google.spanner. v1
Enumerations
Kind
number
The kind of PlanNode. Distinguishes between the two different kinds of nodes that can appear in a query plan.
Value |
|
---|---|
KIND_UNSPECIFIED |
Not specified. |
RELATIONAL |
Denotes a Relational operator node in the expression tree. Relational operators represent iterative processing of rows during query execution. For example, a |
SCALAR |
Denotes a Scalar node in the expression tree. Scalar nodes represent non-iterable entities in the query plan. For example, constants or arithmetic operators appearing inside predicate expressions or references to column names. |
QueryMode
number
Mode in which the statement must be processed.
Value |
|
---|---|
NORMAL |
The default mode. Only the statement results are returned. |
PLAN |
This mode returns only the query plan, without any results or execution statistics information. |
PROFILE |
This mode returns both the query plan and the execution statistics along with the results. |
TypeCode
number
TypeCode
is used as part of Type to indicate the type of a Cloud Spanner value.
Each legal value of a type can be encoded to or decoded from a JSON value, using the encodings described below. All Cloud Spanner values can be null
, regardless of type; null
s are always encoded as a JSON
null
.
Value |
|
---|---|
TYPE_CODE_UNSPECIFIED |
Not specified. |
BOOL |
Encoded as JSON |
INT64 |
Encoded as |
FLOAT64 |
Encoded as |
TIMESTAMP |
Encoded as If the schema has the column option
|
DATE |
Encoded as |
STRING |
Encoded as |
BYTES |
Encoded as a base64-encoded |
ARRAY |
Encoded as |
STRUCT |
Encoded as |
Properties
Kind
number
The kind of PlanNode. Distinguishes between the two different kinds of nodes that can appear in a query plan.
Value |
|
---|---|
KIND_UNSPECIFIED |
Not specified. |
RELATIONAL |
Denotes a Relational operator node in the expression tree. Relational operators represent iterative processing of rows during query execution. For example, a |
SCALAR |
Denotes a Scalar node in the expression tree. Scalar nodes represent non-iterable entities in the query plan. For example, constants or arithmetic operators appearing inside predicate expressions or references to column names. |
QueryMode
number
Mode in which the statement must be processed.
Value |
|
---|---|
NORMAL |
The default mode. Only the statement results are returned. |
PLAN |
This mode returns only the query plan, without any results or execution statistics information. |
PROFILE |
This mode returns both the query plan and the execution statistics along with the results. |
TypeCode
number
TypeCode
is used as part of Type to indicate the type of a Cloud Spanner value.
Each legal value of a type can be encoded to or decoded from a JSON value, using the encodings described below. All Cloud Spanner values can be null
, regardless of type; null
s are always encoded as a JSON
null
.
Value |
|
---|---|
TYPE_CODE_UNSPECIFIED |
Not specified. |
BOOL |
Encoded as JSON |
INT64 |
Encoded as |
FLOAT64 |
Encoded as |
TIMESTAMP |
Encoded as If the schema has the column option
|
DATE |
Encoded as |
STRING |
Encoded as |
BYTES |
Encoded as a base64-encoded |
ARRAY |
Encoded as |
STRUCT |
Encoded as |
Abstract types
BeginTransactionRequest
The request for BeginTransaction.
Properties
Parameter |
|
---|---|
session |
string Required. The session in which the transaction runs. |
options |
Object Required. Options for the new transaction. This object should have the same structure as TransactionOptions |
ChildLink
Metadata associated with a parent-child relationship appearing in a PlanNode.
Properties
Parameter |
|
---|---|
childIndex |
number The node to which the link points. |
type |
string The type of the link. For example, in Hash Joins this could be used to distinguish between the build child and the probe child, or in the case of the child being an output variable, to represent the tag associated with the output variable. |
variable |
string Only present if the child node is SCALAR and corresponds to an output variable of the parent node. The field carries the name of the output variable. For example, a |
CommitRequest
The request for Commit.
Properties
Parameter |
|
---|---|
session |
string Required. The session in which the transaction to be committed is running. |
transactionId |
string Commit a previously-started transaction. |
singleUseTransaction |
Object Execute mutations in a temporary transaction. Note that unlike commit of a previously-started transaction, commit with a temporary transaction is non-idempotent. That is, if the
This object should have the same structure as TransactionOptions |
mutations |
Array of Object The mutations to be executed when this transaction commits. All mutations are applied atomically, in the order they appear in this list. This object should have the same structure as Mutation |
CommitResponse
The response for Commit.
Property
Parameter |
|
---|---|
commitTimestamp |
Object The Cloud Spanner timestamp at which the transaction committed. This object should have the same structure as Timestamp |
CreateSessionRequest
The request for CreateSession.
Properties
Parameter |
|
---|---|
database |
string Required. The database in which the new session is created. |
session |
Object The session to create. This object should have the same structure as Session |
Delete
Arguments to delete operations.
Properties
Parameter |
|
---|---|
table |
string Required. The table whose rows will be deleted. |
keySet |
Object Required. The primary keys of the rows within table to delete. Delete is idempotent. The transaction will succeed even if some or all rows do not exist. This object should have the same structure as KeySet |
DeleteSessionRequest
The request for DeleteSession.
Property
Parameter |
|
---|---|
name |
string Required. The name of the session to delete. |
ExecuteSqlRequest
The request for ExecuteSql and ExecuteStreamingSql.
Properties
Parameter |
|
---|---|
session |
string Required. The session in which the SQL query should be performed. |
transaction |
Object The transaction to use. If none is provided, the default is a temporary read-only transaction with strong concurrency. The transaction to use. For queries, if none is provided, the default is a temporary read-only transaction with strong concurrency. Standard DML statements require a ReadWrite transaction. Single-use transactions are not supported (to avoid replay). The caller must either supply an existing transaction ID or begin a new transaction. Partitioned DML requires an existing PartitionedDml transaction ID. This object should have the same structure as TransactionSelector |
sql |
string Required. The SQL string. |
params |
Object The SQL string can contain parameter placeholders. A parameter placeholder consists of Parameters can appear anywhere that a literal value is expected. The same parameter name can be used more than once, for example:
It is an error to execute an SQL statement with unbound parameters. Parameter values are specified using This object should have the same structure as Struct |
paramTypes |
Object with Object properties It is not always possible for Cloud Spanner to infer the right SQL type from a JSON value. For example, values of type In these cases, |
resumeToken |
string If this request is resuming a previously interrupted SQL statement execution, |
queryMode |
number Used to control the amount of debugging information returned in ResultSetStats. If partition_token is set, query_mode can only be set to QueryMode.NORMAL. The number should be among the values of QueryMode |
partitionToken |
string If present, results will be restricted to the specified partition previously created using PartitionQuery(). There must be an exact match for the values of fields common to this message and the PartitionQueryRequest message used to create this partition_token. |
seqno |
number A per-transaction sequence number used to identify this request. This makes each request idempotent such that if the request is received multiple times, at most one will succeed. The sequence number must be monotonically increasing within the transaction. If a request arrives for the first time with an out-of-order sequence number, the transaction may be aborted. Replays of previously handled requests will yield the same response as the first execution. Required for DML statements. Ignored for queries. |
Field
Message representing a single field of a struct.
Properties
Parameter |
|
---|---|
name |
string The name of the field. For reads, this is the column name. For SQL queries, it is the column alias (e.g., |
type |
Object The type of the field. This object should have the same structure as Type |
GetSessionRequest
The request for GetSession.
Property
Parameter |
|
---|---|
name |
string Required. The name of the session to retrieve. |
KeyRange
KeyRange represents a range of rows in a table or index.
A range has a start key and an end key. These keys can be open or closed, indicating if the range includes rows with that key.
Keys are represented by lists, where the ith value in the list corresponds to the ith component of the table or index primary key. Individual values are encoded as described here.
For example, consider the following table definition:
CREATE TABLE UserEvents (
UserName STRING(MAX),
EventDate STRING(10)
) PRIMARY KEY(UserName, EventDate);
The following keys name rows in this table:
["Bob", "2014-09-23"]
["Alfred", "2015-06-12"]
Since the UserEvents
table's PRIMARY KEY
clause names two columns, each UserEvents
key has two elements; the first is the
UserName
, and the second is the EventDate
.
Key ranges with multiple components are interpreted lexicographically by component using the table or index key's declared sort order. For example, the following range returns all events for user "Bob"
that occurred
in the year 2015:
"start_closed": ["Bob", "2015-01-01"]
"end_closed": ["Bob", "2015-12-31"]
Start and end keys can omit trailing key components. This affects the inclusion and exclusion of rows that exactly match the provided key components: if the key is closed, then rows that exactly match the provided components are included; if the key is open, then rows that exactly match are not included.
For example, the following range includes all events for "Bob"
that occurred during and after the year 2000:
"start_closed": ["Bob", "2000-01-01"]
"end_closed": ["Bob"]
The next example retrieves all events for "Bob"
:
"start_closed": ["Bob"]
"end_closed": ["Bob"]
To retrieve events before the year 2000:
"start_closed": ["Bob"]
"end_open": ["Bob", "2000-01-01"]
The following range includes all rows in the table:
"start_closed": []
"end_closed": []
This range returns all users whose UserName
begins with any character from A to C:
"start_closed": ["A"]
"end_open": ["D"]
This range returns all users whose UserName
begins with B:
"start_closed": ["B"]
"end_open": ["C"]
Key ranges honor column sort order. For example, suppose a table is defined as follows:
CREATE TABLE DescendingSortedTable {
Key INT64,
...
) PRIMARY KEY(Key DESC);
The following range retrieves all rows with key values between 1 and 100 inclusive:
"start_closed": ["100"]
"end_closed": ["1"]
Note that 100 is passed as the start, and 1 is passed as the end, because Key
is a descending column in the schema.
Properties
Parameter |
|
---|---|
startClosed |
Object If the start is closed, then the range includes all rows whose first This object should have the same structure as ListValue |
startOpen |
Object If the start is open, then the range excludes rows whose first
This object should have the same structure as ListValue |
endClosed |
Object If the end is closed, then the range includes all rows whose first This object should have the same structure as ListValue |
endOpen |
Object If the end is open, then the range excludes rows whose first
This object should have the same structure as ListValue |
KeySet
KeySet
defines a collection of Cloud Spanner keys and/or key ranges. All the keys are expected to be in the same table or index. The keys need not be sorted in any particular way.
If the same key is specified multiple times in the set (for example if two ranges, two keys, or a key and a range overlap), Cloud Spanner behaves as if the key were only specified once.
Properties
Parameter |
|
---|---|
keys |
Array of Object A list of specific keys. Entries in This object should have the same structure as ListValue |
ranges |
Array of Object A list of key ranges. See KeyRange for more information about key range specifications. This object should have the same structure as KeyRange |
all |
boolean For convenience |
ListSessionsRequest
The request for ListSessions.
Properties
Parameter |
|
---|---|
database |
string Required. The database in which to list sessions. |
pageSize |
number Number of sessions to be returned in the response. If 0 or less, defaults to the server's maximum allowed page size. |
pageToken |
string If non-empty, |
filter |
string
An expression for filtering the results of the request. Filter rules are case insensitive. The fields eligible for filtering are:
Some examples of using filters are:
|
ListSessionsResponse
The response for ListSessions.
Properties
Parameter |
|
---|---|
sessions |
Array of Object The list of requested sessions. This object should have the same structure as Session |
nextPageToken |
string
|
Mutation
A modification to one or more Cloud Spanner rows. Mutations can be applied to a Cloud Spanner database by sending them in a Commit call.
Properties
Parameter |
|
---|---|
insert |
Object Insert new rows in a table. If any of the rows already exist, the write or transaction fails with error This object should have the same structure as Write |
update |
Object Update existing rows in a table. If any of the rows does not already exist, the transaction fails with error This object should have the same structure as Write |
insertOrUpdate |
Object Like insert, except that if the row already exists, then its column values are overwritten with the ones provided. Any column values not explicitly written are preserved. This object should have the same structure as Write |
replace |
Object Like insert, except that if the row already exists, it is deleted, and the column values provided are inserted instead. Unlike insert_or_update, this means any values not explicitly written become This object should have the same structure as Write |
delete |
Object Delete rows from a table. Succeeds whether or not the named rows were present. This object should have the same structure as Delete |
PartialResultSet
Partial results from a streaming read or SQL query. Streaming reads and SQL queries better tolerate large result sets, large rows, and large values, but are a little trickier to consume.
Properties
Parameter |
|
---|---|
metadata |
Object Metadata about the result set, such as row type information. Only present in the first response. This object should have the same structure as ResultSetMetadata |
values |
Array of Object A streamed result set consists of a stream of values, which might be split into many Most values are encoded based on type as described here. It is possible that the last value in values is "chunked", meaning that the rest of the value is sent in subsequent
Some examples of merging:
For a more complete example, suppose a streaming SQL query is yielding a result set whose rows contain a single string field. The following
This sequence of This object should have the same structure as Value |
chunkedValue |
boolean If true, then the final value in values is chunked, and must be combined with more values from subsequent |
resumeToken |
string Streaming calls might be interrupted for a variety of reasons, such as TCP connection loss. If this occurs, the stream of results can be resumed by re-sending the original request and including
|
stats |
Object Query plan and execution statistics for the statement that produced this streaming result set. These can be requested by setting ExecuteSqlRequest.query_mode and are sent only once with the last response in the stream. This field will also be present in the last response for DML statements. This object should have the same structure as ResultSetStats |
Partition
Information returned for each partition returned in a PartitionResponse.
Property
Parameter |
|
---|---|
partitionToken |
string This token can be passed to Read, StreamingRead, ExecuteSql, or ExecuteStreamingSql requests to restrict the results to those identified by this partition token. |
PartitionedDml
Message type to initiate a Partitioned DML transaction.
PartitionOptions
Options for a PartitionQueryRequest and PartitionReadRequest.
Properties
Parameter |
|
---|---|
partitionSizeBytes |
number Note: This hint is currently ignored by PartitionQuery and PartitionRead requests. The desired data size for each partition generated. The default for this option is currently 1 GiB. This is only a hint. The actual size of each partition may be smaller or larger than this size request. |
maxPartitions |
number Note: This hint is currently ignored by PartitionQuery and PartitionRead requests. The desired maximum number of partitions to return. For example, this may be set to the number of workers available. The default for this option is currently 10,000. The maximum value is currently 200,000. This is only a hint. The actual number of partitions returned may be smaller or larger than this maximum count request. |
PartitionQueryRequest
The request for PartitionQuery
Properties
Parameter |
|
---|---|
session |
string Required. The session used to create the partitions. |
transaction |
Object Read only snapshot transactions are supported, read/write and single use transactions are not. This object should have the same structure as TransactionSelector |
sql |
string The query request to generate partitions for. The request will fail if the query is not root partitionable. The query plan of a root partitionable query has a single distributed union operator. A distributed union operator conceptually divides one or more tables into multiple splits, remotely evaluates a subquery independently on each split, and then unions all results. This must not contain DML commands, such as INSERT, UPDATE, or DELETE. Use ExecuteStreamingSql with a PartitionedDml transaction for large, partition-friendly DML operations. |
params |
Object The SQL query string can contain parameter placeholders. A parameter placeholder consists of Parameters can appear anywhere that a literal value is expected. The same parameter name can be used more than once, for example:
It is an error to execute an SQL query with unbound parameters. Parameter values are specified using This object should have the same structure as Struct |
paramTypes |
Object with Object properties It is not always possible for Cloud Spanner to infer the right SQL type from a JSON value. For example, values of type In these cases, |
partitionOptions |
Object Additional options that affect how many partitions are created. This object should have the same structure as PartitionOptions |
PartitionReadRequest
The request for PartitionRead
Properties
Parameter |
|
---|---|
session |
string Required. The session used to create the partitions. |
transaction |
Object Read only snapshot transactions are supported, read/write and single use transactions are not. This object should have the same structure as TransactionSelector |
table |
string Required. The name of the table in the database to be read. |
index |
string If non-empty, the name of an index on table. This index is used instead of the table primary key when interpreting key_set and sorting result rows. See key_set for further information. |
columns |
Array of string The columns of table to be returned for each row matching this request. |
keySet |
Object Required. It is not an error for the This object should have the same structure as KeySet |
partitionOptions |
Object Additional options that affect how many partitions are created. This object should have the same structure as PartitionOptions |
PartitionResponse
The response for PartitionQuery or PartitionRead
Properties
Parameter |
|
---|---|
partitions |
Array of Object Partitions created by this request. This object should have the same structure as Partition |
transaction |
Object Transaction created by this request. This object should have the same structure as Transaction |
PlanNode
Node information for nodes appearing in a QueryPlan.plan_nodes.
Properties
Parameter |
|
---|---|
index |
number The |
kind |
number Used to determine the type of node. May be needed for visualizing different kinds of nodes differently. For example, If the node is a SCALAR node, it will have a condensed representation which can be used to directly embed a description of the node in its parent. The number should be among the values of Kind |
displayName |
string The display name for the node. |
childLinks |
Array of Object List of child node This object should have the same structure as ChildLink |
shortRepresentation |
Object Condensed representation for SCALAR nodes. This object should have the same structure as ShortRepresentation |
metadata |
Object Attributes relevant to the node contained in a group of key-value pairs. For example, a Parameter Reference node could have the following information in its metadata:
This object should have the same structure as Struct |
executionStats |
Object The execution statistics associated with the node, contained in a group of key-value pairs. Only present if the plan was returned as a result of a profile query. For example, number of executions, number of rows/time per execution etc. This object should have the same structure as Struct |
QueryPlan
Contains an ordered list of nodes appearing in the query plan.
Property
Parameter |
|
---|---|
planNodes |
Array of Object The nodes in the query plan. Plan nodes are returned in pre-order starting with the plan root. Each PlanNode's This object should have the same structure as PlanNode |
ReadOnly
Message type to initiate a read-only transaction.
Properties
Parameter |
|
---|---|
strong |
boolean Read at a timestamp where all previously committed transactions are visible. |
minReadTimestamp |
Object Executes all reads at a timestamp >= This is useful for requesting fresher data than some previous read, or data that is fresh enough to observe the effects of some previously committed transaction whose timestamp is known. Note that this option can only be used in single-use transactions. A timestamp in RFC3339 UTC \"Zulu\" format, accurate to nanoseconds. Example: This object should have the same structure as Timestamp |
maxStaleness |
Object Read data at a timestamp >= Useful for reading the freshest data available at a nearby replica, while bounding the possible staleness if the local replica has fallen behind. Note that this option can only be used in single-use transactions. This object should have the same structure as Duration |
readTimestamp |
Object Executes all reads at the given timestamp. Unlike other modes, reads at a specific timestamp are repeatable; the same read at the same timestamp always returns the same data. If the timestamp is in the future, the read will block until the specified timestamp, modulo the read's deadline. Useful for large scale consistent reads such as mapreduces, or for coordinating many reads against a consistent snapshot of the data. A timestamp in RFC3339 UTC \"Zulu\" format, accurate to nanoseconds. Example: This object should have the same structure as Timestamp |
exactStaleness |
Object Executes all reads at a timestamp that is Guarantees that all writes that have committed more than the specified number of seconds ago are visible. Because Cloud Spanner chooses the exact timestamp, this mode works even if the client's local clock is substantially skewed from Cloud Spanner commit timestamps. Useful for reading at nearby replicas without the distributed timestamp negotiation overhead of This object should have the same structure as Duration |
returnReadTimestamp |
boolean If true, the Cloud Spanner-selected read timestamp is included in the Transaction message that describes the transaction. |
ReadRequest
The request for Read and StreamingRead.
Properties
Parameter |
|
---|---|
session |
string Required. The session in which the read should be performed. |
transaction |
Object The transaction to use. If none is provided, the default is a temporary read-only transaction with strong concurrency. This object should have the same structure as TransactionSelector |
table |
string Required. The name of the table in the database to be read. |
index |
string If non-empty, the name of an index on table. This index is used instead of the table primary key when interpreting key_set and sorting result rows. See key_set for further information. |
columns |
Array of string The columns of table to be returned for each row matching this request. |
keySet |
Object Required. If the partition_token field is empty, rows are yielded in table primary key order (if index is empty) or index key order (if index is non-empty). If the partition_token field is not empty, rows will be yielded in an unspecified order. It is not an error for the This object should have the same structure as KeySet |
limit |
number If greater than zero, only the first |
resumeToken |
string If this request is resuming a previously interrupted read,
|
partitionToken |
string If present, results will be restricted to the specified partition previously created using PartitionRead(). There must be an exact match for the values of fields common to this message and the PartitionReadRequest message used to create this partition_token. |
ReadWrite
Message type to initiate a read-write transaction. Currently this transaction type has no options.
ResultSet
Results from Read or ExecuteSql.
Properties
Parameter |
|
---|---|
metadata |
Object Metadata about the result set, such as row type information. This object should have the same structure as ResultSetMetadata |
rows |
Array of Object Each element in This object should have the same structure as ListValue |
stats |
Object Query plan and execution statistics for the SQL statement that produced this result set. These can be requested by setting ExecuteSqlRequest.query_mode. DML statements always produce stats containing the number of rows modified, unless executed using the ExecuteSqlRequest.QueryMode.PLAN ExecuteSqlRequest.query_mode. Other fields may or may not be populated, based on the ExecuteSqlRequest.query_mode. This object should have the same structure as ResultSetStats |
ResultSetMetadata
Metadata about a ResultSet or PartialResultSet.
Properties
Parameter |
|
---|---|
rowType |
Object Indicates the field names and types for the rows in the result set. For example, a SQL query like
This object should have the same structure as StructType |
transaction |
Object If the read or SQL query began a transaction as a side-effect, the information about the new transaction is yielded here. This object should have the same structure as Transaction |
ResultSetStats
Additional statistics about a ResultSet or PartialResultSet.
Properties
Parameter |
|
---|---|
queryPlan |
Object QueryPlan for the query associated with this result. This object should have the same structure as QueryPlan |
queryStats |
Object Aggregated statistics from the execution of the query. Only present when the query is profiled. For example, a query could return the statistics as follows:
This object should have the same structure as Struct |
rowCountExact |
number Standard DML returns an exact count of rows that were modified. |
rowCountLowerBound |
number Partitioned DML does not offer exactly-once semantics, so it returns a lower bound of the rows modified. |
RollbackRequest
The request for Rollback.
Properties
Parameter |
|
---|---|
session |
string Required. The session in which the transaction to roll back is running. |
transactionId |
string Required. The transaction to roll back. |
Session
A session in the Cloud Spanner API.
Properties
Parameter |
|
---|---|
name |
string The name of the session. This is always system-assigned; values provided when creating a session are ignored. |
labels |
Object with string properties
The labels for the session.
|
createTime |
Object Output only. The timestamp when the session is created. This object should have the same structure as Timestamp |
approximateLastUseTime |
Object Output only. The approximate timestamp when the session is last used. It is typically earlier than the actual last use time. This object should have the same structure as Timestamp |
ShortRepresentation
Condensed representation of a node and its subtree. Only present for
SCALAR
PlanNode(s).
Properties
Parameter |
|
---|---|
description |
string A string representation of the expression subtree rooted at this node. |
subqueries |
Object with number properties A mapping of (subquery variable name) -> (subquery node id) for cases where the |
StructType
StructType
defines the fields of a STRUCT type.
Property
Parameter |
|
---|---|
fields |
Array of Object The list of fields that make up this struct. Order is significant, because values of this struct type are represented as lists, where the order of field values matches the order of fields in the StructType. In turn, the order of
fields matches the order of columns in a read request, or the order of fields in the This object should have the same structure as Field |
Transaction
A transaction.
Properties
Parameter |
|
---|---|
id |
string
Single-use read-only transactions do not have IDs, because single-use transactions do not support multiple requests. |
readTimestamp |
Object For snapshot read-only transactions, the read timestamp chosen for the transaction. Not returned by default: see TransactionOptions.ReadOnly.return_read_timestamp. A timestamp in RFC3339 UTC \"Zulu\" format, accurate to nanoseconds. Example: This object should have the same structure as Timestamp |
TransactionOptions
Transactions
Each session can have at most one active transaction at a time. After the active transaction is completed, the session can immediately be re-used for the next transaction. It is not necessary to create a new session for each transaction.
Transaction Modes
Cloud Spanner supports three transaction modes:
-
Locking read-write. This type of transaction is the only way to write data into Cloud Spanner. These transactions rely on pessimistic locking and, if necessary, two-phase commit. Locking read-write transactions may abort, requiring the application to retry.
-
Snapshot read-only. This transaction type provides guaranteed consistency across several reads, but does not allow writes. Snapshot read-only transactions can be configured to read at timestamps in the past. Snapshot read-only transactions do not need to be committed.
-
Partitioned DML. This type of transaction is used to execute a single Partitioned DML statement. Partitioned DML partitions the key space and runs the DML statement over each partition in parallel using separate, internal transactions that commit independently. Partitioned DML transactions do not need to be committed.
For transactions that only read, snapshot read-only transactions provide simpler semantics and are almost always faster. In particular, read-only transactions do not take locks, so they do not conflict with read-write transactions. As a consequence of not taking locks, they also do not abort, so retry loops are not needed.
Transactions may only read/write data in a single database. They may, however, read/write data in different tables within that database.
Locking Read-Write Transactions
Locking transactions may be used to atomically read-modify-write data anywhere in a database. This type of transaction is externally consistent.
Clients should attempt to minimize the amount of time a transaction is active. Faster transactions commit with higher probability and cause less contention. Cloud Spanner attempts to keep read locks active as long as the transaction continues to do reads, and the transaction has not been terminated by Commit or Rollback. Long periods of inactivity at the client may cause Cloud Spanner to release a transaction's locks and abort it.
Conceptually, a read-write transaction consists of zero or more reads or SQL statements followed by Commit. At any time before Commit, the client can send a Rollback request to abort the transaction.
Semantics
Cloud Spanner can commit the transaction if all read locks it acquired are still valid at commit time, and it is able to acquire write locks for all writes. Cloud Spanner can abort the transaction for any reason. If a commit attempt returns
ABORTED
, Cloud Spanner guarantees that the transaction has not modified any user data in Cloud Spanner.
Unless the transaction commits, Cloud Spanner makes no guarantees about how long the transaction's locks were held for. It is an error to use Cloud Spanner locks for any sort of mutual exclusion other than between Cloud Spanner transactions themselves.
Retrying Aborted Transactions
When a transaction aborts, the application can choose to retry the whole transaction again. To maximize the chances of successfully committing the retry, the client should execute the retry in the same session as the original attempt. The original session's lock priority increases with each consecutive abort, meaning that each attempt has a slightly better chance of success than the previous.
Under some circumstances (e.g., many transactions attempting to modify the same row(s)), a transaction can abort many times in a short period before successfully committing. Thus, it is not a good idea to cap the number of retries a transaction can attempt; instead, it is better to limit the total amount of wall time spent retrying.
Idle Transactions
A transaction is considered idle if it has no outstanding reads or SQL queries and has not started a read or SQL query within the last 10 seconds. Idle transactions can be aborted by Cloud Spanner so that they don't hold on to locks indefinitely.
In that case, the commit will fail with error ABORTED
.
If this behavior is undesirable, periodically executing a simple SQL query in the transaction (e.g., SELECT 1
) prevents the transaction from becoming idle.
Snapshot Read-Only Transactions
Snapshot read-only transactions provides a simpler method than locking read-write transactions for doing several consistent reads. However, this type of transaction does not support writes.
Snapshot transactions do not take locks. Instead, they work by choosing a Cloud Spanner timestamp, then executing all reads at that timestamp. Since they do not acquire locks, they do not block concurrent read-write transactions.
Unlike locking read-write transactions, snapshot read-only transactions never abort. They can fail if the chosen read timestamp is garbage collected; however, the default garbage collection policy is generous enough that most applications do not need to worry about this in practice.
Snapshot read-only transactions do not need to call Commit or Rollback (and in fact are not permitted to do so).
To execute a snapshot transaction, the client specifies a timestamp bound, which tells Cloud Spanner how to choose a read timestamp.
The types of timestamp bound are:
- Strong (the default).
- Bounded staleness.
- Exact staleness.
If the Cloud Spanner database to be read is geographically distributed, stale read-only transactions can execute more quickly than strong or read-write transaction, because they are able to execute far from the leader replica.
Each type of timestamp bound is discussed in detail below.
Strong
Strong reads are guaranteed to see the effects of all transactions that have committed before the start of the read. Furthermore, all rows yielded by a single read are consistent with each other -- if any part of the read observes a transaction, all parts of the read see the transaction.
Strong reads are not repeatable: two consecutive strong read-only transactions might return inconsistent results if there are concurrent writes. If consistency across reads is required, the reads should be executed within a transaction or at an exact read timestamp.
See TransactionOptions.ReadOnly.strong.
Exact Staleness
These timestamp bounds execute reads at a user-specified timestamp. Reads at a timestamp are guaranteed to see a consistent prefix of the global transaction history: they observe modifications done by all transactions with a commit timestamp <= the read timestamp, and observe none of the modifications done by transactions with a larger commit timestamp. They will block until all conflicting transactions that may be assigned commit timestamps <= the read timestamp have finished.
The timestamp can either be expressed as an absolute Cloud Spanner commit timestamp or a staleness relative to the current time.
These modes do not require a "negotiation phase" to pick a timestamp. As a result, they execute slightly faster than the equivalent boundedly stale concurrency modes. On the other hand, boundedly stale reads usually return fresher results.
See TransactionOptions.ReadOnly.read_timestamp and TransactionOptions.ReadOnly.exact_staleness.
Bounded Staleness
Bounded staleness modes allow Cloud Spanner to pick the read timestamp, subject to a user-provided staleness bound. Cloud Spanner chooses the newest timestamp within the staleness bound that allows execution of the reads at the closest available replica without blocking.
All rows yielded are consistent with each other -- if any part of the read observes a transaction, all parts of the read see the transaction. Boundedly stale reads are not repeatable: two stale reads, even if they use the same staleness bound, can execute at different timestamps and thus return inconsistent results.
Boundedly stale reads execute in two phases: the first phase negotiates a timestamp among all replicas needed to serve the read. In the second phase, reads are executed at the negotiated timestamp.
As a result of the two phase execution, bounded staleness reads are usually a little slower than comparable exact staleness reads. However, they are typically able to return fresher results, and are more likely to execute at the closest replica.
Because the timestamp negotiation requires up-front knowledge of which rows will be read, it can only be used with single-use read-only transactions.
See TransactionOptions.ReadOnly.max_staleness and TransactionOptions.ReadOnly.min_read_timestamp.
Old Read Timestamps and Garbage Collection
Cloud Spanner continuously garbage collects deleted and overwritten data in the background to reclaim storage space. This process is known as "version GC". By default, version GC reclaims versions after they are one hour old. Because
of this, Cloud Spanner cannot perform reads at read timestamps more than one hour in the past. This restriction also applies to in-progress reads and/or SQL queries whose timestamp become too old while executing. Reads and SQL queries
with too-old read timestamps fail with the error FAILED_PRECONDITION
.
Partitioned DML Transactions
Partitioned DML transactions are used to execute DML statements with a different execution strategy that provides different, and often better, scalability properties for large, table-wide operations than DML in a ReadWrite transaction. Smaller scoped statements, such as an OLTP workload, should prefer using ReadWrite transactions.
Partitioned DML partitions the keyspace and runs the DML statement on each partition in separate, internal transactions. These transactions commit automatically when complete, and run independently from one another.
To reduce lock contention, this execution strategy only acquires read locks on rows that match the WHERE clause of the statement. Additionally, the smaller per-partition transactions hold locks for less time.
That said, Partitioned DML is not a drop-in replacement for standard DML used in ReadWrite transactions.
-
The DML statement must be fully-partitionable. Specifically, the statement must be expressible as the union of many statements which each access only a single row of the table.
-
The statement is not applied atomically to all rows of the table. Rather, the statement is applied atomically to partitions of the table, in independent transactions. Secondary index rows are updated atomically with the base table rows.
-
Partitioned DML does not guarantee exactly-once execution semantics against a partition. The statement will be applied at least once to each partition. It is strongly recommended that the DML statement should be idempotent to avoid unexpected results. For instance, it is potentially dangerous to run a statement such as
UPDATE table SET column = column + 1
as it could be run multiple times against some rows. -
The partitions are committed automatically - there is no support for Commit or Rollback. If the call returns an error, or if the client issuing the ExecuteSql call dies, it is possible that some rows had the statement executed on them successfully. It is also possible that statement was never executed against other rows.
-
Partitioned DML transactions may only contain the execution of a single DML statement via ExecuteSql or ExecuteStreamingSql.
-
If any error is encountered during the execution of the partitioned DML operation (for instance, a UNIQUE INDEX violation, division by zero, or a value that cannot be stored due to schema constraints), then the operation is stopped at that point and an error is returned. It is possible that at this point, some partitions have been committed (or even committed multiple times), and other partitions have not been run at all.
Given the above, Partitioned DML is good fit for large, database-wide, operations that are idempotent, such as deleting old rows from a very large table.
Properties
Parameter |
|
---|---|
readWrite |
Object Transaction may write. Authorization to begin a read-write transaction requires
This object should have the same structure as ReadWrite |
partitionedDml |
Object Partitioned DML transaction. Authorization to begin a Partitioned DML transaction requires
This object should have the same structure as PartitionedDml |
readOnly |
Object Transaction will not write. Authorization to begin a read-only transaction requires
This object should have the same structure as ReadOnly |
TransactionSelector
This message is used to select the transaction in which a Read or ExecuteSql call runs.
See TransactionOptions for more information about transactions.
Properties
Parameter |
|
---|---|
singleUse |
Object Execute the read or SQL query in a temporary transaction. This is the most efficient way to execute a transaction that consists of a single SQL query. This object should have the same structure as TransactionOptions |
id |
string Execute the read or SQL query in a previously-started transaction. |
begin |
Object Begin a new transaction and execute this read or SQL query in it. The transaction ID of the new transaction is returned in ResultSetMetadata.transaction, which is a Transaction. This object should have the same structure as TransactionOptions |
Type
Type
indicates the type of a Cloud Spanner value, as might be stored in a table cell or returned from an SQL query.
Properties
Parameter |
|
---|---|
code |
number Required. The TypeCode for this type. The number should be among the values of TypeCode |
arrayElementType |
Object If code == ARRAY, then This object should have the same structure as Type |
structType |
Object If code == STRUCT, then This object should have the same structure as StructType |
Write
Arguments to insert, update, insert_or_update, and replace operations.
Properties
Parameter |
|
---|---|
table |
string Required. The table whose rows will be written. |
columns |
Array of string The names of the columns in table to be written. The list of columns must contain enough columns to allow Cloud Spanner to derive values for all primary key columns in the row(s) to be modified. |
values |
Array of Object The values to be written. This object should have the same structure as ListValue |