I need some inputs on how to handle this issue on a Percona XtraDB 5.7.18 cluster (2 master + 1 arbitrator).
We are trying to copy a table using “INSERT INTO tbname_new SELECT * FROM tbname;”. This has some 20+ Million rows and it keeps erroring out “WSREP detected deadlock/conflict and aborted the transaction.”
We tried even splitting it into smaller chunks of 200k rows and still it gets that error. Please note we are writing only to one node.
Any input on this is highly appreciate