I have two galera clusters C1 and C2 each with two nodes. I have async replication set between node ‘C1N1’ of cluster C1 and ‘C2N1’ of cluster C2 (C1N1 —> C2N1).
The binlog format for both C1 and C2 is ROW. I have tested the replication with basic create table and insert table commands and it works as expected.
As soon as I run the pt-table-checksum command on C1N1, I see that the replication link C1N1 —> C2N1 has broken. Running the ‘show slave status\G’ command on C2N1 shows this error message: Last_SQL_Errno: 1047
Last_SQL_Error: Node has dropped from cluster
The pt-table-checksum was run with the following parameters: ./pt-table-checksum h=10.65.99.221 --user=root --password=pass123 --replicate mydb.checksums --no-check-binlog-format --ignore-databases=mysql
Any pointers please