class Google::Apis::SpannerV1::TransactionOptions

as deleting old rows from a very large table.
DML is good fit for large, database-wide, operations that are idempotent, such
and other partitions have not been run at all. Given the above, Partitioned
point, some partitions have been committed (or even committed multiple times),
stopped at that point and an error is returned. It is possible that at this
value that cannot be stored due to schema constraints), then the operation is
operation (for instance, a UNIQUE INDEX violation, division by zero, or a
If any error is encountered during the execution of the partitioned DML
execution of a single DML statement via ExecuteSql or ExecuteStreamingSql. -
against other rows. - Partitioned DML transactions may only contain the
on them successfully. It is also possible that statement was never executed
ExecuteSql call dies, it is possible that some rows had the statement executed
Rollback. If the call returns an error, or if the client issuing the
The partitions are committed automatically - there is no support for Commit or
column = column + 1` as it could be run multiple times against some rows. -
it is potentially dangerous to run a statement such as ‘UPDATE table SET
DML statement should be idempotent to avoid unexpected results. For instance,
applied at least once to each partition. It is strongly recommended that the
exactly-once execution semantics against a partition. The statement will be
atomically with the base table rows. - Partitioned DML does not guarantee
table, in independent transactions. Secondary index rows are updated
of the table. Rather, the statement is applied atomically to partitions of the
single row of the table. - The statement is not applied atomically to all rows
must be expressible as the union of many statements which each access only a
The DML statement must be fully-partitionable. Specifically, the statement
not a drop-in replacement for standard DML used in ReadWrite transactions. -
partition transactions hold locks for less time. That said, Partitioned DML is
match the WHERE clause of the statement. Additionally, the smaller per-
lock contention, this execution strategy only acquires read locks on rows that
automatically when complete, and run independently from one another. To reduce
partition in separate, internal transactions. These transactions commit
Partitioned DML partitions the keyspace and runs the DML statement on each
such as an OLTP workload, should prefer using ReadWrite transactions.
wide operations than DML in a ReadWrite transaction. Smaller scoped statements,
provides different, and often better, scalability properties for large, table-
are used to execute DML statements with a different execution strategy that
change stream TVFs. Partitioned DML transactions: Partitioned DML transactions
google.com/spanner/docs/change-streams for more details on how to query the
be discarded and not used for any subsequent queries. Please see cloud.
the transaction, instead of a valid read timestamp. This special value should
value of 2^63 - 2 will be returned in the Transaction message that describes
TransactionOptions.read_only.return_read_timestamp is set to true, a special
TransactionOptions are invalid for change stream queries. In addition, if
accessible using the strong read-only timestamp_bound. All other
range of interest. All change records within the retention period is
allows users to specify the start_timestamp and end_timestamp for the time
transaction with a strong read-only timestamp_bound. The change stream TVF
must be executed using the ExecuteStreamingSql API with a single-use read-only
from the name of the change stream: READ_. All queries on change stream TVFs
ExecuteStreamingSql API. The name of the TVF for a change stream is generated
be used to query the change records in the associated change stream using the
automatically defines a corresponding SQL Table-Valued Function (TVF) that can
set of columns in a database. When a change stream is created, Spanner
configured to watch data changes on the entire database, a set of tables, or a
Querying change Streams: A Change Stream is a schema object that can be
which allows Cloud Spanner to perform reads up to one week in the past.
VERSION_RETENTION_PERIOD` of a database up to a period as long as one week,
error `FAILED_PRECONDITION`. You can configure and extend the `
executing. Reads and SQL queries with too-old read timestamps fail with the
progress reads and/or SQL queries whose timestamp become too old while
timestamps more than one hour in the past. This restriction also applies to in-
one hour old. Because of this, Cloud Spanner cannot perform reads at read
known as “version GC”. By default, version GC reclaims versions after they are
overwritten data in the background to reclaim storage space. This process is
garbage collection: Cloud Spanner continuously garbage collects deleted and
TransactionOptions.ReadOnly.min_read_timestamp. Old read timestamps and
transactions. See TransactionOptions.ReadOnly.max_staleness and
which rows will be read, it can only be used with single-use read-only
replica. Because the timestamp negotiation requires up-front knowledge of
to return fresher results, and are more likely to execute at the closest
slower than comparable exact staleness reads. However, they are typically able
of the two phase execution, bounded staleness reads are usually a little
the second phase, reads are executed at the negotiated timestamp. As a result
phase negotiates a timestamp among all replicas needed to serve the read. In
inconsistent results. Boundedly stale reads execute in two phases: the first
the same staleness bound, can execute at different timestamps and thus return
Boundedly stale reads are not repeatable: two stale reads, even if they use
the read observes a transaction, all parts of the read see the transaction.
blocking. All rows yielded are consistent with each other – if any part of
allows execution of the reads at the closest available replica without
Cloud Spanner chooses the newest timestamp within the staleness bound that
Spanner to pick the read timestamp, subject to a user-provided staleness bound.
exact_staleness. Bounded staleness: Bounded staleness modes allow Cloud
TransactionOptions.ReadOnly.read_timestamp and TransactionOptions.ReadOnly.
other hand, boundedly stale reads usually return fresher results. See
slightly faster than the equivalent boundedly stale concurrency modes. On the
require a “negotiation phase” to pick a timestamp. As a result, they execute
timestamp or a staleness relative to the current time. These modes do not
The timestamp can either be expressed as an absolute Cloud Spanner commit
that may be assigned commit timestamps <= the read timestamp have finished.
larger commit timestamp. They will block until all conflicting transactions
timestamp, and observe none of the modifications done by transactions with a
all transactions with a commit timestamp less than or equal to the read
prefix of the global transaction history: they observe modifications done by
specified timestamp. Reads at a timestamp are guaranteed to see a consistent
strong. Exact staleness: These timestamp bounds execute reads at a user-
also specify the strong read timestamp bound. See TransactionOptions.ReadOnly.
read timestamp. Queries on change streams (see below for more details) must
required, the reads should be executed within a transaction or at an exact
results if there are concurrent writes. If consistency across reads is
two consecutive strong read-only transactions might return inconsistent
all parts of the read see the transaction. Strong reads are not repeatable:
consistent with each other – if any part of the read observes a transaction,
the start of the read. Furthermore, all rows yielded by a single read are
guaranteed to see the effects of all transactions that have committed before
timestamp bound is discussed in detail below. Strong: Strong reads are
because they are able to execute far from the leader replica. Each type of
transactions can execute more quickly than strong or read-write transactions,
database to be read is geographically distributed, stale read-only
the default). - Bounded staleness. - Exact staleness. If the Cloud Spanner
how to choose a read timestamp. The types of timestamp bound are: - Strong (
transaction, the client specifies a timestamp bound, which tells Cloud Spanner
Rollback (and in fact are not permitted to do so). To execute a snapshot
practice. Snapshot read-only transactions do not need to call Commit or
is generous enough that most applications do not need to worry about this in
timestamp is garbage collected; however, the default garbage collection policy
snapshot read-only transactions never abort. They can fail if the chosen read
concurrent read-write transactions. Unlike locking read-write transactions,
reads at that timestamp. Since they do not acquire locks, they do not block
Instead, they work by choosing a Cloud Spanner timestamp, then executing all
transaction does not support writes. Snapshot transactions do not take locks.
transactions for doing several consistent reads. However, this type of
only transactions provides a simpler method than locking read-write
transaction from becoming idle. Snapshot read-only transactions: Snapshot read-
a simple SQL query in the transaction (for example, `SELECT 1`) prevents the
with error `ABORTED`. If this behavior is undesirable, periodically executing
locks indefinitely. If an idle transaction is aborted, the commit will fail
transactions can be aborted by Cloud Spanner so that they don’t hold on to
and has not started a read or SQL query within the last 10 seconds. Idle
A transaction is considered idle if it has no outstanding reads or SQL queries
is better to limit the total amount of time spent retrying. Idle transactions:
good idea to cap the number of retries a transaction can attempt; instead, it
many times in a short period before successfully committing. Thus, it is not a
transactions attempting to modify the same row(s)), a transaction can abort
of success than the previous. Under some circumstances (for example, many
each consecutive abort, meaning that each attempt has a slightly better chance
as the original attempt. The original session’s lock priority increases with
committing the retry, the client should execute the retry in the same session
retry the whole transaction again. To maximize the chances of successfully
aborted transactions: When a transaction aborts, the application can choose to
exclusion other than between Cloud Spanner transactions themselves. Retrying
held for. It is an error to use Cloud Spanner locks for any sort of mutual
Cloud Spanner makes no guarantees about how long the transaction’s locks were
not modified any user data in Cloud Spanner. Unless the transaction commits,
attempt returns ‘ABORTED`, Cloud Spanner guarantees that the transaction has
writes. Cloud Spanner can abort the transaction for any reason. If a commit
still valid at commit time, and it is able to acquire write locks for all
Cloud Spanner can commit the transaction if all read locks it acquired are
the client can send a Rollback request to abort the transaction. Semantics:
more reads or SQL statements followed by Commit. At any time before Commit,
locks and abort it. Conceptually, a read-write transaction consists of zero or
inactivity at the client may cause Cloud Spanner to release a transaction’s
transaction has not been terminated by Commit or Rollback. Long periods of
locks active as long as the transaction continues to do reads, and the
probability and cause less contention. Cloud Spanner attempts to keep read
amount of time a transaction is active. Faster transactions commit with higher
transaction is externally consistent. Clients should attempt to minimize the
atomically read-modify-write data anywhere in a database. This type of
Locking read-write transactions: Locking transactions may be used to
may, however, read-write data in different tables within that database.
not needed. Transactions may only read-write data in a single database. They
consequence of not taking locks, they also do not abort, so retry loops are
do not take locks, so they do not conflict with read-write transactions. As a
semantics and are almost always faster. In particular, read-only transactions
transactions that only read, snapshot read-only transactions provide simpler
independently. Partitioned DML transactions do not need to be committed. For
partition in parallel using separate, internal transactions that commit
Partitioned DML partitions the key space and runs the DML statement over each
type of transaction is used to execute a single Partitioned DML statement.
TransactionOptions.ReadOnly.strong for more details. 3. Partitioned DML. This
the snapshot read-only transaction mode, specifying a strong read. Please see
do not need to be committed. Queries on change streams must be performed with
have committed before the start of the read). Snapshot read-only transactions
such that the read is guaranteed to see the effects of all transactions that
configured to perform a strong read (where Spanner will select a timestamp
read-only transactions can be configured to read at timestamps in the past, or
guaranteed consistency across several reads, but do not allow writes. Snapshot
to retry. 2. Snapshot read-only. Snapshot read-only transactions provide
commit. Locking read-write transactions may abort, requiring the application
These transactions rely on pessimistic locking and, if necessary, two-phase
This type of transaction is the only way to write data into Cloud Spanner.
modes: Cloud Spanner supports three transaction modes: 1. Locking read-write.
is not necessary to create a new session for each transaction. Transaction
completed, the session can immediately be re-used for the next transaction. It
count towards the one transaction limit). After the active transaction is
note that standalone reads and queries use a transaction internally and do
Transactions: Each session can have at most one active transaction at a time (