Scheduling actions in multiple replicas of computing environment using consensus protocol
Embodiments of the present invention provide systems, methods, and software to manage actions across a cluster of replica computing nodes for a data structure. In one implementation, a first replica in the cluster generates a first action notification and receives one or more additional action notifications from at least one other replica in the cluster. The first replica further determines that consensus exists for the first action notification based on the one or more additional action notifications and, in response to determining that consensus exists, identifies one or more rules associated with the first action notification and initiates one or more actions corresponding to the one or more rules.
1. A method of operating replica in a cluster of replicas, the method comprising:
generating a first action notification at the replica by a processor;
receiving one or more additional action notifications, via a communication interface, from one or more other replicas in the cluster of replicas;
determining, by the processor, that consensus exists for the first action notification based on the one or more additional action notifications, the consensus being determined when a threshold number of replicas generate a same action notification;
in response to determining that consensus exists for the first action notification:
identifying one or more rules associated with the first action notification; and
performing one or more actions associated with the one or more rules on a data structure supported by the replica by a processing system causing a change to a storage system, the data structure being stored in the storage system.
2. The method of claim 1 , wherein the cluster of replicas comprises a cluster of virtual machines.
3. The method of claim 1 , wherein the one or more actions comprise a prune action, a transaction action, or a test action.
4. The method of claim 1 further comprising:
communicating the first action notification to at least one other replica in the cluster of replicas.
5. The method of claim 1 , wherein the first action notification comprises a timestamp and an identifier associated with at least the one or more rules.
6. The method of claim 1 further comprising:
receiving a new rule from a second replica;
determining that the new rule is valid based on a consensus determination from the cluster of replicas; and
initiating a trigger operation, wherein the trigger operation generates one or more action notifications in accordance with timing information associated with the new rule.
7. The method of claim 6 , wherein the timing information comprises periodic intervals for generating the one or more action notifications for the new rule.
8. The method of claim 1 , wherein generating the first action notification at the replica comprises generating the first action notification at an interval defined by the one or more rules.
9. A method of operating replica in a cluster of replicas, the method comprising:
generating a first action notification at the replica by a processor;
receiving one or more additional action notifications, via a communication interface, from one or more other replicas in the cluster of replicas;
determining, by the processor, that consensus exists for the first action notification based on the one or more additional action notifications, wherein the received action notifications include signatures, hashed values, timestamps, identifiers, or other information that can be used in determining consensus for locally generated action notifications;
in response to determining that consensus exists for the first action notification:
identifying one or more rules associated with the first action notification; and
performing one or more actions associated with the one or more rules on a data structure supported by the replica by a processing system causing a change to a storage system, the data structure being stored in the storage system.
10. The method of claim 9 , wherein the cluster of replicas comprises a cluster of virtual machines.
11. The method of claim 9 , wherein the one or more actions comprise at least one of a prune action, a transaction action, or a test action.
12. The method of claim 9 , further comprising communicating the first action notification to at least one other replica in the cluster of replicas.
13. The method of claim 9 , wherein the first action notification comprises a timestamp and an identifier associated with at least the one or more rules.
14. The method of claim 9 , wherein determining that consensus exists comprises applying a Byzantine Fault Tolerant (BFT) consensus protocol.
15. The method of claim 9 , wherein identifying the one or more rules comprises accessing at least one Cron table that schedules commands for the replica.
16. The method of claim 9 , further comprising scheduling implementation of at least one of the one or more actions at a defined period following a timestamp associated with the first action notification.
17. The method of claim 9 , wherein performing the one or more actions comprises implementing the one or more actions as State Machine Replication (SMR) operations.
18. The method of claim 9 , further comprising triggering the one or more actions in a sequence associated with a priority for each of the one or more actions, wherein the sequence is common amongst all replicas in the cluster of replicas.
19. The method of claim 9 , wherein generating the first action notification comprises generating the first action notification at an interval defined by at least one of the one or more rules.