Here’s a way to detect the sql query causing a lock or a session
to fail, and also to identify the o.s.pid if need be (btw, no
rocket science). “a” way.. I’m sure there are many others, so
feel free to suggest, please.
So, we’re using MCM, and have created a MySQL Cluster like
mentioned in the cluster intro session (in Spanish I’m afraid),
using 7.4.6, which comes with 5.6.24.
With the env up and running, set up a schema, some data and run a
few queries:
mysql> create database world;
mysql> use world;
Database changed
mysql> source world_ndb.sql
(world_ndb.sql, as you might guess, is the world_innodb tables
script, with a little adjustment as to which storage engine to be
used.)
Once created, let’s lock things up in Cluster:
mysql -uroot -h127.0.0.1 -P3306
mysql> use test; …
[Read more]