I see that ports from 47100 to 47109 enlisted in ip finder. By default the
addresses used by CommunicationSpi. Could you leave only one port range?
<property name="addresses">
<list>
<value>192.168.10.231:47500..47509</value>
</list>
</property>
> However, it still takes a very long time to connect.
It's related with windows specific. Long waiting when trying to connect to
the vacant port. For avoiding it need to decrease port range. Also if you
start node under powershell could you set java.encoding (-Dfile.encoding=UTF
-8
On Wed, Jan 4, 2017 at 4:01 PM, mark_balmer <
[email protected]> wrote:
> An update on this, it seems I can only get the exception if I have the
> second
> node in the cluster. If I just run the Java code against a single server
> node it doesnt throw the error. However, it still takes a very long time to
> connect.
>
> Here is the output from the topology change in the console, it seems a bit
> odd that its assigning 10 cpus for such a simple hello world operation!
>
> [12:56:41] Topology snapshot [ver=6, servers=1, clients=0, CPUs=2,
> heap=1.0GB]
> [12:58:34] Topology snapshot [ver=7, servers=1, clients=1, CPUs=10,
> heap=4.5GB]
> [12:58:35] Topology snapshot [ver=8, servers=1, clients=0, CPUs=2,
> heap=1.0GB]
>
>
>
> --
> View this message in context: http://apache-ignite-users.
> 70518.x6.nabble.com/Invalid-message-type-84-error-tp9869p9872.html
> Sent from the Apache Ignite Users mailing list archive at Nabble.com.
>
