[SOLVED] SNMPd stop binding on port after node fail

[Przemysław Knycz]

Hello.

I see that I've strange problem with SNMPd after my node with ID 1 stops working.

# pvecm status
Quorum information
------------------
Date: Fri Jun 14 14:49:46 2019
Quorum provider: corosync_votequorum
Nodes: 4
Node ID: 0x00000001
Ring ID: 4/21584
Quorate: Yes
Votequorum information
----------------------
Expected votes: 4
Highest expected: 4
Total votes: 4
Quorum: 3
Flags: Quorate
Membership information
----------------------
Nodeid Votes Name
0x00000004 1 192.168.7.48
0x00000002 1 192.168.7.50
0x00000001 1 192.168.7.52 (local)
0x00000003 1 192.168.7.53

When node 192.168.7.52 goes down it disables binding SNMPd on other machines of cluster.

On 7.53 node you can see that snmpd start working as usually but i see nothing more:

# netstat -anp | grep snmp
unix 3 [ ] STREAM CONNECTED 65015715 1339/snmpd

Stopping service takes more than normally (up to 1-2 minutes).

When 7.52 is up, all starting working without problems on 7.53:

# netstat -anp | grep snmp
udp 0 0 0.0.0.0:161 0.0.0.0:* 13028/snmpd
udp6 0 0 ::1:161 :::* 13028/snmpd
unix 2 [ ACC ] STREAM LISTENING 66815744 13028/snmpd /var/agentx/master
unix 2 [ ] DGRAM 66815741 13028/snmpd
unix 3 [ ] STREAM CONNECTED 66816807 13028/snmpd

and snmpd service can be stopped below 1 second.

Does anybody have this issue? I wonder if it isn't connected with that 7.52 is a first node of cluster.

Cheers.

 

[Przemysław Knycz]

Hi.

After research seems that problem was with not unmounted NFS share (not seen in Storage at GUI, but it was mounted on all nodes).

https://www.youtube.com/watch?v=J2-2K5ldhUE