Dynatrace Managed feature update, version 1.166

Notification of pending license expiration

The menu bar across the top of the Cluster Management Console (CMC) now includes a License expires in … days! alert that counts down the days remaining in your contracted license period (see example below). The counter appears 60 days in advance of the expiration of your license, which gives you plenty of advance notice that it’s time to extend your Dynatrace Managed license.

License expiry notice in CMC

30 days after the license has expired, your Dynatrace Managed cluster will no longer receive updates (i.e., your cluster will no longer receive updated packages for new product releases). You’ll also miss out on email notifications, cluster health checks, proactive support, access to the Dynatrace Community portal, and any domains or certificates that are managed by Dynatrace. You will be notified on specific preferences pages within the CMC as to any discontinued functionality once your license has expired (see example below from the Automatic update preferences page).

Automatic update expiry notification

Once your license has been extended, all deactivated settings will be reverted to their previous states. Dynatrace will then provide you with the required updates to upgrade your cluster to the most current, supported version. At that time, you will regain access to the full Dynatrace Managed experience.

More efficient SSL/TLS handling for OneAgent traffic

By default, all OneAgent traffic is now routed to your embedded ActiveGate via NGINX on port 443. As announced with the release of Dynatrace Managed version 1.150, we now route all incoming traffic through NGINX in an effort to increase performance and ease configuration effort. For OneAgent traffic, the default port has been set to 443 for new clusters only so as to ensure backward compatibility.

Additional audit logs for Nodekeeper

Because the Nodekeeper is our central component for managing cluster nodes, we now log details about which users make changes to cluster components and exactly what those changes are. This relates to changes to Elasticsearch, Cassandra, configuration files, and more. We now also log component management operations (for example, start, stop, restart, reload) and include the ID of the responsible user in the audit trail.

More detail about ActiveGates that reside behind load balancers

In scenarios where a load balancer forwarded traffic to Cluster ActiveGates, the ActiveGate instances weren’t distinguishable on the CMC Home page as the nodes showed the domain name rather than the individual IP address. We now map the actual IP address to the node ID and display it in the deployment status section of the Home page, as shown in the example below.

Cluster ActiveGate information on CMC Home page

Also in this release

  • Node size categories in the CMC have been updated to match the node type, as documented in Dynatrace Managed hardware and system requirements.
  • Previously, when a Cluster ActiveGate on a cluster node was updated, launcheruserconfig.conf was reset to its default state, which overwrote any changes that had been made to it. We now preserve the current configuration during updates.
  • We’ve optimized backups by shrinking their size by excluding the same set of column families in the databases. We’ve also extended the backup timeout from 12 to 20 hours as 12 hours was barely sufficient for larger nodes.
  • The import_trusted_certificate.sh script (for details, see How to add a certificate to the server trust store) is now available in the utils directory of each Dynatrace Managed installation. This makes it easier for you to import certificates to the cluster key store.
  • When you’re logged into a specific node on your cluster via the Dynatrace web UI, for example, n01.<cluster-domain>/cmc, you previously weren’t able to remove the node. We’ve removed this restriction so you can now perform node removals regardless of how you’re logged in.

Resolved issues

  • As we faced a number of Nodekeeper instances in production that weren’t working correctly, we now explicitly fail cluster upgrades when a Nodekeeper upgrade can’t be successfully completed. Cluster health checks have been extended with the version number of the Nodekeeper instance.
  • We removed disk space alerts for Session Replay when Session Replay is inactive.
  • When users opened the OpenID configuration page, password managers tended to automatically fill in fields with incorrect user and password data. Wrong values were saved and configurations became invalid. This issue has been solved by adding a confirmation panel that forces each user to actively confirm any changes that are made on the OpenID configuration page.

Other new features

Additionally, all new features introduced with Dynatrace SaaS version 1.165 and version 1.166 are now also supported by Dynatrace Managed.