GridGain node cannot be in one cluster with Ignite node

Question

I was running local ignite(8.7.21) server in local below is the confirmation log that ignite server is up and running. however when I'm trying to connect from Java client facing below issue any idea ?

local server log:

Topology snapshot [ver=11, locNode=34f4abec, servers=1, clients=0, state=ACTIVE, CPUs=2, offheap=2.0GB, heap=6.0GB] ^-- Baseline [id=0, size=1, online=1, offline=0]

Issue:

Caused by: class org.gridgain.grid.internal.processors.nodevalidation.NodeValidationException: GridGain node cannot be in one cluster with Ignite node [locNodeAddrs=[machine name/0:0:0:0:0:0:0:1, /10.136.68.128, /127.0.0.1], rmtNodeAddrs=[ machine name/0:0:0:0:0:0:0:1, /10.136.68.128, /127.0.0.1]]

defualt-cache.xml from Java client

   <bean id="igniteBean" class="org.apache.ignite.configuration.IgniteConfiguration" abstract="true">
        <property name="discoverySpi">
            <bean class="org.apache.ignite.spi.discovery.tcp.TcpDiscoverySpi">
                <property name="localPort" value="48500"/>
                <property name="localPortRange" value="5"/>
            </bean>
        </property>
    </bean>

default-xml which I used to start ignite server

<beans xmlns="http://www.springframework.org/schema/beans"
       xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
       xsi:schemaLocation="
       http://www.springframework.org/schema/beans
       http://www.springframework.org/schema/beans/spring-beans.xsd">

    <bean id="private.ignite.cfg" class="org.apache.ignite.configuration.IgniteConfiguration" abstract="true">
        <property name="discoverySpi">
            <bean class="org.apache.ignite.spi.discovery.tcp.TcpDiscoverySpi">
                <property name="localPort" value="48500"/>
                <property name="localPortRange" value="5"/>
            </bean>
        </property>
    </bean>

score 0 · Answer 1 · answered Dec 30 '21 at 00:52

Based on the IPs being the same I would surmise that you are running a GridGain java instance and an Ignite java instance on the same machine.
Use jps -v/jinfo to find all the java process running on the machine and shutdown the GridGain process.

IPs are the same:

 [locNodeAddrs=[machine name/0:0:0:0:0:0:0:1, /10.136.68.128, /127.0.0.1], rmtNodeAddrs=[ machine name/0:0:0:0:0:0:0:1, /10.136.68.128, /127.0.0.1]]

You can tell that whether the process is a GridGain/Ignite one using the classpath/startup location/logs

Thank you for responding, I tried stopping all Java process still its the same issue. — coder999, Dec 30 '21 at 15:26

score 0 · Answer 2 · answered Jan 02 '22 at 21:32

Caused by: class org.gridgain.grid.internal.processors.nodevalidation.NodeValidationException: GridGain node cannot be in one cluster with Ignite node [locNodeAddrs=[machine name/0:0:0:0:0:0:0:1, /10.136.68.128, /127.0.0.1], rmtNodeAddrs=[ machine name/0:0:0:0:0:0:0:1, /10.136.68.128, /127.0.0.1]]

2 possible causes:

you are running Apache Ignite and GridGain edition at the same time
you are running different GridGain editions at the same time, like Community and Ultimate one.

Note that your grid version is 8.7.21 which means, it's not Ignite, but GridGain edition. Check that you are not running another grid in Kubernetes/Docker environments on your machine. This is something that happens to me, though.

Alternatively, configure static IpFinder explicitely:

<bean class="org.apache.ignite.spi.discovery.tcp.TcpDiscoverySpi">
    <property name="ipFinder">
          <bean class="org.apache.ignite.spi.discovery.tcp.ipfinder.vm.TcpDiscoveryVmIpFinder">
              <property name="addresses">
                 <list>
                    <value>localhost:48500..48509</value>
                 </list>
              </property>

             .... <your values>

             
           </bean>
    </property></bean>

Thank you for you're response. I guess you're right I'm running Ignite (from application as client connecting to server) and Gridgain(8.7.21 community version running locally) at same time. — coder999, Jan 04 '22 at 19:17
Yes, unless it's not a thin client, this mixed configuration is prohibited. Use single edition instead: either GG or Ignite — Alexandr Shapkin, Jan 05 '22 at 11:26
If gridgain itself says the same mistake what I'm doing it would be nice. — coder999, Jan 05 '22 at 15:29

score 0 · Answer 3 · edited Feb 06 '23 at 19:45

We got this resolved by tidying up our dependencies, as we are working the the GridGain Community Edition 8.8.22 and really only need the core dependencies we changed ours to ~

def igniteVersion = '8.8.22'

// Ignite
compile group: 'org.gridgain', name: 'ignite-core', version: igniteVersion
compile group: 'org.gridgain', name: 'ignite-spring', version: igniteVersion
compile group: 'org.gridgain', name: 'ignite-indexing', version: igniteVersion
compile group: 'org.gridgain', name: 'ignite-h2', version: igniteVersion
compile group: 'org.gridgain', name: 'ignite-kubernetes', version: igniteVersion

We previously had the gridgain-ultimate dependency included as well, it may have been causing the conflict and really shouldn't have been in there as we are using the CE.

score -1 · Answer 4 · edited Feb 07 '23 at 15:50

I struggled with this for awhile until I found the following: Initially I was successfully executing Java compiled code against a 2 node cluster running on my laptop using gridgain-community-8.8.21.

I then downloaded gridgain-ultimate-8.8.24.
I then updated the gridgain version in my Maven build file to 8.8.24 and received the above exception.

Looking at the GridGain website I found that I had to add additional Maven dependencies for the ultimate edition.

I did that and still experienced the above exception!

After about an hour I shutdown everything and rebooted my laptop.
I then started my local 2 node cluster and I was able to successfully execute the newly compiled Java code against the gridgain-ultimate-8.8.24 code based.

GridGain node cannot be in one cluster with Ignite node

4 Answers4