Functionality Added or Changed
An Export as PNG option was added to the graphs. Hovering over a graph will reveal this option, to go along with the Export as CSV option. It includes the graph as a PNG image, and includes the title of the graph. (Bug #16925223)
The MySQL Enterprise Service Manager installation failed if there was a space in the service manager's repository password. (Bug #17490022)
custom.xml would append to the
configuration service's state stores, instead of replacing it.
--migration-agent option was failing with an
unclear error message when the Agent could not connect to at
least one of the previously configured MySQL instances.
The display name and instance notes could not be edited for MySQL instances monitored by a 2.3.x Agent. This information is edited from the MySQL Instances dashboard. (Bug #17464967)
Upgrading an Agent would overwrite a custom
custom.xml configuration file, instead of
saving the custom values.
The Purge Operations graph would display a straight line if the selected time interval was longer than the purge configuration value. (Bug #17383397)
On the MySQL Instances dashboard, the emergency icons next to the MySQL instance names did not refresh automatically, and instead required manual page reloading. (Bug #17375926)
Upgrading a MySQL Enterprise Monitor Agent from 3.0.x to 3.0.x would produce an
file when choosing "No" for "Backup the current installation
[Y/n]" during the Agent upgrade.
Under certain circumstances, the SNMP trap Agent address for Advisor traps was sometimes 0.0.0.0, when the "Use the remote MySQL agent host IP address as the SNMP trap agent address for Advisor traps" setting was enabled. (Bug #17343962)
The InnoDB Buffer Pool Usage graph would sometimes fail to load the first time it was selected.
The First Time Setup page now links to the related help section in the MySQL Enterprise Monitor manual.
The Asset Selector did not indicate when a MySQL Instance is not monitored due to a downed Agent. It is now red and italicized.
After clicking Configure Query Analysis on the Query Analysis page, the message on top of the Advisors page would refer to closed events when events were not closed.
The Host Details message was improved for remote hosts, and now includes the remotely monitored host address that is being monitored.
When a pop-up was active, pressing tab past the Cancel button would lose focus to the greyed out page in the background, instead of cycling back to the first element in the pop-up.
Unhandled exceptions would be generated (and fill the logs) when an older event was processed after a newer event of the same type was closed.
Now when multiple Advisors are selected, only the schedule options that all Advisors can support are visible in the Scheduling configuration section. Most commonly, this will mean the "Default", "Fixed Rate" and "Disabled" options will be present.
The MySQL Backup Failed Advisor now raises an "Emergency" level error instead of the "Critical" level.
For a 3.0.x to 3.0.x upgrade, the generated
configuration_report.txt could contain
incorrect MySQL server connection information, as multiple MySQL
instances might be monitored. Now,
configuration_report.txt does not contain
MySQL server connection information for upgrades, but does for
new installations if MySQL server information is entered during