The Router takes care of forwarding the request to the correct component.
It is the entry point of the Grid, all external requests will be received by it. The Router behaves differently depending on the request. If it is a new session request, the Router will add it to the New Session Queue. The Distributor regularly checks if there is a free slot. If so, the first matching request is removed from the New Session Queue. will receive the event and poll the New Session Queue to get the new session request. If the request belongs to an existing session, the Router will send the session id to the Session Map, and the Session Map will return the Node where the session is running. After this, the Router will forward the request to the Node.
The Router aims to balance the load in the Grid by sending the requests to the component that is able to handle them better, without overloading any component that is not needed in the process.
The Distributor is aware of all the Nodes and their capabilities. Its main role is to receive a new session request and find a suitable Node where the session can be created. After the session is created, the Distributor stores in the Session Map the relation between the session id and Node where the session is being executed.
A Node can be present several times in a Grid. Each Node takes care of managing the slots for the available browsers of the machine where it is running.
The Node registers itself to the Distributor through the Event Bus, and its configuration is sent as part of the registration message.
By default, the Node auto-registers all browser drivers available on the path of the machine where it runs. It also creates one slot per available CPU for Chromium based browsers and Firefox. For Safari and Internet Explorer, only one slot is created. Through a specific configuration, it can run sessions in Docker containers or relay commands. You can see more configuration details in the next section.
A Node only executes the received commands, it does not evaluate, make judgments, or control anything. The machines where the Node is running does not need to have the same operating system as the other components. For example, A Windows Node might have the capability of offering Internet Explorer as a browser option, whereas this would not be possible on Linux or Mac.
The Session Map is a data store that keeps the information of the session id and the Node where the session is running. It serves as a support for the Router in the process of forwarding a request to the Node. The Router will ask the Session Map for the Node associated to a session id.
New Session Queue
New Session Queue holds all the new session requests in a FIFO order. It has configurable parameters for setting the request timeout and request retry interval.
The Router adds the new session request to the New Session Queue and waits for the response. The New Session Queue regularly checks if any request in the queue has timed out, if so the request is rejected and removed immediately.
The Distributor regularly checks if a slot is available. If so, the Distributor requests the New Session Queue for the first matching request. The Distributor then attempts to create a new session.
Once the requested capabilities match the capabilities of any of the free Node slots, the Distributor attempts to get the available slot. If all the slots are busy, the Distributor will ask the queue to add the request to the front of the queue. If request times out while retrying or adding to the front of the queue it is rejected.
After a session is created successfully, the Distributor sends the session information to the New Session Queue. The New Session Queue sends the response back to the client.
The Event Bus serves as a communication path between the Nodes, Distributor, New Session Queue, and Session Map. The Grid does most of its internal communication through messages, avoiding expensive HTTP calls. When starting the Grid in its fully distributed mode, the Event Bus is the first component that should be started.