If this is for PartitionedRead or PartitionedQuery and this field is set to true, the request
will be executed via Spanner independent compute resources.
Option to request DirectedRead for ReadOnlyTransaction and SingleUseTransaction.
The DirectedReadOptions can be used to indicate which replicas or regions should be used for
non-transactional reads or queries. Not all requests can be sent to non-leader replicas. In
particular, some requests such as reads within read-write transactions must be sent to a
designated leader replica. These requests ignore DirectedReadOptions.
Specifying this will help in optimistic concurrency control as a way to help prevent
simultaneous deletes of an instance config from overwriting each other. Operations that support
this option are:
Specifying this instructs the transaction to be excluded from being recorded in change streams
with the DDL option allow_txn_exclusion=true. This does not exclude the transaction from
being recorded in the change streams with the DDL option allow_txn_exclusion being false or
unset.
Specifying this instructs the transaction to request Optimistic Lock from the backend. In this
concurrency mode, operations during the execution phase, i.e., reads and queries, are performed
without acquiring locks, and transactional consistency is ensured by running a validation
process in the commit phase (when any needed locks are acquired). The validation process
succeeds only if there are no conflicting committed transactions (that committed mutations to
the read data at a commit timestamp after the read timestamp).
Specifying this will allow the client to prefetch up to prefetchChunks
PartialResultSet chunks for read and query. The data size of each chunk depends on the server
implementation but a good rule of thumb is that each chunk will be up to 1 MiB. Larger values
reduce the likelihood of blocking while consuming results at the cost of greater memory
consumption. prefetchChunks should be greater than 0. To get good performance choose a
value that is large enough to allow buffering of chunks for an entire row. Apart from the
buffered chunks, there can be at most one more row buffered in the client.
[[["Easy to understand","easyToUnderstand","thumb-up"],["Solved my problem","solvedMyProblem","thumb-up"],["Other","otherUp","thumb-up"]],[["Hard to understand","hardToUnderstand","thumb-down"],["Incorrect information or sample code","incorrectInformationOrSampleCode","thumb-down"],["Missing the information/samples I need","missingTheInformationSamplesINeed","thumb-down"],["Other","otherDown","thumb-down"]],["Last updated 2025-02-06 UTC."],[],[]]
