I have an NDB cluster made of two management instances, two ndbd(data nodes) instances and two SQL nodes instances. The management nodes are sharing a VM with the data nodes and the SQL nodes hold their own VMs.
I am attempting to use dynamic floating IPs for the cluster rather than internal IPs. I attempted to configure a new NIC address for the various VMs I am running the services on which works as the nodes can now bind on those newly configured "local addresses", though the nodes (Data Nodes, SQL Nodes) themselves don't stay connected to the management nodes but rather continuously attempt to reconnected.
Could anyone please help me out?
Attaching the following logs:
INFO -- Node 3: Initial start, waiting for node 4 to connect, nodes [ all: 3 and 4 connected: 3 no-wait: ]
INFO -- Node 4: Initial start, waiting for node 3 to connect, nodes [ all: 3 and 4 connected: 4 no-wait: ]
Attaching the config.ini configurations:
[ndb_mgmd default]
DataDir=/var/lib/mysql-cluster/cluster
[ndb_mgmd]
HostName=[some1.dns.pointing.at.floating.ip]
NodeId=1
PortNumber=1186
PortNumberStats=2186
[ndb_mgmd]
HostName=[some2.dns.pointing.at.floating.ip]
NodeId=2
PortNumber=1186
PortNumberStats=2186
[ndbd default]
NoOfReplicas=2
DataMemory=8148M
MaxNoOfAttributes=40000
MaxNoOfOrderedIndexes=20000
DataDir=/var/lib/mysql
[ndbd]
HostName=[some3.dns.pointing.at.floating.ip]
NodeId=3
TcpBind_INADDR_ANY=1
[ndbd]
HostName=[some4.dns.pointing.at.floating.ip]
NodeId=4
TcpBind_INADDR_ANY=1
[mysqld]
HostName=[some5.dns.pointing.at.floating.ip]
AutoReconnect=true
NodeId=5
[mysqld]
HostName=[some6.dns.pointing.at.floating.ip]
AutoReconnect=true
NodeId=6
[tcp default]
TcpBind_INADDR_ANY=1
PreferIPVersion=4
Attempted to configure new NIC addresses for the nodes, the NIC configuration is:
BOOTPROTO=dhcp
DEVICE=eth0:1
MTU=1500
ONBOOT=yes
MN_CONTROLLED=no
IPADDR=[address of floating ip]
NETMASK=255.255.255.255
STARTMODE=auto
TYPE=Ethernet
USERCTL=no
Attempted to use wan configurations for the cluster.
