Percona Monitoring and Management (PMM) is a free and open-source platform for managing and monitoring MySQL, MongoDB, and PostgreSQL performance.
## Release Highlights
- Switch to push metrics by default
In PMM 2.12.0, Percona replaced its metrics collection engine (formerly Prometheus) with VictoriaMetrics. Historically, PMM used a *pull* method with Prometheus while VictoriaMetrics can operate in either a *pull* or *push* method. When PMM 2.12.0 was released, Percona kept the default method as *pull*. Now with PMM 2.14.0, Percona is shifting the default to *push* for all newly-added instances. This [blog post](https://www.percona.com/blog/2020/12/01/foiled-by-the-firewall-a-tale-of-transition-from-prometheus-to-victoriametrics/) describes the two methods and why *push* benefits users. Also, [here is a post](https://www.percona.com/blog/2020/12/16/percona-monitoring-and-management-migration-from-prometheus-to-victoriametrics-faq/) by Peter Zaitzev of FAQs relating to the move to VictoriaMetrics and the push model. Documentation on the push method is [here](https://www.percona.com/doc/percona-monitoring-and-management/2.x/details/victoria-metrics.html).
**Note**: Installing the 2.14.0 or newer PMM server will change the default behavior on 2.12.0 and 2.13.0 clients from "pull" method to "push" for any newly added services. Existing services will remain in whatever mode they were prior to upgrade.
- DBaaS Preview phase 1.0 (Technical Preview)
In 2.13.0 we introduced Percona's Database as a Service (DBaaS) which enables non-DBAs (software architects, developers, site reliability engineers, etc.) to perform typical DBA tasks to manage an organization's database environment via user interfaces and automation orchestration. This release contains several enhancements and fixes, many directly from user feedback.
**Note**: This capability is feature-flagged and turned off by default. Users require a variable to be passed to PMM to expose this functionality.
- External services presentation on node summary dashboard
Improvements to the user experience for adding and viewing external services (any data that can be monitored by a Prometheus exporter such as: non-Percona supported databases like Redis, ElasticSearch, Cassandra, etc. or an organization's external application) on the Node Summary dashboard of PMM.
## New Features
- [PMM-5765](https://jira.percona.com/browse/PMM-5765): Ability to monitor External Services for situations where PMM Client can't be installed -- Uses a new command `pmm-admin add external-serverless`. (See [pmm-admin](../details/commands/pmm-admin.md).) (This is a **Technical Preview** feature)
- [PMM-7015](https://jira.percona.com/browse/PMM-7015): DBaaS Preview: Create DB cluster with randomly-generated password
- [PMM-7007](https://jira.percona.com/browse/PMM-7007): Integrated Alerting: Ability to copy (duplicate) alert rules
- [PMM-7006](https://jira.percona.com/browse/PMM-7006): Integrated Alerting: Ability to delete alert rules
- [PMM-6941](https://jira.percona.com/browse/PMM-6941): Integrated Alerting: Ability to delete alert rule templates
## Improvements
- [PMM-6985](https://jira.percona.com/browse/PMM-6985): DBaaS: Ability to force unregister Kubernetes cluster
- [PMM-7145](https://jira.percona.com/browse/PMM-7145): 'Push' metrics mode is default when adding services and nodes (All agents collecting data from Services and Nodes will now use PUSH model if not specified explicitly. You will still be able to use `--metrics-mode` flag to use Pull metrics if needed. All previously set up agents will keep their existing mode. To change these you need to remove and re-add them.)
- [PMM-7282](https://jira.percona.com/browse/PMM-7282): Integrated Alerting: Ability to create rule without channels and filters
- [PMM-7226](https://jira.percona.com/browse/PMM-7226): Integrated Alerting: Validate parameters during rule creation/update
- [PMM-7082](https://jira.percona.com/browse/PMM-7082): Integrated Alerting: Severity levels are color-coded
- [PMM-7065](https://jira.percona.com/browse/PMM-7065): Integrated Alerting: Show rule details for items in Alert Rules list
- [PMM-7048](https://jira.percona.com/browse/PMM-7048): DBaaS: Simplify Cluster creation by moving Create Cluster button to earlier steps
- [PMM-6993](https://jira.percona.com/browse/PMM-6993): Protect against possible problems with EXPLAIN of stored functions in MySQL -- We are fixing possible problems caused by an attempt to analyze queries covered in .
## Bugs Fixed
- [PMM-7312](https://jira.percona.com/browse/PMM-7312): Error when accessing Metrics data on Dashboards for large installations
- [PMM-7310](https://jira.percona.com/browse/PMM-7310): VictoriaMetrics consuming 100's Gb's of disk in `/tmp/searchResults` in PMM 2.13.0
- [PMM-5137](https://jira.percona.com/browse/PMM-5137): Swagger page redirect isn't working
- [PMM-7144](https://jira.percona.com/browse/PMM-7144): DBaaS: Creating DB cluster with same name (Thanks to Beata Handzelova for reporting this issue)
- [PMM-7323](https://jira.percona.com/browse/PMM-7323): DBaaS: 'Remove DB Cluster from Kubernetes Cluster' removes wrong one
- [PMM-7251](https://jira.percona.com/browse/PMM-7251): Integrated Alerting: Error `Rule with ID "mysql_version" not found` if both Security Threat Tool and Integrated Alerting enabled
- [PMM-7247](https://jira.percona.com/browse/PMM-7247): DBaaS: Disk size is always 0 for Percona XtraDB cluster
- [PMM-7178](https://jira.percona.com/browse/PMM-7178): `pg_stat_monitor` integration is broken with version 0.6.0 of the plugin
- [PMM-7169](https://jira.percona.com/browse/PMM-7169): Old data (from Prometheus) not deleted when Retention period expires
- [PMM-7105](https://jira.percona.com/browse/PMM-7105): Query Analytics: no 'Example' or 'Explain' data for MariaDB
- [PMM-7239](https://jira.percona.com/browse/PMM-7239): Integrated Alerting: Validate Slack channel names in Notification Channels
- [PMM-7213](https://jira.percona.com/browse/PMM-7213): MySQL InnoDB Details dashboard: remove color-coding on 'Data Buffer Pool Fit' element
- [PMM-7167](https://jira.percona.com/browse/PMM-7167): Some panels not visible when using long time intervals (e.g. 30 days)
- [PMM-7133](https://jira.percona.com/browse/PMM-7133): Incorrect descriptions for data links in dashboards
- [PMM-7103](https://jira.percona.com/browse/PMM-7103): VictoriaMetrics build logs not deleted from PMM Server Docker image
- [PMM-6904](https://jira.percona.com/browse/PMM-6904): `pmm-admin annotate` command crashes for non-generic node types
- [PMM-6902](https://jira.percona.com/browse/PMM-6902): No query Examples on PostgreSQL 12 with pg_stat_monitor
- [PMM-6838](https://jira.percona.com/browse/PMM-6838): ProxySQL Instance Summary dashboard: Incorrect "Hostgroup Size" formula
- [PMM-6490](https://jira.percona.com/browse/PMM-6490): `rds_exporter` crashes when more than 100 AWS RDS instances added (Thanks to for fixing this)
- [PMM-6096](https://jira.percona.com/browse/PMM-6096): `pmm-agent` connection checker does not check authentication for MongoDB
- [PMM-7303](https://jira.percona.com/browse/PMM-7303): Disk Details, Nodes Compare dashboards: 'Disk Utilization' description is confusing