MySQL Forums
Forum List  »  Replication

group replication out of order
Posted by: jifa Xu
Date: February 23, 2018 01:30AM

I have a multi-primary cluster with 3 nodes.

I write a program to insert into a table with infinite loop.

Then I follow below steps:

1. run program on node1
2. wait for a moment and kill the program, the data count of node1 is A, and node2 & node3 is lag behind node1 and still apply events
3. run truncate on node2 and this operation will be blocked for a while

At final, data count on node1 and node3 is 0, that is ok, but on node2, the data count is larger than 0 and smaller than A

And when I check binlog events, I found that the binlog evnets is not the same order on each node, the truncate event is the last event on node1 and node3, but on node2 the truncate event is in middle of binlog events.

Why could such thing happend?


Below is my main config:

log_bin=ON
sync_binlog=1
binlog_format=ROW
gtid_mode=ON
enforce_gtid_consistency=ON
binlog_checksum=NONE
log_slave_updates=ON

loose-group_replication_single_primary_mode=OFF
loose-group_replication_enforce_update_everywhere_checks=ON


Thanks for your reply!

Options: ReplyQuote


Subject
Views
Written By
Posted
group replication out of order
740
February 23, 2018 01:30AM


Sorry, you can't reply to this topic. It has been closed.

Content reproduced on this site is the property of the respective copyright holders. It is not reviewed in advance by Oracle and does not necessarily represent the opinion of Oracle or any other party.