tsgrpc_kvstore Key-Value Store driver

The tsgrpc_kvstore driver connects to a gRPC server which implements tensorstore_grpc.kvstore.KvStoreService.

json kvstore/tsgrpc_kvstore : object

Read/write key-value store using the tensorstore-specific gRPC protocol.

JSON specification of the key-value store.

Extends:

  • KvStore — Key-value store specification.

Required members:
driver : "tsgrpc_kvstore"
address : string

gRPC Service Address.

An address of the grpc service.

Optional members:
path : string

Key prefix within the key-value store.

If the prefix is intended to correspond to a Unix-style directory path, it should end with "/".

context : Context

Specifies context resources that augment/override the parent context.

timeout : string

Timeout for requests to the gRPC service.

tsgrpc_request_retries : ContextResource = "tsgrpc_request_retries"

Specifies or references a previously defined Context.tsgrpc_request_retries. It is typically more convenient to specify a default tsgrpc_request_retries in the context.

data_copy_concurrency : ContextResource = "data_copy_concurrency"

Specifies or references a previously defined Context.data_copy_concurrency. It is typically more convenient to specify a default data_copy_concurrency in the context.

grpc_client_credentials : ContextResource = "grpc_client_credentials"

Specifies or references a previously defined Context.grpc_client_credentials.

json Context.data_copy_concurrency : object

Specifies a limit on the number of CPU cores used concurrently for data copying/encoding/decoding.

Optional members:
limit : integer[1, +∞) | "shared" = "shared"

The maximum number of CPU cores that may be used. If the special value of "shared" is specified, a shared global limit equal to the number of CPU cores/threads available applies.

json Context.tsgrpc_request_retries : object

Specifies retry parameters for handling transient network errors.

Optional members:
max_retries : integer[1, +∞) = 32

Maximum number of attempts in the case of transient errors.

initial_delay : string = "1s"

Initial backoff delay for transient errors.

max_delay : string = "32s"

Maximum backoff delay for transient errors.

Limitations

Note

This is an experimental driver and should be considered a work-in-progress. There are known limitations around large values and authenticated connections.