Connectors and other external entities log any errors while
accessing servers so that MySQL Fabric can monitor server health and
act accordingly. For example, MySQL Fabric promotes a new master after
notifications from the number of
clients configured in
within the time interval configured in
notification_interval. If a server is
considered unstable but it is not a master, it is simply marked as
faulty. To avoid making the system unstable, a new master can only
be automatically promoted after the
failover_interval has been elapsed since the
last promotion. In order to ease the adoption of MySQL Fabric, a
built-in failure detector is provided. If the failure detector is
enabled to monitor a group, a new master is promoted after
3 failed successive attempts to access the
current master within the time interval configured in
failover_interval. The failure detection
routine tries to connect to servers in a group and uses the value
detection_timeout as timeout.
How often the internal event log is pruned, in seconds and also the age of events in the event log that is used to present statistics.
Number of issues before the server is considered unstable.
Number of different sources that should report issues on a server before it is considered unstable.
Amount of time in seconds that is used when deciding if a server is unstable. Issues older than this are not considered when deciding.
Minimum time in seconds between subsequent automatic promotions. This parameter is used to prevent the system entering a sequence of promotions that could disable the system.
Number of successive failed attempts to contact the server after which the built-in failure detector considers the server unstable.
Timeout in seconds used when contacting the server. If the server does not respond within this time period, it is recorded as a failed attempt to contact the server.
Maximum age in seconds for reported issues in the error log. Issues older than this are removed from the error log.