galera cluster does not start

1,879 views
Skip to first unread message

Apostolis Type R

unread,
Mar 6, 2019, 7:31:05 AM3/6/19
to codership
hello all,

i have installed 3 nodes in order to create a galera cluster.
in the 1st node "master" the gcomm is set to null and i can start mariadb with galera_new_cluster

in the other two nodes i cannot start mariadb with multiple errors showing 
"gcomm://x.x.x.x,c.c.c.c,v.v.v.v."

error log

systemctl status mariadb.service -l
● mariadb.service - MariaDB database server
   Loaded: loaded (/usr/lib/systemd/system/mariadb.service; enabled; vendor preset: disabled)
  Drop-In: /etc/systemd/system/mariadb.service.d
           └─migrated-from-my.cnf-settings.conf
   Active: activating (auto-restart) (Result: signal) since Wed 2019-03-06 13:50:12 EET; 720ms ago
  Process: 20749 ExecStartPost=/bin/sh -c systemctl unset-environment _WSREP_START_POSITION (code=exited, status=0/SUCCESS)
  Process: 54893 ExecStart=/usr/sbin/mysqld $MYSQLD_OPTS $_WSREP_NEW_CLUSTER $_WSREP_START_POSITION (code=killed, signal=ABRT)
  Process: 54813 ExecStartPre=/bin/sh -c [ ! -e /usr/bin/galera_recovery ] && VAR= ||   VAR=`/usr/bin/galera_recovery`; [ $? -eq 0 ]   && systemctl set-environment _WSREP_START_POSITION=$VAR || exit 1 (code=exited, status=0/SUCCESS)
  Process: 54811 ExecStartPre=/bin/sh -c systemctl unset-environment _WSREP_START_POSITION (code=exited, status=0/SUCCESS)
 Main PID: 54893 (code=killed, signal=ABRT)
   CGroup: /system.slice/mariadb.service
           ├─54902 /bin/sh -ue /usr//bin/wsrep_sst_rsync --role joiner --address    10.1.0.172" --datadir /var/lib/mysql/ --parent 54893
           ├─54976 rsync --daemon --no-detach --port 4444 --config /var/lib/mysql//rsync_sst.conf
           └─55027 sleep 0.5

Mar 06 13:50:12 tms-galeracl2 systemd[1]: mariadb.service: main process exited, code=killed, status=6/ABRT
Mar 06 13:50:12 tms-galeracl2 systemd[1]: Failed to start MariaDB database server.
Mar 06 13:50:12 tms-galeracl2 systemd[1]: Unit mariadb.service entered failed state.
Mar 06 13:50:12 tms-galeracl2 systemd[1]: mariadb.service failed.

Xiao Guo-An

unread,
Mar 13, 2019, 2:52:06 AM3/13/19
to codership
You need to add all IP addresses of the nodes to the gcomm on the all 3 nodes.

Then on the first node, stop MariaDB and start the cluster with: galera_new_cluster.

Finally, restart MariaDB server on the other two nodes.

Saeid Bostandoust

unread,
Apr 4, 2019, 3:30:18 PM4/4/19
to codership
If you want to create new Galera cluster lets do these steps:

1- on SERVER1 stop mariadb and set galera configuration with "gcomm://"
2- run galera_new_cluster, mariadb starts automatically with galera cluster initiated.
3- on SERVER2 stop mariadb and set galera configuration with "gcomm://SERVER1"
4- start mariadb on SERVER2, it join to cluster automatically.
5- do two previous steps on SERVER3.
6- update gcomm on all nodes with value "gcomm://SERVER1,SERVER2,SERVER3"

Done.

lucky

unread,
Apr 24, 2019, 9:19:52 AM4/24/19
to codership
You are not clear
>>1- on SERVER1 stop mariadb and set galera configuration with "gcomm://"
you mean to set this in config file and start galera and mariadb from node1,

LinuxBabe

unread,
Apr 24, 2019, 9:25:32 AM4/24/19
to codership
Actually, you can put the IP addresses of all nodes in gcomm:// before you start Galera cluster on node 1.
Reply all
Reply to author
Forward
0 new messages