-
Notifications
You must be signed in to change notification settings - Fork 26
Queue Server Configuration
This configuration is described by QueueServerConfiguration.swift source file. It defines how workers and queue operate.
This is an analytics configuration object
checkAgainTimeInterval defines how often workers poll the queue for buckets to execute. Use a value of 30 seconds.
Defines where a queue should be looked for, or where it should be started in case if it needs to be started.
Version 14 and below: The key is called queueServerDeploymentDestination (singular). Emcee will first try to scan the provided destination in order to find a running queue there. If queue won't be located, it will try to start the new queue on that destination.
Version 15 and above: The key is called "queueServerDeploymentDestinations" (plural). This is an array with destinations. Emcee will first try to scan all destinations in order to find a running queue there. If no running queue will be located, it will try to start the new queue on destinations in specified order.
Example below is for v.15 and above.
{
"queueServerDeploymentDestinations": [
{
"host": "example.com",
"port": 22,
"username": "ssh login",
"password": "ssh password",
"remoteDeploymentPath": "/path/to/working/dir/"
}
]
}This is an array of computers that will be used as Emcee workers. The computer that is running shared queue can also be Emcee worker.
Each element of this array describes how Emcee can reach the computer.
This is an array of Deployment Destination objects.
This is map from host name to Worker Specific Configuration object.
{
"workerSpecificConfigurations": {
"worker1.example.com": {"numberOfSimulators": 3},
"worker2.example.com": {"numberOfSimulators": 2},
}
}Note: currently worker id and host name are the same things.
Defines when queue server will automatically terminate itself.
Detailed explanation is given on Queue Termination Policy page.