Version: Eucalyptus 1.6.1.
Installation: Compiled from source code.
H/W: CLC (10.146.172.96, 2Gb mem), separate CC (10.146.172.97, 2Gb mem), 4 x Nodes running on Dell 1950's (16Gb mem)
O/S : Oracle Enterprise Linux 5.3 x64. (plus latest release of Java JDK)
Hypervisor type: Xen 3.4
Topology: Separate CLC and 1 CC. 4 Nodes.
Networking mode: STATIC.
Looking at the CLC log files I am seeing an issue with the heartbeat from CLC to CC. It looks as if the CLC port (8773) is being used as the dst port when communicating with the CC (10.146.172.97).
Am I correct in assuming the port should be 8774 as specified in the eucalyptus.conf file.
If so, can you recommend a solution?
(I have refrained from digging around src code as I wanted to confirm that 8773 is never used as dst port when communicating with CC)
** cloud-output.log **
05:58:14 INFO ClusterUtil | ---------------------------------------------------------------
05:58:14 INFO ClusterUtil | -> [ EASD-CC ] Cluster certificate valid=true
05:58:14 INFO ClusterUtil | -> [ EASD-CC ] Node certificate valid=true
05:58:14 INFO ClusterUtil | ---------------------------------------------------------------
05:58:15 WARN NetworkStateHandler | Response from cluster [EASD-CC]: NOTSUPPORTED
05:58:20 WARN HeartbeatClient | Failed to connect to heartbeat service at /10.146.172.97:8773: Connection refused
05:58:20 ERROR RemoteBootstrapperClient | java.net.ConnectException: Connection refused
CC eucalyptus.conf contains
VNET_LOCALIP="10.146.172.97"
VNET_CLOUDIP="10.146.172.96"
** cloud-debug.log **
05:58:20 WARN [HeartbeatClient:New I/O client boss #1] Failed to connect to heartbeat service at /10.146.172.97:8773: Connection refused
05:58:20 ERROR [RemoteBootstrapperClient:New I/O client boss #1] java.net.ConnectException: Connection refused
java.net.ConnectException: Connection refused
at sun.nio.ch.SocketChannelImpl.checkConnect(Native Method)
at sun.nio.ch.SocketChannelImpl.finishConnect(SocketChannelImpl.java:574)
at org.jboss.netty.channel.socket.nio.NioClientSocketPipelineSink$Boss.connect(NioClientSocketPipelineSink.java:351)
at org.jboss.netty.channel.socket.nio.NioClientSocketPipelineSink$Boss.processSelectedKeys(NioClientSocketPipelineSink.java:322)
at org.jboss.netty.channel.socket.nio.NioClientSocketPipelineSink$Boss.run(NioClientSocketPipelineSink.java:247)
at org.jboss.netty.util.internal.IoWorkerRunnable.run(IoWorkerRunnable.java:53)
at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908)
at java.lang.Thread.run(Thread.java:619)
Any guidance is much appreciated
Mike.
Hello Mike,
can you elaborate a bit more on your installation procedure and where you put the services? IN particular: where is walrus and the SC? How did you register the services and if you did see any error during the registration process? Can you also report the output of euca_conf --list-walruses --list-scs --list-clusters?
cheers
graziano
Graziano
Thanks for replying!
Walrus exists on the same server as the CLC with the SC sharing the same server as the CC.
The output of euca_conf verifies this.
euca_conf --list-walruses --list-scs --list-clusters
registered walruses:
**** UPDATE ****
Issue solved.
Walrus 10.146.172.96
registered storage controllers:
EASD-CC 10.146.172.97
registered clusters:
EASD-CC 10.146.172.97
In terms of service registration, Walrus, CC and the SC were registered via the CLC UI and euca_conf --enable cloud --enable walrus run on the CLC
As for the installation, no, I can't recall any relevant errors.
Do you advising a recompilation?
Mike.
My issue was related to running the SC service on a different tier than the Cloud/Walrus tier. I run the CC and SC on the same tier.
My heartbeat error was due to the fact that although I had 'enabled' the sc service on the CC/SC tier via euca_conf (but ensured the cloud and walrus services were disabled), I had not started the service. Thus the CLC could not hand shake with the SC
In other words on the CC/SC tier, not only did i have to start /etc/init.d/eucalyptus-cc but also /etc/init.d/eucalyptus-cloud
Make sense?
Not really. Exactly what are you referring to in reference to "tier"s?
I have /etc/init.d/eucalyptus-cloud started on my SC/CC node but I still don't have anything listening on port 8773 and am getting continual
20:36:40 WARN [HeartbeatClient:New I/O client boss #1] Failed to connect to heartbeat service at /10.8.2.3:8773: Connection refused
20:36:40 ERROR [RemoteBootstrapperClient:New I/O client boss #1] java.net.ConnectException: Connection refused
java.net.ConnectException: Connection refused
at sun.nio.ch.SocketChannelImpl.checkConnect(Native Method)
at sun.nio.ch.SocketChannelImpl.finishConnect(SocketChannelImpl.java:592)
at org.jboss.netty.channel.socket.nio.NioClientSocketPipelineSink$Boss.connect(NioClientSocketPipelineSink.java:351)
at org.jboss.netty.channel.socket.nio.NioClientSocketPipelineSink$Boss.processSelectedKeys(NioClientSocketPipelineSink.java:322)
at org.jboss.netty.channel.socket.nio.NioClientSocketPipelineSink$Boss.run(NioClientSocketPipelineSink.java:247)
at org.jboss.netty.util.internal.IoWorkerRunnable.run(IoWorkerRunnable.java:53)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1110)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:603)
at java.lang.Thread.run(Thread.java:636)