Class DCAwareRoundRobinPolicy

cassandra/policies.py:203–318

Similar to :class:`.RoundRobinPolicy`, but prefers hosts in the local datacenter and only uses nodes in remote datacenters as a last resort.

class DCAwareRoundRobinPolicy(LoadBalancingPolicy):
    """
    Similar to :class:`.RoundRobinPolicy`, but prefers hosts
    in the local datacenter and only uses nodes in remote
    datacenters as a last resort.
    """

    local_dc = None
    used_hosts_per_remote_dc = 0

    def __init__(self, local_dc='', used_hosts_per_remote_dc=0):
        """
        The `local_dc` parameter should be the name of the datacenter
        (such as is reported by ``nodetool ring``) that should
        be considered local. If not specified, the driver will choose
        a local_dc based on the first host among :attr:`.Cluster.contact_points`
        having a valid DC. If relying on this mechanism, all specified
        contact points should be nodes in a single, local DC.

        `used_hosts_per_remote_dc` controls how many nodes in
        each remote datacenter will have connections opened
        against them. In other words, `used_hosts_per_remote_dc` hosts
        will be considered :attr:`~.HostDistance.REMOTE` and the
        rest will be considered :attr:`~.HostDistance.IGNORED`.
        By default, all remote hosts are ignored.
        """
        self.local_dc = local_dc
        self.used_hosts_per_remote_dc = used_hosts_per_remote_dc
        self._dc_live_hosts = {}
        self._position = 0
        self._endpoints = []
        LoadBalancingPolicy.__init__(self)

    def _dc(self, host):
        return host.datacenter or self.local_dc

    def populate(self, cluster, hosts):
        for dc, dc_hosts in groupby(hosts, lambda h: self._dc(h)):
            self._dc_live_hosts[dc] = tuple(set(dc_hosts))

        if not self.local_dc:
            self._endpoints = [
                endpoint
                for endpoint in cluster.endpoints_resolved]

        self._position = randint(0, len(hosts) - 1) if hosts else 0

    def distance(self, host):
        dc = self._dc(host)
        if dc == self.local_dc:
            return HostDistance.LOCAL

        if not self.used_hosts_per_remote_dc:
            return HostDistance.IGNORED
        else:
            dc_hosts = self._dc_live_hosts.get(dc)
            if not dc_hosts:
                return HostDistance.IGNORED