Re: Auto-Advancing Master Bin-Log Position
Thanks for the help! I'm actually doing the following at the moment:
CREATE EVENT mysql.keep_alive ON SCHEDULE EVERY 60 SECOND DO ALTER DATABASE <DB_NAME> CHARACTER SET latin1;
I haven't done really in-depth degradation tests for about two years. But at that time, we were testing with 10 machines, though with a special case of replication:
- 2 parent servers, who would replicate from each other on both a Main instance (3307) and a Logs instance (3308)
- 4 "sites" of slave servers, 2 per-site, where each site's server pair would again replicate from each other on 3307 and 3308 (like parents), and each slave server would also slave off of each parents' 3307 instances to local 3309 and 3310 instances
(4 repl/child * 4 children) + (2 repl/parent * 2 parents) = 16 + 4 = 20 repls
So, while we don't have "dozens" of slaves, the replication itself is not as straightforward as you'd think. Could this possibly compound the heartbeat problem?
Subject
Views
Written By
Posted
2129
March 24, 2014 02:43PM
998
March 25, 2014 06:58PM
Re: Auto-Advancing Master Bin-Log Position
836
April 09, 2014 07:07AM
1027
April 10, 2014 03:53PM
923
April 11, 2014 07:35AM
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.