Galera ne fonctionne pas avec la réplication
J'utilise Mariahbb 10.2
J'ai node2 (2.2.2.2) en cours d'exécution comme esclave à une base de données externe. J'ai bootsTrapped Node2 comme premier membre d'un groupe Galera "my_cluster".
Node1 (1.1.1.1) Joignée de RSYNC SST mais après une heure, je vois que c'est derrière Node2.
J'ai essayé de courir des insertions manuellement et cela fonctionne dans les deux sens. Cependant, tout ce qui est ajouté par la réplication à Node2 dans complètement ignoré par Node1.
nœud1:
MariaDB [db1]> SHOW STATUS LIKE 'wsrep%';
+------------------------------+--------------------------------------+
| Variable_name | Value |
+------------------------------+--------------------------------------+
| wsrep_apply_oooe | 0.000000 |
| wsrep_apply_oool | 0.000000 |
| wsrep_apply_window | 0.000000 |
| wsrep_causal_reads | 0 |
| wsrep_cert_deps_distance | 0.000000 |
| wsrep_cert_index_size | 0 |
| wsrep_cert_interval | 0.000000 |
| wsrep_cluster_conf_id | 6 |
| wsrep_cluster_size | 2 |
| wsrep_cluster_state_uuid | 09e3b6c8-343c-11e8-87cf-07a9813fdf95 |
| wsrep_cluster_status | Primary |
| wsrep_commit_oooe | 0.000000 |
| wsrep_commit_oool | 0.000000 |
| wsrep_commit_window | 0.000000 |
| wsrep_connected | ON |
| wsrep_desync_count | 0 |
| wsrep_evs_delayed | |
| wsrep_evs_evict_list | |
| wsrep_evs_repl_latency | 0/0/0/0/0 |
| wsrep_evs_state | OPERATIONAL |
| wsrep_flow_control_paused | 0.000000 |
| wsrep_flow_control_paused_ns | 0 |
| wsrep_flow_control_recv | 0 |
| wsrep_flow_control_sent | 0 |
| wsrep_gcomm_uuid | 8854b393-3713-11e8-8cfd-f7a101a4c6bf |
| wsrep_incoming_addresses | 1.1.1.1:3306,2.2.2.2:3306 |
| wsrep_last_committed | 0 |
| wsrep_local_bf_aborts | 0 |
| wsrep_local_cached_downto | 18446744073709551615 |
| wsrep_local_cert_failures | 0 |
| wsrep_local_commits | 0 |
| wsrep_local_index | 0 |
| wsrep_local_recv_queue | 0 |
| wsrep_local_recv_queue_avg | 0.000000 |
| wsrep_local_recv_queue_max | 1 |
| wsrep_local_recv_queue_min | 0 |
| wsrep_local_replays | 0 |
| wsrep_local_send_queue | 0 |
| wsrep_local_send_queue_avg | 0.000000 |
| wsrep_local_send_queue_max | 1 |
| wsrep_local_send_queue_min | 0 |
| wsrep_local_state | 4 |
| wsrep_local_state_comment | Synced |
| wsrep_local_state_uuid | 09e3b6c8-343c-11e8-87cf-07a9813fdf95 |
| wsrep_protocol_version | 8 |
| wsrep_provider_name | Galera |
| wsrep_provider_vendor | Codership Oy <[email protected]> |
| wsrep_provider_version | 25.3.23(r3789) |
| wsrep_ready | ON |
| wsrep_received | 3 |
| wsrep_received_bytes | 219 |
| wsrep_repl_data_bytes | 0 |
| wsrep_repl_keys | 0 |
| wsrep_repl_keys_bytes | 0 |
| wsrep_repl_other_bytes | 0 |
| wsrep_replicated | 0 |
| wsrep_replicated_bytes | 0 |
| wsrep_thread_count | 2 |
+------------------------------+--------------------------------------+
nœud2:
MariaDB [db1]> SHOW STATUS LIKE 'wsrep%';
+------------------------------+--------------------------------------+
| Variable_name | Value |
+------------------------------+--------------------------------------+
| wsrep_apply_oooe | 0.000000 |
| wsrep_apply_oool | 0.000000 |
| wsrep_apply_window | 0.000000 |
| wsrep_causal_reads | 0 |
| wsrep_cert_deps_distance | 0.000000 |
| wsrep_cert_index_size | 0 |
| wsrep_cert_interval | 0.000000 |
| wsrep_cluster_conf_id | 6 |
| wsrep_cluster_size | 2 |
| wsrep_cluster_state_uuid | 09e3b6c8-343c-11e8-87cf-07a9813fdf95 |
| wsrep_cluster_status | Primary |
| wsrep_commit_oooe | 0.000000 |
| wsrep_commit_oool | 0.000000 |
| wsrep_commit_window | 0.000000 |
| wsrep_connected | ON |
| wsrep_desync_count | 0 |
| wsrep_evs_delayed | |
| wsrep_evs_evict_list | |
| wsrep_evs_repl_latency | 0/0/0/0/0 |
| wsrep_evs_state | OPERATIONAL |
| wsrep_flow_control_paused | 0.000000 |
| wsrep_flow_control_paused_ns | 0 |
| wsrep_flow_control_recv | 0 |
| wsrep_flow_control_sent | 0 |
| wsrep_gcomm_uuid | d1198d28-367a-11e8-a0ac-2382228e259f |
| wsrep_incoming_addresses | 1.1.1.1:3306,2.2.2.2:3306 |
| wsrep_last_committed | 0 |
| wsrep_local_bf_aborts | 0 |
| wsrep_local_cached_downto | 18446744073709551615 |
| wsrep_local_cert_failures | 0 |
| wsrep_local_commits | 0 |
| wsrep_local_index | 1 |
| wsrep_local_recv_queue | 0 |
| wsrep_local_recv_queue_avg | 0.100000 |
| wsrep_local_recv_queue_max | 2 |
| wsrep_local_recv_queue_min | 0 |
| wsrep_local_replays | 0 |
| wsrep_local_send_queue | 0 |
| wsrep_local_send_queue_avg | 0.000000 |
| wsrep_local_send_queue_max | 1 |
| wsrep_local_send_queue_min | 0 |
| wsrep_local_state | 4 |
| wsrep_local_state_comment | Synced |
| wsrep_local_state_uuid | 09e3b6c8-343c-11e8-87cf-07a9813fdf95 |
| wsrep_protocol_version | 8 |
| wsrep_provider_name | Galera |
| wsrep_provider_vendor | Codership Oy <[email protected]> |
| wsrep_provider_version | 25.3.23(r3789) |
| wsrep_ready | ON |
| wsrep_received | 10 |
| wsrep_received_bytes | 1081 |
| wsrep_repl_data_bytes | 0 |
| wsrep_repl_keys | 0 |
| wsrep_repl_keys_bytes | 0 |
| wsrep_repl_other_bytes | 0 |
| wsrep_replicated | 0 |
| wsrep_replicated_bytes | 0 |
| wsrep_thread_count | 2 |
+------------------------------+--------------------------------------+
nœud1 my.cnf:
log_bin = /var/mysql/log/mysql-bin.log
max_binlog_size = 100M
expire_logs_days=3
max_binlog_cache_size = 2G
binlog_cache_size = 32K
max_binlog_stmt_cache_size = 2G
binlog_stmt_cache_size = 32K
binlog_format=row
default-storage-engine=InnoDB
innodb_autoinc_lock_mode=2
query_cache_size=0
query_cache_type=0
innodb_flush_log_at_trx_commit=0
wsrep_on=ON
wsrep_slave_threads=1
wsrep_provider=/usr/lib/galera/libgalera_smm.so
wsrep_cluster_name="my_cluster"
wsrep_cluster_address="gcomm://2.2.2.2"
# also tried below
#wsrep_cluster_address="gcomm://2.2.2.2,1.1.1.1"
wsrep_sst_receive_address=1.1.1.1:4444
wsrep_provider_options='ist.recv_addr=1.1.1.1:4568;'
wsrep_sst_method=rsync
wsrep_sst_donor="node2,"
wsrep_node_address="1.1.1.1"
wsrep_node_name="node1"
nœud2 my.cnf:
log_bin = /var/mysql/log/mysql-bin.log
max_binlog_size = 100M
expire_logs_days=3
max_binlog_cache_size = 2G
binlog_cache_size = 32K
max_binlog_stmt_cache_size = 2G
binlog_stmt_cache_size = 32K
server-id = 10
relay-log = /var/mysql/log/mysql-relay-bin.log
replicate-ignore-db = mysql
binlog_format=row
default-storage-engine=InnoDB
innodb_autoinc_lock_mode=2
query_cache_size=0
query_cache_type=0
innodb_flush_log_at_trx_commit=0
wsrep_on=ON
wsrep_slave_threads=1
wsrep_provider=/usr/lib/galera/libgalera_smm.so
wsrep_cluster_name="my_cluster"
# for bootstrapping
wsrep_cluster_address="gcomm://"
#wsrep_cluster_address="gcomm://2.2.2.2,1.1.1.1"
wsrep_sst_receive_address=2.2.2.2:4444
wsrep_provider_options='ist.recv_addr=2.2.2.2:4568;'
wsrep_sst_method=rsync
wsrep_node_address="2.2.2.2"
wsrep_node_name="node2"
Votre node2 doit avoir log-slave-updates=1.
Voir aussi article de plusieurs naines sur le sujet qui indique:
log-esclaves-mises à jour = 1 est nécessaire pour les écritures de se propager de l'esclave aux autres nœuds Galera