I've been trying to set an NDB Cluster with docker-compose so I've tried to translate the instructions found in docker hub of MySQL-cluster (https://hub.docker.com/r/mysql/mysql-cluster/) and I'm having problems that I think is related to the amount of memory allocated to the MySQL nodes because they're unstable and restarting every couple of seconds.
I haven't changed the amount of memory in the configuration file "mysql-cluster.cnf" which by the way throws a parsing error when there's an empty line in the file even though the original file in the mysql cluster repo has empty lines.
This is what ndb_mgm shows:
$ docker exec -it mysql_cluster_mysql-manager-1_1 bash
bash-4.2# ndb_mgm
-- NDB Cluster -- Management Client --
ndb_mgm> show
Connected to Management Server at: mysql-manager-1:1186
Cluster Configuration
---------------------
[ndbd(NDB)] 2 node(s)
id=11 @172.19.0.6 (mysql-8.0.22 ndb-8.0.22, starting, Nodegroup: 0)
id=12 (not connected, accepting connect from data-node-2)
[ndb_mgmd(MGM)] 2 node(s)
id=1 @172.19.0.2 (mysql-8.0.22 ndb-8.0.22)
id=2 @172.19.0.3 (mysql-8.0.22 ndb-8.0.22)
[mysqld(API)] 2 node(s)
id=21 (not connected, accepting connect from mysql-server-1)
id=22 (not connected, accepting connect from mysql-server-2)
ndb_mgm> Node 11: Forced node shutdown completed. Occurred during startphase 0. Initiated by signal 9.
This is part of the output log from the data node:
2020-12-18 23:36:46 [ndbd] INFO -- Clearing filesystem in initial start
2020-12-18 23:36:47 [ndbd] INFO -- Start phase 0 completed
2020-12-18 23:36:47 [ndbd] INFO -- Phase 0 has made some file system initialisations
2020-12-18 23:36:47 [ndbd] INFO -- Watchdog KillSwitch off.
2020-12-18 23:36:47 [ndbd] INFO -- Starting QMGR phase 1
2020-12-18 23:36:47 [ndbd] INFO -- DIH reported initial start, now starting the Node Inclusion Protocol
2020-12-18 23:37:05 [ndbd] INFO -- Child process terminated by signal 9
2020-12-18 23:37:05 [ndbd] ALERT -- Node 11: Forced node shutdown completed. Occurred during startphase 0. Initiated by signal 9
I still haven't got to the analysis of why the mysqld API is failing as well.
I've posted the compose and related files in this repository
https://github.com/rsegecin/mysql-cluster.
Any help is appreciated.
Edited 1 time(s). Last edit at 01/02/2021 09:22AM by Rinaldi Segecin.