JPPF, java, parallel computing, distributed computing, grid computing, parallel, distributed, cluster, grid, cloud, open source, android, .net
JPPF

The open source
grid computing
solution

 Home   About   Features   Download   Documentation   On Github   Forums 
April 14, 2021, 10:43:41 PM *
Welcome,
Please login or register.

Login with username, password and session length
Advanced search  
News: New users, please read this message. Thank you!
  Home Help Search Login Register  
Pages: [1]   Go Down

Author Topic: It seems that nodes can not work if I restart the server?  (Read 2155 times)

javacfish

  • Guest
It seems that nodes can not work if I restart the server?
« on: August 09, 2013, 11:56:32 AM »

Hi lolo,

Thanks for your help. Now I meet another problem:

I modify the configure of server. In order to let the modified configure work, I should restart the server. The process which I operate as below:

First, I use the command: "ant" to start the server, and 10 nodes connect to the server. At the same, I submit the tasks, the nodes work well. Then, I modify the configure of server.

Second, In order to let the server update, I use the keyboard "Ctrl + C" to terminate the server. The 10 nodes begin trying to explore the server. And then I repeat to input the "ant" to start the server.  In this time, the 10 nodes reconnect to the server. But when I submit the tasks, the nodes seems no work!

could you give me the way to solve this problem?

thanks!

Sincererly,
javacfish
« Last Edit: August 10, 2013, 05:16:32 AM by javacfish »
Logged

Colin

  • JPPF Padawan
  • *
  • Posts: 1
Re: It seems that nodes can not work if I restart the server?
« Reply #1 on: September 02, 2013, 11:47:42 PM »

I've run into this same problem.  Using JPPF 3.3.5 (though noticed the same behavior with 3.3.3).

If the driver is stopped and restarted but the nodes are not restarted, job submission from the client never gets executed.

Also, I have a ClientListener set on the JPPFClient, but neither connectionFailed() nor newConnection() is called when the driver is stopped or restarted.

Right AFTER I submit a job after restarting the driver I'll see multiples of this:
[client: driver1 - TasksServer] Attempting connection to the JPPF task server at x.x.x.x:11111

and then:

[client: driver1 - TasksServer] Reconnected to the JPPF task server


But the task is never executed on any of the nodes.

Thanks for any help!

-Colin


Logged

lolo

  • Administrator
  • JPPF Council Member
  • *****
  • Posts: 2268
    • JPPF Web site
Re: It seems that nodes can not work if I restart the server?
« Reply #2 on: September 03, 2013, 07:00:04 AM »

Hi Colin, javacfish

I'm having trouble reproducing this behavior. Is there any way you could send over your client, node and driver configuration files, so I can figure what I may be missing?
Also, could you tell us the minimum number of nodes you've seen this problem with?

Regarding the ClientListener notifications, I believe what you are observing is the expected behavior. Each connection from the client to the driver has a built-in recovery mechanism (the same mechanism applies to the nodes). This will cause the client to retry connecting for a specified time and at specified intervals. The time during which the client will retry connecting is specified with the configuration property "reconnect.max.time = time_in_seconds", the interval time is specified with "reconnect.interval = time_in_seconds". While "reconnect.max.time" has not been reached, the client considers that the connection has not yet failed. Only after that time is reached, the connectionFailed() notification will be sent. After that, a successful connection attempt will trigger a newConnection() notification.

Thus, if you want to be immediately notified of a disconnection or reconnection, you will need to use a ClientConnectionStatusListener, as illustrated here.

Thanks,
-Laurent
Logged

javacfish

  • Guest
Re: It seems that nodes can not work if I restart the server?
« Reply #3 on: September 05, 2013, 07:36:22 PM »

The situation really happen. The test was 10 nodes. I will tell you the specific err another day.

Now I restart the server, and the node was something wrong:

Code: [Select]
     [java] Exception in thread "main" java.lang.NoClassDefFoundError: org/jppf/management/spi/JPPFMBeanProviderManager
     [java] at org.jppf.server.node.JPPFNode.reset(JPPFNode.java:409)
     [java] at org.jppf.server.node.JPPFNode.run(JPPFNode.java:128)
     [java] at org.jppf.node.NodeRunner.main(NodeRunner.java:142)
     [java] Caused by: java.lang.ClassNotFoundException: could not find resource org.jppf.management.spi.JPPFMBeanProviderManager
     [java] at org.jppf.classloader.AbstractJPPFClassLoader.findClass(AbstractJPPFClassLoader.java:145)
     [java] at org.jppf.classloader.AbstractJPPFClassLoader.findClass(AbstractJPPFClassLoader.java:113)
     [java] at java.lang.ClassLoader.loadClass(ClassLoader.java:423)
     [java] at org.jppf.classloader.AbstractJPPFClassLoader.loadClass(AbstractJPPFClassLoader.java:432)
     [java] at java.lang.ClassLoader.loadClass(ClassLoader.java:356)
     [java] ... 3 more

The test version of JPPF was 3.1.
 
What is the problem?

Thanks!
javacfish
« Last Edit: September 06, 2013, 05:04:12 AM by javacfish »
Logged

lolo

  • Administrator
  • JPPF Council Member
  • *****
  • Posts: 2268
    • JPPF Web site
Re: It seems that nodes can not work if I restart the server?
« Reply #4 on: September 06, 2013, 08:26:25 AM »

Hello javacfish,

I have registered a bug for this issue: JPPF-185 NoClassDefFoundError in JPPFNode.reset()
JPPF v3.1 is not the most up to date version, is there any chance you could upgrade to the latest version v3.3.5?
In any case I will keep investigating and I will let you know of my findings ion this thread.

Sincerely,
-Laurent
Logged
Pages: [1]   Go Up
 
JPPF Powered by SMF 2.0 RC5 | SMF © 2006–2011, Simple Machines LLC Get JPPF at SourceForge.net. Fast, secure and Free Open Source software downloads