Support Forums : DataStax

DataStax Support Forums » DataStax Community

The ./nodetool -h localhost ring doesnt list all the node in a 4 node cluster

(12 posts) (2 voices)

Started 8 months ago by abrahamjohnson
Latest reply from abrahamjohnson
This topic is resolved

Tags:

No tags yet.

abrahamjohnson
Member

I am using cassasndra 1.1.5 and opscenter2.1.2.

I have followed all the steps mentioned in the link :http://www.datastax.com/resources/articles/multi-node-cassandra-cluster-on-linux

But after starting cassasndra in all nodes,the nodes are not listed when using the command
./nodetool -h localhost ring
On all individual nodes only the current one is listed.
For eg: In the seed node with Ip 192.168.1.2

[root@masternode bin]# ./nodetool -h 192.168.1.2 -p 7199 ring
Address DC Rack Status State Load Effective-Ownership Token
192.168.1.2 datacenter1 rack1 Up Normal 54.25 KB 100.00% 0

In the other nodes with 192.168.1.3

[root@datanode1 bin]# ./nodetool -h 192.168.1.3 -p 7199 ring
Note: Ownership information does not include topology, please specify a keyspace.
Address DC Rack Status State Load Owns Token
192.168.1.3 datacenter1 rack1 Up Normal 31.26 KB 100.00% 42535295865117307932921825928971026432

Posted 8 months ago #
nickmbailey
Key Master

The first thing to check is to make sure your nodes can connect on the ports cassandra uses. You should check your firewall/network settings.

http://www.datastax.com/docs/1.1/references/firewall_ref

Posted 8 months ago #
abrahamjohnson
Member

Hi nick,
thanx for the reply..
bt i tried enabling all the ports mentioned in the list using :
iptables -A INPUT -p tcp --dport <port number> -j ACCEPT in all nodes.

I tried to ping all the system and am getting the reply too.

even if i use the ./nodetool -h local host join command
it gives the message 'Already Joined the ring'

Still the all nodes are not listed yet...

Posted 8 months ago #
nickmbailey
Key Master

How are your seed nodes configured in cassandra.yaml?

Posted 8 months ago #
abrahamjohnson
Member

I am having four nodes and these are main properties in cassandra.yaml in each node :

Node1: ip add/name (192.168.1.2/masternode)
listen_address:192.168.1.2
rpc_address:192.168.1.2
seeds:"192.168.1.2"

Node2:ip add/name (192.168.1.3/datanode1)
listen_address:192.168.1.3
rpc_address:192.168.1.3
seeds:"192.168.1.2"

Node3:ip add/name (192.168.1.4/datanode2)
listen_address:192.168.1.4
rpc_address:192.168.1.4
seeds:"192.168.1.2"

Node4:ip add/name (192.168.1.5/datanode3)
listen_address:192.168.1.5
rpc_address:192.168.1.5
seeds:"192.168.1.2"

Posted 7 months ago #
nickmbailey
Key Master

Those settings appear fine. The only other advice I can give is to check the log for errors, and possibly verify connectivity between the nodes.

Can you run:

telnet <ip> 7000

From each node to every other node?

Posted 7 months ago #
abrahamjohnson
Member

i tried the command in all node :i get the following from all

[root@masternode bin]# telnet 192.168.1.3 7000
Trying 192.168.1.3...
telnet: connect to address 192.168.1.3: No route to host
[root@masternode bin]# telnet 192.168.1.4 7000
Trying 192.168.1.4...
telnet: connect to address 192.168.1.4: No route to host
[root@masternode bin]# telnet 192.168.1.5 7000
Trying 192.168.1.5...
telnet: connect to address 192.168.1.5: No route to host

also when i searched for the process using port 7000
using the command : netstat -tunlap | grep 7000
i got following in nodes:

Node1:
[root@masternode bin]# netstat -tunlap | grep 7000
tcp 0 0 192.168.1.2:7000 0.0.0.0:* LISTEN 25245/java

Node 2:
[root@datanode1 bin]# netstat -tunlap | grep 7000
tcp 0 0 192.168.1.3:7000 0.0.0.0:* LISTEN 23825/java
tcp 0 1 192.168.1.3:53055 192.168.1.2:7000 SYN_SENT 23825/java

Does this signify anything?

Also there is no error logged in the logs... :(

Posted 7 months ago #
nickmbailey
Key Master

The netstat is showing the cassandra process listening on port 7000. The fact that telnet is failing means that even though cassandra is listening, something is blocking access to that port.

Most likely, you have a firewall/port blocking enabled either on your individual servers or on your networking hardware.

Posted 7 months ago #
abrahamjohnson
Member

i have tried evrything ,but still no progress :-( .The nodes are not listed yet.

There is no firewall for the system.They are just 4 sytems in an independent network.

Posted 7 months ago #
nickmbailey
Key Master

Perhaps your router/networking software is blocking certain ports?

Posted 7 months ago #
abrahamjohnson
Member

I stopped iptable service

command-- iptables servide stop

Now the node are listening to each other.

Thanks nick for yor support

Cheeerssss!!!!!!!!!!!! .. :)

Posted 7 months ago #
abrahamjohnson
Member

as small correction

The command is : service iptables stop

Posted 7 months ago #

RSS feed for this topic

Reply

You must log in to post.