maria-discuss team mailing list archive

[jocelyn fournier <jocelyn.fournier@xxxxxxxxx>]

Hi,


After upgrading from TokuDB Enterprise with MariaDB 5.5 to MariaDB 
10.1.14, I tried to enable the parallel replication 
(parallel_mode=optimistic, slave_parallel_threads=4) on a GTID enabled 
RFR slave, but it leads to random failures.

(both master & slave are running 10.1.14)

E.g :

Could not execute Update_rows_v1 event on table sc_2.sc_param_index; 
Can't find record in 'sc_param_index', Error_code: 1032; handler error 
HA_ERR_KEY_NOT_FOUND; the event's master log mysql-bin.008418, 
end_log_pos 217608066

If I issue a start slave (with sql_slave_skip_counter=0) :

Could not execute Update_rows_v1 event on table sc_2.sc_log_cron; Can't 
find record in 'sc_log_cron', Error_code: 1032; handler error 
HA_ERR_KEY_NOT_FOUND; the event's master log mysql-bin.008418, 
end_log_pos 218275984

If I put back slave_parallel_threads to 0, the slave keeps replicating 
happily without any issue.


Is this a known limitation with TokuDB ?


Tokudb configuration :

tokudb_rpl_lookup_rows=0

tokudb_rpl_unique_checks=0

tokudb_pk_insert_mode=2

tokudb_directio = 1


Thanks and regards,

  Jocelyn Fournier