Techniques for workload spawning
View Patent ↗Techniques for spawning workloads are provided. A single repository is read once to obtain an image for a workload or files and resources for the image. The read operation spawns multiple, and in some cases, concurrent write operations, to instantiate the workload over a network as multiple occurrences or instances of the workload in multiple processing environments.
1. A method implemented in a non-transitory machine-readable storage medium and processed by one or more processors configured to perform the method, comprising:
configuring, by a first workload, a workload with a deployment policy, an identity-based authentication mechanism, and an access-control mechanism;
deploying, by the first workload, a first instance of the workload in a first processing environment utilizing the deployment policy, the identity-based authentication mechanism, and the access-control mechanism;
updating, by the first workload, the deployment policy to reflect an occurrence of the first instance; and
causing, by the first workload, the deployment of additional instances of the workload to additional processing environments using and updating the deployment policy for each additional occurrence of the workload and enforcing, by the first workload, the identity-based authentication mechanism and the access-control mechanism for the instance and the additional instances, and operating the instance and the additional instances independently but acting together as a single unit as managed by the first workload.
2. The method of claim 1 , wherein configuring further includes representing the workload as an image.
3. The method of claim 1 , wherein configuring further includes representing the workload as one or more configuration files.
4. The method of claim 1 , wherein configuring further includes receiving a request from a principal, the request identifies a total number of instances for the workload to deploy, the total number of instances updated in the deployment policy.
5. The method of claim 1 , wherein deploying the first instance further includes multicasting files or an image for the workload over a network to the first processing environment, the files or the image also multicast to the additional processing environments.
6. The method of claim 5 , wherein multicasting further includes initiating multiple heads on a hard drive in the first processing environment to write the files or the image on the hard drive of the first processing environment or using solid state drives or multiple buses in the first processing environment to write the files or the image to storage in the first processing environment.
7. The method of claim 6 , wherein initiating further includes starting additional multiple heads on each additional hard drive in the additional processing environments to write the files or the image on the additional hard drives of the additional processing environments.
8. The method of claim 1 , wherein updating further includes, decrementing a total number of instances for the workload by one within the deployment policy to reflect the first instance being deployed in the first processing environment.
9. The method of claim 1 , wherein causing further includes having the first instance from the first processing environment recursively initiate the deployment of the additional instances based on the deployment policy.
10. The method of claim 9 , wherein having further includes instructing each instance to update the deployment policy to reflect successful deployment of that instance within an image or files that are used for that instance for deployment to a particular additional processing environment.
11. The method of claim 1 further comprising, sending, by the first workload, messages to each instance within each instance's processing environment to monitor each instance.
12. A method implemented in a non-transitory machine-readable storage medium and processed by one or more processors configured to perform the method, comprising:
initiating an instance of a workload in a first processing environment as a workload image for the workload;
authenticating, by the instance, with an external identity service;
enforcing, by the instance, access control defined in the workload;
decrementing a total number of instances for the workload defined in a deployment policy by one; and
deploying a next instance of the workload when the total number of instances is non zero to a next processing environment, deployment occurs from the first processing environment to the next processing environment, and operating the instance and the next instance independently but acting together as a single unit as managed by the instance.
13. The method of claim 12 , wherein initiating further includes acquiring the instance from a remote repository that uses a multicast or broadcast protocol over a network from the remote repository to the first processing environment.
14. The method of claim 12 , wherein authentication further includes indirectly authenticating with the external identity service via communication by the instance with a first instance of the workload deployed over a network.
15. The method of claim 12 , wherein enforcing further includes obtaining access control policies for the access control from the external identity service.
16. The method of claim 12 , wherein decrementing further includes modifying the deployment policy as metadata carried with the instance.
17. The method of claim 12 , wherein decrementing further includes modifying the deployment policy as metadata associated with the workload and centrally housed in a remote repository over a network.
18. A system, comprising:
a deployment manager implemented in a non-transitory computer-readable storage medium for execution on one or more processors of a network; and
an instance of a workload implemented in a non-transitory computer-readable storage medium and to execute on one or more of the processors;
the deployment configured as a first workload and the deployment manager configured to deploy at least the instance to a target processing environment, and the instance including a deployment policy that identifies a total number of instances of the workload for deployment over the network, the instance configured to cause additional deployments of remaining instances of the workload over the network to additional target processing environments using an updated deployment policy, and operating the instance and the remaining instances independently but acting together as a single unit as managed by the instance.
19. The system of claim 18 , wherein the deployment manager configured to include authentication and access control services in configuration data associated with the workload.
20. The system of claim 18 , wherein each of the remaining instances decrement the total number of instances once operable in a particular additional target processing environment within the deployment policy.