The OpenStack Telemetry service (Ceilometer) is responsible for aggregating the usage and performance data across all the services deployed in an OpenStack environment. The data collected by the Telemetry service helps to monitor and analyze usage of the cloud resources across dozens of data points, as well as enables cloud administrators to view metrics both globally or by individually deployed resources. You can use the data gathered by the Telemetry service for capacity planning, alarming service, or in an external billing system. The Ceilometer framework provides capabilities to expand measurements for other needs. The Telemetry service collects a lot of data, as well as performs a large volume of database writes. For example, with 100 resources and default configuration Ceilometer collects around 16,000 samples per hour. You must plan to assign enough storage to address store this data. You can deploy MongoDB database for the Telemetry service on a dedicated node using the MongoDB virtual role provided by Fuel. Deploying MongoDB on a separate node resolves the Ceilometer performance issues caused by the volume of concurrent read/write operations.
A group of users or accounts that use services offered by a private or public cloud. A tenant may have no users or many, as well as each user can be associated with multiple tenants. A role can be assigned to each tenant and user pairing.
A Top of Rack switch, or TOR switch, is a switch to which all servers within one rack connect. A TOR switch in turn connects to the aggregation switch through the fiber cable. TOR switches are typical in multi-rack data centers that deploy the Top of Rack network architecture. Although, in many data centers TOR switches are located at the top of server racks for easier cabling management, a TOR switch can be placed in the middle or at the bottom of a server rack.