Identifying a Computation
Last updated
Last updated
Before computations can be executed, they need to be identified as Computation Definitions, which exist within the context of an associated . The joint computing environment establishes the state that the computation can interact with. Each Computation Definition includes details on the outputs of the computation and contains a version field to ensure that any future executions are based on the correct version of the definition.
Also, the computation definition specifies who has the authority to execute the computation and authority can be restricted in several ways: no one may have access (Null), a single entity may be authorized (Private), specific entities may be allowed (Reserved), or anyone may execute the computation (Public).
Each Computation Definition also includes parameters. These fall into two categories: Data Objects, which come from on-chain or off-chain sources and can be mutable or immutable, and Binary Data, which has a specified length in bits for raw data that will be used during the computation.
Shared key generation computation: These computations handle the shared key generation when a new is added to a and the computation customer can set high priority fees upon setup.
Transition system computation: These computations manage both forced and planned migrations of clusters and each in the cluster can contribute a variable amount to the priority fees based on how important fast execution is to them.
Inactivity detection computation: These computations occur when a cluster experiences non-participation issues that can’t be resolved internally. They require a broader consensus of nodes, and priority fees are carried over from the original computation causing the dispute.