Percona Monitoring and Management 2.2.1

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

For PMM install instructions, see Installing PMM Server and Installing PMM Client.

⚠️ Caution
PMM 2 is designed to be used as a new installation — please don’t try to upgrade your existing PMM 1 environment.

PMM Server version 2.2.0 suffered an unauthenticated denial of service vulnerability (CVE-2020-7920). Any other PMM versions do not carry the same code logic, and are thus unaffected by this issue. Users who have already deployed PMM Server 2.2.0 are advised to upgrade to version 2.2.1 which resolves this issue.

Improvements and new features

• PMM-5229: The new RDS Exporter section added to the Prometheus Exporter Status dashboard shows singlestats and charts related to the rds_exporter
• PMM-5228 and PMM-5238: The Prometheus dashboard and the Exporters Overview dashboard were updated to include the rds_exporter metrics in their charts, allowing better understanding of the impacts of monitoring RDS instances
• PMM-4830: The consistency of the applied filters between the Query Analytics and the Overview dashboards was implemented, and now filters selected in QAN will continue to be active after the switch to any of the Overview dashboards available in the Services menu
• PMM-5235: The DB uptime singlestats in node rows on the Home dashboard were changed to show minimal values instead of average ones to be consistent with the top row
• PMM-5127: The “Search by” bar on the Query Analytics dashboard was renamed to “Filter by” to make its purpose more clear
• PMM-5131: The Filter panel on the Query Analytics dashboard now shows the total number of available Labels within the “See all” link, which appears if the Filter panel section shows only top 5 of its Labels

Fixed bugs

• PMM-5232: The pmm-managed component of the PMM Server 2.2.0 is vulnerable to DoS attacks, that could be carried out by anyone who knows the PMM Server IP address (CVE-2020-7920). Versions other than 2.2.0 are not affected.
• PMM-5226: The handlebars package was updated to version 4.5.3 because of the Prototype Pollution vulnerability in it (CVE-2019-19919). Please note PMM versions were not affected by this vulnerability, as handlebars package is used as a build dependency only.
• PMM-5206: Switching to the Settings dashboard was breaking the visual style of some elements on the Home dashboard
• PMM-5139: The breadcrumb panel, which shows all dashboards visited within one session starting from the root, was unable to fully show breadcrumb longer than one line
• PMM-5212: The explanatory text was added to the Download PMM Server Logs button in the Diagnostic section of the PMM Settings dashboard, and a link to it was added to the Prometheus dashboard which was the previous place to download logs
• PMM-5215: The unneeded mariadb-libs package was removed from the PMM Server 2.2.0 OVF image, resulting in both faster updating with the yum update command and avoiding dependency conflict messages in the update logs
• PMM-5216: PMM Server Upgrade to 2.2.0 was showing Grafana Update Error page with the Refresh button which had to be clicked to start using the updated version
• PMM-5211: The “Where do I get the security credentials for my Amazon RDS DB instance” link in the Add AWS RDS MySQL or Aurora MySQL instance dialog was not targeted at the appropriate instruction
• PMM-5217: PMM 2.x OVF Image memory size was increased from 1 Gb to 4 Gb with the additional 1 Gb swap space because the previous amount was hardly housing the PMM Server, and it wasn’t enough in some cases like performing an upgrade
• PMM-5271: LVM logical volumes were wrongly resized on AWS deployment, resulting in “no space left on device” errors
• PMM-5295: InnoDB Transaction Rollback Rate values on the MySQL InnoDB Details dashboard were calculated incorrectly
• PMM-5270: PXC/Galera Cluster Summary dashboard was showing empty Cluster drop-down list, making it impossible to choose the cluster name
• PMM-4769: The wrongly named “Timeout value used for retransmitting” singlestat on the Network Details dashboard was renamed to “The algorithm used to determine the timeout value” and updated to show the algorithm name instead of a digital code
• PMM-5260: Extensive resource consumption by pmm-agent took place in case of Query Analytics for PostgreSQL; it was fixed by a number of optimizations in the code, resulting in about 4 times smaller memory usage
• PMM-5261: CPU usage charts on all dashboards which contain them have undergone colors update to make softIRQ and Steal curves better differentiated
• PMM-5244: High memory consumption in the PMM Server with a large number of agents sending data simultaneously was fixed by improving bulk data insertion to the ClickHouse database