mysqld 不会连接到 MySQL 集群或数据节点

Question

I've created a v8.0MySQL cluster running on 4 VMs. VMs 1 & 2 have the MySQL and Cluster Manager daemon installed. VMs 3 & 4 have the Data Node daemon installed.

Configuration of VMs:

 ╓──────────────────────╥─────────────────────╥────────────────────╥────────────────────╖
 ║  [VM1]               ║ [VM2]               ║ [VM3]              ║  [VM4]             ║
 ║  10.103.0.11         ║ 10.103.0.12         ║ 10.103.0.23        ║  10.103.0.24       ║
 ║  mysql-clustermgr-01 ║ mysql-clustermgr-02 ║ mysql-datanode-01  ║  mysql-datanode-02 ║
 ║  ------------------- ║ ------------------- ║ ------------------ ║  ----------------- ║
 ║  mysqld  (NodeId=5)  ║ mysqld  (NodeId=6)  ║ ndbd (NodeId=3)    ║  ndbd (NodeId=4)   ║
 ║  ndb_mgmd(NodeId=1)  ║ ndb_mgmd(NodeId=2)  ║                    ║                    ║
 ╙──────────────────────╨─────────────────────╨────────────────────╨────────────────────╜

I have ufw disabled on all nodes to alleviate any FW issues. I can ping between all the servers in question without issue. All servers belong to the same su.net. I have set the permissions for /etc/mysql from root:root to mysql:mysql . I have tried several how-to's on the Inte.net that diagnose around mysqld not connecting. Nothing seems to work over the last couple days. I have tried several configurations and this is as far as I can get with output from ndb_mgm :

emerg@mysql-clustermgr-01:~$ ndb_mgm -e show
Connected to Management Server at: mysql-clustermgr-01:1186
Cluster Configuration
---------------------
[ndbd(NDB)]     2 node(s)
id=3    @10.103.0.23  (mysql-8.0.28 ndb-8.0.28, starting, Nodegroup: 0)
id=4    @10.103.0.24  (mysql-8.0.28 ndb-8.0.28, starting, Nodegroup: 0)

[ndb_mgmd(MGM)] 2 node(s)
id=1    @10.103.0.11  (mysql-8.0.28 ndb-8.0.28)
id=2    @10.103.0.12  (mysql-8.0.28 ndb-8.0.28)

[mysqld(API)]   2 node(s)
id=5 (not connected, accepting connect from mysql-clustermgr-01)
id=6 (not connected, accepting connect from mysql-clustermgr-02)

As you can see, the two cluster daemons can see each other just fine, however mysqld and ndbd on the data nodes refuse to connect/start respectively.

Here's my configuration files:

[VM1] /etc/mysql/my.cnf

!includedir /etc/mysql/conf.d/
!includedir /etc/mysql/mysql.conf.d/

[mysqld]
ndbcluster

[mysql_cluster]
ndb-connectstring=NodeId=5,mysql-clustermgr-01,mysql-clustermgr-02

[VM2] /etc/mysql/my.cnf

!includedir /etc/mysql/conf.d/
!includedir /etc/mysql/mysql.conf.d/

[mysqld]
ndbcluster

[mysql_cluster]
ndb-connectstring=NodeId=6,mysql-clustermgr-01,mysql-clustermgr-02

[VM1 & VM2] /var/lib/mysql-cluster/config.ini

[ndbd default]
NoOfReplicas=2
DataMemory=80M
DataDir=/var/lib/mysql-cluster

[ndb_mgmd]
HostName=mysql-clustermgr-01
DataDir=/var/lib/mysql-cluster
NodeId=1

[ndb_mgmd]
HostName=mysql-clustermgr-02
DataDir=/var/lib/mysql-cluster
NodeId=2

[ndbd]
HostName=mysql-datanode-01
NodeId=3
datadir=/usr/local/mysql/data
LockExecuteThreadToCPU=1
LockMaintThreadsToCPU=0

[ndbd]
HostName=mysql-datanode-02
NodeId=4
datadir=/usr/local/mysql/data

[mysqld]
HostName=mysql-clustermgr-01
NodeId=5

[mysqld]
HostName=mysql-clustermgr-02
NodeId=6

[VM3] /etc/my.cnf

[mysqld]
ndbcluster

[mysql_cluster]
ndb-connectstring=NodeId=3,mysql-clustermgr-01,mysql-clustermgr-02

[VM4] /etc/my.cnf

[mysqld]
ndbcluster

[mysql_cluster]
ndb-connectstring=NodeId=4,mysql-clustermgr-01,mysql-clustermgr-02

Services start fine, does anything within the configuration or architecture jump out at anyone?

Answer 1

Figured it out, you can't use hostnames, you have to use IPs in all config files. Man that set me back two days.