|
The target machine has IPv6 installed, and if you look at the daemon's gui, it says its name is rosyth_0:0:0:0:0:1
[smartfrog@rosyth ~]$ /sbin/ifconfig eth0 Link encap:Ethernet HWaddr 00:0C:29:C9:14:95 inet addr:16.25.169.167 Bcast:16.25.171.255 Mask:255.255.252.0 inet6 addr: fe80::20c:29ff:fec9:1495/64 Scope:Link UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:76034 errors:5 dropped:5 overruns:0 frame:0 TX packets:5945 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:1000 RX bytes:45860297 (43.7 MiB) TX bytes:425988 (416.0 KiB) Interrupt:67 Base address:0x2000 eth1 Link encap:Ethernet HWaddr 00:0C:29:C9:14:9F UP BROADCAST MULTICAST MTU:1500 Metric:1 RX packets:0 errors:0 dropped:0 overruns:0 frame:0 TX packets:0 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:1000 RX bytes:0 (0.0 b) TX bytes:0 (0.0 b) Interrupt:67 Base address:0x2080 lo Link encap:Local Loopback inet addr:127.0.0.1 Mask:255.0.0.0 inet6 addr: ::1/128 Scope:Host UP LOOPBACK RUNNING MTU:16436 Metric:1 RX packets:2066 errors:0 dropped:0 overruns:0 frame:0 TX packets:2066 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:0 RX bytes:3382840 (3.2 MiB) TX bytes:3382840 (3.2 MiB) The local developer machine that is trying to deploy does have ipV6 enabled. -the obvious workaround is to disable IPv6. But it raises some questions
1. Why is RMI defaulting to IPv6 and ::1 as the address? 2. If it is, why isn't it listening on that port/ turn off ipv6 by editing /etc/modprobe.conf and adding two lines
alias net-pf-10 off alias ipv6 off Part of the problem is /etc/hosts, that binds localhost to a ::1 address. This must be deleted
Also, remove the "rosyth" entry there, as that was what was binding the local hostname to ::1. Hostname lookup should be via DNS, or, if you are in a VM-centric world, via explicit /etc/hosts entries everywhere. Not with machine names bound to localhost. This is wrong. the solution is to insert the following option into setSFEnvVariables.sh
#IPv4 only, please export SF_OPTS="-Djava.net.preferIPv4Stack=true" and in build.properties (for testing) daemon.jvmargs="-Djava.net.preferIPv4Stack=true" this switches SF to IPv4 only. I'm not going to stick this setting into bin/default.ini (yet), though it is tempting. Linking to the Hadoop problem. The fix is the same.
resolving as wontfix; there is a workaround.
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
Bug
Resolved
Major
SF is trying to deploy over IPv6, when it shouldn't
[deploy] - FAILED when trying DEPLOY of 'cluster', [/org/smartfrog/extras/hadoop/cluster/cluster.sf], host:16.25.169.167
[deploy] Result:
[deploy] * Exception: 'SmartFrogException:: java.rmi.ConnectIOException: Exception creating connection to: 0:0:0:0:0:0:0:1; nested exception is:
[deploy] java.net.SocketException: Protocol family unavailable
[deploy] SmartFrog 3.17.008 (2009-04-29 17:20:32 BST)'
[deploy] * StackTrace: '
[deploy] SmartFrogException:: java.rmi.ConnectIOException: Exception creating connection to: 0:0:0:0:0:0:0:1; nested exception is:
[deploy] java.net.SocketException: Protocol family unavailable, SmartFrog 3.17.008 (2009-04-29 17:20:32 BST)
[deploy] at org.smartfrog.sfcore.common.SmartFrogException.forward(SmartFrogException.java:92)
[deploy] at org.smartfrog.sfcore.common.ActionDeploy.Deploy(ActionDeploy.java:192)
[deploy] at org.smartfrog.sfcore.common.ActionDeploy.doDeploy(ActionDeploy.java:297)
[deploy] at org.smartfrog.sfcore.common.ActionDeploy.execute(ActionDeploy.java:272)
[deploy] at org.smartfrog.sfcore.common.ConfigurationAction.execute(ConfigurationAction.java:113)
[deploy] at org.smartfrog.sfcore.common.ConfigurationDescriptor.execute(ConfigurationDescriptor.java:1076)
[deploy] at org.smartfrog.SFSystem.runConfigurationDescriptor(SFSystem.java:367)
[deploy] at org.smartfrog.SFSystem.runConfigurationDescriptor(SFSystem.java:346)
[deploy] at org.smartfrog.SFSystem.runConfigurationDescriptors(SFSystem.java:303)
[deploy] at org.smartfrog.SFSystem.runSmartFrog(SFSystem.java:591)
[deploy] at org.smartfrog.SFSystem.execute(SFSystem.java:434)
[deploy] at org.smartfrog.SFSystem.main(SFSystem.java:395)
[deploy] Caused by: java.rmi.ConnectIOException: Exception creating connection to: 0:0:0:0:0:0:0:1; nested exception is:
[deploy] java.net.SocketException: Protocol family unavailable
[deploy] at sun.rmi.transport.tcp.TCPEndpoint.newSocket(TCPEndpoint.java:614)
[deploy] at sun.rmi.transport.tcp.TCPChannel.createConnection(TCPChannel.java:198)
[deploy] at sun.rmi.transport.tcp.TCPChannel.newConnection(TCPChannel.java:184)
[deploy] at sun.rmi.server.UnicastRef.invoke(UnicastRef.java:110)
[deploy] at org.smartfrog.sfcore.processcompound.ProcessCompoundImpl_Stub.sfDeployComponentDescription(Unknown Source)
[deploy] at org.smartfrog.sfcore.common.ActionDeploy.Deploy(ActionDeploy.java:155)
[deploy] ... 10 more
[deploy] Caused by: java.net.SocketException: Protocol family unavailable
[deploy] at java.net.PlainSocketImpl.socketConnect(Native Method)
[deploy] at java.net.PlainSocketImpl.doConnect(PlainSocketImpl.java:333)
[deploy] at java.net.PlainSocketImpl.connectToAddress(PlainSocketImpl.java:195)
[deploy] at java.net.PlainSocketImpl.connect(PlainSocketImpl.java:182)
[deploy] at java.net.SocksSocketImpl.connect(SocksSocketImpl.java:366)
[deploy] at java.net.Socket.connect(Socket.java:518)
[deploy] at java.net.Socket.connect(Socket.java:468)
[deploy] at java.net.Socket.<init>(Socket.java:365)
[deploy] at java.net.Socket.<init>(Socket.java:179)
[deploy] at sun.rmi.transport.proxy.RMIDirectSocketFactory.createSocket(RMIDirectSocketFactory.java:22)
[deploy] at sun.rmi.transport.proxy.RMIMasterSocketFactory.createSocket(RMIMasterSocketFactory.java:128)
[deploy] at sun.rmi.transport.tcp.TCPEndpoint.newSocket(TCPEndpoint.java:595)
[deploy] ... 15 more
[deploy] ALL: SmartFrogException: java.rmi.ConnectIOException: Exception creating connection to: 0:0:0:0:0:0:0:1; nested exception is: