Comment 9 for bug 1581580

Reviewed: https://review.openstack.org/379578
Committed: https://git.openstack.org/cgit/openstack/neutron/commit/?id=5c1516e1fb07bb3e026a3569396567c3907b31c6
Submitter: Jenkins
Branch: stable/newton

commit 5c1516e1fb07bb3e026a3569396567c3907b31c6
Author: venkata anil <email address hidden>
Date: Tue May 17 16:30:13 2016 +0000

    New option for num_threads for state change server

    Currently max number of client connections(i.e greenlets spawned at
    a time) opened at any time by the WSGI server is set to 100 with
    wsgi_default_pool_size[1].

    This configuration may be fine for neutron api server. But with
    wsgi_default_pool_size(=100) requests, state change server
    is creating heavy cpu load on agent.
    So this server(which run on agents) need lesser value i.e
    can be configured to half the number of cpu on agent

    We use "ha_keepalived_state_change_server_threads" config option
    to configure number of threads in state change server instead of
    wsgi_default_pool_size.

    [1] https://review.openstack.org/#/c/278007/

    DocImpact: Add new config option -
    ha_keepalived_state_change_server_threads, to configure number
    of threads in state change server.

    Closes-Bug: #1581580
    Change-Id: I822ea3844792a7731fd24419b7e90e5aef141993
    (cherry picked from commit 70ea188f5d87c45fb60ace8b8405274e5f6dd489)