Struct pg_async::Cluster

pub struct Cluster { /* private fields */ }

A cluster of several replicated PostgreSQL nodes.

Implementations

impl Cluster

pub fn new() -> Result<Cluster, String>

Start the thread.

pub fn connect(&self, dsn: String, mul: u8) -> Result<(), String>

Setup a database connection to a [replicated] node.

Replicated clusters have several nodes usually. This method should be used for each node.

Nodes can be added on the go. E.g. add a first node, fire some queries, add a second node.

• dsn - A keyword=value connection string.
• mul - Pipelining support in libpq is currently limited to one batch of queries per connection, but parallelism can be increased by adding the same connection several times.

pub fn execute<I: IntoQueryPieces>(&self, sql: I) -> PgFuture

Schedule an SQL command to be executed on one of the nodes.

SQL will be run inside a transaction.

When using multiple statements, the library user must specify the exact number of top-level statements that the PostgreSQL server is going to see. For example:

  // Running two statements as a single op:
  cluster.execute ((2, "SELECT 1; SELECT 2"));
  // Running three statements:
  cluster.execute ((3, "\
    DELETE FROM foo; \
    INSERT INTO foo VALUES (1); \
    INSERT INTO foo VALUES (2)"));

To avoid SQL injection one might use the escapes provided by PgQueryPiece:

  use pg_async::PgQueryPiece::{Static as S, Plain as P, Literal as L, Bytea as B};
  cluster.execute (vec! [S ("SELECT * FROM foo WHERE bar = "), L (bar)]);

Trait Implementations

impl Drop for Cluster

fn drop(&mut self)

Executes the destructor for this type.