How does HA-JDBC handle merge? I am using HA-JDBC 2.0.15.
My setup is two servers connected by a switch. Each node is running one Tomcat
with one local MySQL database. I cut power to a switch which isolates the
nodes. When I restart the switch, I see jGroups initiate merge but it
eventually times out.
I see in HA-JDBC code that it listens for jGroups merge notification. I assume
HA-JDBC must decide which of the two database nodes is valid. How does it
decide, or do I have to implement some hook into HA-JDBC or jGroups to decide
and then notify completion?
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
The short answer (to which I elaborated in the previous post), HA-JDBC doesn't
handle this deployment scenario particularly well. Currently, HA-JDBC doesn't
choose winners upon merge - it just stops. I plan to do something a little
less crude in the current development branch.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
How does HA-JDBC handle merge? I am using HA-JDBC 2.0.15.
My setup is two servers connected by a switch. Each node is running one Tomcat
with one local MySQL database. I cut power to a switch which isolates the
nodes. When I restart the switch, I see jGroups initiate merge but it
eventually times out.
I see in HA-JDBC code that it listens for jGroups merge notification. I assume
HA-JDBC must decide which of the two database nodes is valid. How does it
decide, or do I have to implement some hook into HA-JDBC or jGroups to decide
and then notify completion?
The short answer (to which I elaborated in the previous post), HA-JDBC doesn't
handle this deployment scenario particularly well. Currently, HA-JDBC doesn't
choose winners upon merge - it just stops. I plan to do something a little
less crude in the current development branch.