6

We are implementing a 3-node openstack cloud using glusterfs for storage solution. 3 nodes : controller compute and network are peers in gluster. We need to add another compute node as a peer in gluster but it is showing the following error :

[root@newcompute2 ~]# gluster peer probe 192.168.10.3
peer probe: failed: Probe returned with Transport endpoint is not connected

where 192.168.10.3 is the IP of controller node. Logs are also showing the same error. Please suggest what may be the reason for this and the required solution.

cloud learner
  • 61
  • 1
  • 1
  • 4

9 Answers9

5

This indicate that their is no port open for proper communication Node1 and Node2 host.

For that you need to fire these commands on servers

1) firewall-cmd --zone=public --add-port=24007-24008/tcp --permanent

2) firewall-cmd --zone=public --add-port=24009/tcp --permanent

3) firewall-cmd --zone=public --add-service=nfs --add-service=samba --add-service=samba-client --permanent

4) firewall-cmd --zone=public --add-port=111/tcp --add-port=139/tcp --add-port=445/tcp --add-port=965/tcp --add-port=2049/tcp --add-port=38465-38469/tcp --add-port=631/tcp --add-port=111/udp --add-port=963/udp --add-port=49152-49251/tcp --permanent

5) firewall-cmd --reload

Then try again command => gluster peer probe x.y.z.n (IP)
peer probe: success.

Please check and revert if any other query.

maheshp1987
  • 146
  • 1
  • 3
3

You must be sure that glusterd service has been started on the probed node. you can use the following command to start the glusterd service:

sudo service glusterd start

or you can restart the service with the following command:

sudo service glusterd restart
Amirhossein72
  • 491
  • 6
  • 7
2

I have had this error come up when I have GlusterFS configured to use SSL and there is a problem with loading the certificates as well (expired cert, etc.) You can check this by removing the file: /var/lib/glusterd/secure-access. If this file does not exist, then you probably do not have SSL configured and need to look for the issue elsewhere.

Chris Mooney
  • 21
  • 3
1

I know its very late but I too got stuck with this problem and somehow solved it. So for others facing this issue, one reason could be that once the cluster is formed new nodes can be added only from one of the trusted nodes.

The quick start doc very clearly says

Note: Once this pool has been established, only trusted members may probe new servers into the pool. A new server cannot probe the pool, it must be probed from the pool.

for example:

node-1 and node-2: part of the trusted cluster group

node-3: needs to be added to this cluster

then from either node-1 or node-2 write

sudo gluster peer probe <node-3>
coda
  • 2,188
  • 2
  • 22
  • 26
  • you are right. Seems strange, but probings from two different sides are not the same. Thank you for the hint. – rth Jan 17 '18 at 20:02
  • @rth why does it seem strange? if this was not the case, any server could inject itself into an existing cluster group. – pzkpfw Jul 21 '22 at 10:22
1

This is because of firewall issue.

peer probe: failed: Probe returned with Transport endpoint is not connected

You need to run the following command in all of your peers.

systemctl stop firewalld
Ayush Ujjwal
  • 11
  • 1
1

From link, https://docs.gluster.org/en/latest/Quick-Start-Guide/Quickstart/ below step should solve problem.

Step 4 - Configure the firewall The gluster processes on the nodes need to be able to communicate with each other. To simplify this setup, configure the firewall on each node to accept all traffic from the other node.

iptables -I INPUT -p all -s -j ACCEPT

chetan more
  • 11
  • 3
0

iptables -I INPUT -p all -s 192.168.10.3 -j ACCEPT (accpet rom one and the second it works on my machines :)

it is on instruction -> http://gluster.readthedocs.org/en/latest/Install-Guide/Configure/

user3619277
  • 21
0

Running below commands from worker nodes solved the issue

firewall-cmd --zone=public --add-port=24007-24008/tcp --permanent
firewall-cmd --reload
GLM
  • 59
  • 2
-4

Turn off firewall on both the nodes.

systemctl status firewalld.service

shekhar berry
  • 9
  • 1