Yesterday, colleagues in the boss from the time, from the library reported the following error:
Got fatal error 1236 from master when reading data from binary log: ' misconfigured master-server ID is not set '
Rough looks like the master Server-id is not set, but colleagues do the following query:
Repository acquisition: [email protected] Fri may 14:18:59 14:18:59 [(none)]> show variables like '%server_id% '; +---------------+- ------+| variable_name | Value |+---------------+-------+| server_id | 2 | +---------------+-------+ Main Library capture:mysql> show variables like '%server% '; +-------------------- --+-------------------+| Variable_name | Value |+----------------------+-------------------+| character_set_server | latin1 | | collation_server | latin1_swedish_ci | | server_id | 1 |+----------------------+-------------------+
At this point, the Server-id of the replication environment is truly globally unique. Why the error?
Since the business is the metadata used to store Hadoop, the previous MySQL version was lower, and colleagues used a higher slave than the master version, which was thought to be the reason, but found on the MySQL bug list:
The findings negate our previous judgments, tested by colleagues, and indeed the bug:-)
Have a good weekend,
Good luck!