NAME

    Database::Async::Engine::PostgreSQL - support for PostgreSQL databases
    in Database::Async

DESCRIPTION

METHODS

 connection

    Returns a Future representing the database connection, and will attempt
    to connect if we are not already connected.

 ssl

    Whether to try SSL or not, expected to be one of the following values
    from Protocol::Database::PostgreSQL::Constants:

      * SSL_REQUIRE

      * SSL_PREFER

      * SSL_DISABLE

 read_len

    Buffer read length. Higher values mean we will attempt to read more
    data for each I/O loop iteration.

    Defaults to 2 megabytes.

 write_len

    Buffer write length. Higher values mean we will attempt to write more
    data for each I/O loop iteration.

    Defaults to 2 megabytes.

 connect

    Establish a connection to the server.

    Returns a Future which resolves to the IO::Async::Stream once ready.

 negotiate_ssl

    Apply SSL negotiation.

 uri_for_dsn

    Returns a URI corresponding to the given database source name
    <https://en.wikipedia.org/wiki/Data_source_name>.

    May throw an exception if we don't have a valid string.

 stream

    The IO::Async::Stream representing the database connection.

 on_read

    Process incoming database packets.

    Expects the following parameters:

      * $stream - the IO::Async::Stream we are receiving data on

      * $buffref - a scalar reference to the current input data buffer

      * $eof - true if we have reached the end of input

 ryu

    Provides a Ryu::Async instance.

 outgoing

    Ryu::Source representing outgoing packets for the current database
    connection.

 incoming

    Ryu::Source representing incoming packets for the current database
    connection.

 authenticated

    Resolves once database authentication is complete.

 protocol

    Returns the Protocol::Database::PostgreSQL instance, creating it and
    setting up event handlers if necessary.

 set_parameter

    Marks a parameter update from the server.

 idle

    Resolves when we are idle and ready to process the next request.

Implementation notes

    Query sequence is essentially:

      * receive ReadyForQuery

      * send frontend_query

      * Row Description

      * Data Row

      * Command Complete

      * ReadyForQuery

    The DB creates an engine. The engine does whatever connection handling
    required, and eventually should reach a "ready" state. Once this
    happens, it'll notify DB to say "this engine is ready for queries". If
    there are any pending queries, the next in the queue is immediately
    assigned to this engine. Otherwise, the engine is pushed into the pool
    of available engines, awaiting query requests.

    On startup, the pool `min` count of engine instances will be
    instantiated. They start in the pending state.

    Any of the following:

      * tx

      * query

      * copy etc.

    is treated as "queue request". It indicates that we're going to send
    one or more commands over a connection.

    "next_engine" resolves with an engine instance:

      * check for engines in `available` queue - these are connected and
      waiting, and can be assigned immediately

      * next look for engines in `unconnected` - these are instantiated but
      need a ->connection first

AUTHOR

    Tom Molesworth <TEAM@cpan.org>

LICENSE

    Copyright Tom Molesworth 2011-2019. Licensed under the same terms as
    Perl itself.