10月28日星期三 UTC+8上午12:44:09，Bin Fan写道：

[Bin Fan]

Hi Renzhi,

Is the document here useful?

http://tachyon-project.org/documentation/Configuring-Tachyon-with-GlusterFS.html

- Bin

[renzhix...@gmail.com]

Hi Bin，

    Telling you a good news ,I have passed the tests based on hadoop2.4.1 cluster  on Tachyon0.8.0 in VirtualBox5.0 locally.

When create this topic "how to run TachyonWorker in tachyon0.8.0&hadoop1.0.4?",then I just

encountered only no TachyonWorker running in physical machine cluster .Sorry for the trouble.  Thanks for your help!

Today, I am very happy for that I can passed all the tests in 3 vitual machine guest OS in locally machine.There are some part

of the result as follows:

One TachyonMaster & one TachyonWorker on de21 virtual machine,

Each one TachyonWorker on de22 & de23 virtual machine.

So sorry for the late reply. Thanks again! 

Best wishes

Renzhi

在 2015年10月28日星期三 UTC+8下午12:43:40，Bin Fan写道：

[renzhix...@gmail.com]

Hi all,

 Additionally ,I have solved one error encountered in VirtualBox5.0.The error as follows:

Exception in thread "main" java.lang.OutofMemoryError:java heap space,

I can find the reason why it is wrong,because the java heap default space is 64MB in virtual machine in VirtualBox.

So I changed the java heap space useing the command is "export _JAVA_OPTIONS="-Xmx1g"",

after this command , all 9 tests haved passed through lately.

Cheers.

Renzhi 

在 2015年10月26日星期一 UTC+8下午10:41:55，renzhix...@gmail.com写道：

Hi everyone!

[Haoyuan Li]

Great. Thanks for letting us know!

Haoyuan

[Bin Fan]

Really appreciate your feedback and solution!

Meanwhile we will definitely improve the documentation on the website!

Thanks,

- Bin

[renzhix...@gmail.com]

Hi Bin，

    You are welcome. But I still have a confusion, I just see your official configureation file only  running Tachyon cluster on VirtualBox.

My question is that whether do you configure it successfully in physical machine cluster instead in virtual machine.Looking forward to

your reply. Thank you.

Best wishes

Renzhi

在 2015年11月5日星期四 UTC+8上午1:42:31，Bin Fan写道：

[renzhix...@gmail.com]

Hi haoyuan,

     Not at all.  I have already send a private email to your google email. You can see the problem if you have spare time.

I just don't know why it is some wrong in physical machine cluster. Please see the private email which is sent from xiaoren...@163.com.

Thanks!

Renzhi

在 2015年11月5日星期四 UTC+8上午1:38:56，Haoyuan Li写道：