The Perl Toolchain Summit needs more sponsors. If your company depends on Perl, please support this very important event.
 / 
Elasticsearch-1.00
8 ++
/ Elasticsearch::CxnPool::Sniff

NAME

Elasticsearch::CxnPool::Sniff - A CxnPool for connecting to a local cluster with a dynamic node list

VERSION

version 1.00

SYNOPSIS

    $e = Elasticsearch->new(
        cxn_pool => 'Sniff',
        nodes    => [
            'search1:9200',
            'search2:9200'
        ],
    );

DESCRIPTION

The Sniff connection pool should be used when you do have direct access to the Elasticsearch cluster, eg when your web servers and Elasticsearch servers are on the same network. The nodes that you specify are used to discover the cluster, which is then sniffed to find the current list of live nodes that the cluster knows about.

This sniff process is repeated regularly, or whenever a node fails, to update the list of healthy nodes. So if you add more nodes to your cluster, they will be auto-discovered during a sniff.

If all sniffed nodes fail, then it falls back to sniffing the original seed nodes that you specified in new().

For HTTP Cxn classes, this module will also dynamically detect the max_content_length which the nodes in the cluster will accept.

This class does Elasticsearch::Role::CxnPool::Sniff and Elasticsearch::Role::Is_Sync.

CONFIGURATION

nodes

The list of nodes to use to discover the cluster. Can accept a single node, multiple nodes, and defaults to localhost:9200 if no nodes are specified. See "node" in Elasticsearch::Role::Cxn::HTTP for details of the node specification.

sniff_interval

    $e = Elasticsearch->new(
        cxn_pool       => 'Sniff',
        nodes          => [...],
        sniff_interval => 300,
    );

How often to perform a sniff to detect new nodes in the cluster. Defaults to 300 seconds. Note: In case of node failure, the cluster will be resniffed before the next request to update the list of healthy nodes.

sniff_max_content_length

    $e = Elasticsearch->new(
        cxn_pool                 => 'Sniff',
        nodes                    => [...],
        sniff_max_content_length => 0,
    );

Whether max_content_length should be dynamically updated based on the value returned from each node in the cluster. This defaults to true unless you have manually set "max_content_length" in Elasticsearch::Cxn::HTTP.

See also

METHODS

next_cxn()

    $cxn = $cxn_pool->next_cxn

Returns the next available live node (in round robin fashion), or throws a NoNodes error if no nodes can be sniffed from the cluster.

schedule_check()

    $cxn_pool->schedule_check

Forces a sniff before the next Cxn is returned, to updated the list of healthy nodes in the cluster.

sniff()

    $bool = $cxn_pool->sniff

Sniffs the cluster and returns true if the sniff was successful.

should_accept_node()

    $host = $cxn_pool->should_accept_node($host,$node_id,\%node_data)

This method serves as a hook which can be overridden by the user. When a sniff is performed, this method is called with the host (eg 192.168.5.100:9200), the node_id (the ID assigned to the node by Elasticsearch) and the node_data which contains the information about the node that Elasticsearch has returned, eg:

    {
        "transport_address" => "inet[192.168.5.100/192.168.5.100:9300]",
        "http" : {
           "publish_address"    => "inet[/192.168.5.100:9200]",
           "max_content_length" => "100mb",
           "bound_address"      => "inet[/0:0:0:0:0:0:0:0:9200]",
           "max_content_length_in_bytes" : 104857600
        },
        "version"       => "0.90.4",
        "name"          => "Silver Sable",
        "hostname"      => "search1.domain.com",
        "http_address"  => "inet[/192.168.5.100:9200]"
    }

If the node should be accepted (ie used to serve data), then it should return the host value which to use. By default, nodes are always accepted.

AUTHOR

Clinton Gormley <drtech@cpan.org>

COPYRIGHT AND LICENSE

This software is Copyright (c) 2014 by Elasticsearch BV.

This is free software, licensed under:

  The Apache License, Version 2.0, January 2004