June#

June 08#

Changes#

  • You can create node groups with GPUs in Kubernetes clusters.

June 18#

Changes#

  • Now, PaaS services can be scaled out. In certain high availability configurations, you can add nodes up to the maximum supported number of nodes, without stopping the service.

    Scale-out is supported for the following configurations that may have:

    • 3 to 6 master nodes (MongoDB and PostgreSQL database services);

    • 3 to 9 master nodes (Elasticsearch search and analytics service);

    • 3 to 9 master nodes (ELK logging service);

    • 3 to 9 master nodes (RabbitMQ message broker service);

    • 3 to 9 broker coordinator nodes or 3 to 9 broker nodes with 3 coordinator nodess (Apache Kafka message broker service).

  • You can view monitoring metrics for the installed PaaS services in the web interface. Metric graphs are displayed in the Monitoring tab on the service page. The set of metrics varies with the service type.

    Metrics are available for the following service types:

    • MongoDB, MySQL, PostgreSQL and Redis databases;

    • Apache Kafka and RabbitMQ message brokers;

    • Elasticsearch search and analytics service;

    • ELK logging.

  • Network interfaces of virtual switches can be used as a mirror source.

API#

  • The AddInstances method allows you to add instances to a PaaS service.

  • The GetConsoleAccesses method displays a list of permissions for access to the specified instance remote console.

Fixed#

  • Thresholds are displayed correctly on volume metric graphs.

  • If there are no metrics (for example, for a newly created instance), the alarm creation dialog or the alarm page returns no error.