Percona Monitoring and Management 2.4.0

Percona Monitoring and Management (PMM) is a free and open-source platform for managing and monitoring MySQL, MongoDB, and PostgreSQL performance.

New Features

• PMM-3387: Prometheus custom configuration is now supported by PMM Server. The feature is targeted at experienced users and is done by adding the base configuration file into the PMM Server container to be parsed and included into the managed Prometheus configuration.
• PMM-5186: Including –-pprof option in the pmm-admin summary command adds pprof debug profiles to the diagnostics data archive
• PMM-5102: The new “Node Details” dashboard now displays data from the hardware monitoring sensors in hwmon. The new dashboard is based on the hwmon collector data from the node_exporter. Please note that data may be unavailable for some nodes because of the configuration or virtualization parameters.

Improvements

• PMM-4915: The Query Analytics dashboard now shows Time Metrics in the Profile Section as “AVG per query” instead of “AVG per second”

• PMM-5470: ClickHouse query optimized for Query Analytics to improve its speed and reduce the load on the back-end

• PMM-5448: The default high and medium metrics resolutions were changed to 1-5-30 and 5-10-60 sec. To reduce the effect of this change on existing installations, systems having the “old” high resolution chosen on the PMM Settings page (5-5-60 sec.) will be automatically re-configured to the medium one during an upgrade. If the resolution was changed to some custom values via API, it will not be affected

• PMM-5531: A health check indicator was implemented for the PMM Server Docker image. It is based on the Docker HEALTHCHECK. This feature can be used as follows:

  docker inspect -f {% raw %}{{.State.Health.Status}}{% endraw %}
  until [ "`docker inspect -f {% raw %}{{.State.Health.Status}}{% endraw %} pmm-server`" == "healthy" ]; do sleep 1; done

• PMM-5489: The “Total” line in all charts is now drawn with the same red color for better consistency

• PMM-5461: Memory graphs on the node-related dashboards were adjusted to have fixed colors that are more distinguishable from each other

• PMM-5329: Prometheus in PMM Server was updated to version 2.16.0. This update has brought several improvements. Among them are significantly reduced memory footprint of the loaded TSDB blocks, lower memory footprint for the compaction process (caused by the more balanced choice of what to buffer during compaction), and improved query performance for the queries that only touch the most recent 2 hours of data.

• PMM-5210: Data Retention is now specified in days instead of seconds on the PMM Settings page. Please note this is a UI-only change, so the actual data retention precision is not changed

• PMM-5182: The logs.zip archive available on the PMM Settings page now includes additional self-monitoring information in a separate client subfolder. This subfolder contains information collected on the PMM Server and is equivalent to the one collected on a node by the pmm-admin summary command.

• PMM-5112: The Inventory API List requests now can be filtered by the Node/Service/Agent type

Bugs Fixed

• PMM-5178: Query Detail Section of the Query Analytics dashboard didn’t show tables definitions and indexes for the internal PostgreSQL database
• PMM-5465: MySQL Instance related dashboards had row names not always matching the actual contents. To fix this, elements were re-ordered and additional rows were added for better matching of the row name and the corresponding elements
• PMM-5455: Dashboards from the Insight menu were fixed to work correctly when the low resolution is set on the PMM Settings page
• PMM-5446: A number of the Compare Dashboards were fixed to work correctly when the low resolution is set on the PMM Settings page
• PMM-5430: MySQL Exporter section on the Prometheus Exporter Status dashboard now collapsed by default to be consistent with other database-related sections
• PMM-5445, PMM-5439, PMM-5427, PMM-5426, PMM-5419: Labels change (which occurs e.g. when the metrics resolution is changed on the PMM Settings page) was breaking dashboards
• PMM-5347: Selecting queries on the Query Analytics dashboard was generating errors in the browser console
• PMM-5305: Some applied filters on the Query Analytics dashboard were not preserved after changing the time range
• PMM-5267: The Refresh button was not working on the Query Analytics dashboard
• PMM-5003: pmm-admin list and status use different JSON naming for the same data
• PMM-5526: A typo was fixed in the Replication Dashboard description tooltip

Help us improve our software quality by reporting any bugs you encounter using our bug tracking system.