Avoid unnecessary restarts for replication

[Morgan Jones]

When a slave is attached to a master, the master site is configured as a master and restarted. This is unnecessary if the master site already has a slave attached (and therefore binary logging is already enabled). Avoid restarting the master in such cases.

[Nikhil Manchanda (slicknik)]

Spoke to Morgan regarding this, and we're good with moving it out to Kilo since it's not critical.

[Nikhil Manchanda (slicknik)]

Hi Morgan -- any updates on this?

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.openstack.org/169890

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/169890
Committed: https://git.openstack.org/cgit/openstack/trove/commit/?id=267a81625f46ad9dd6d0f2ad6fe516c86820c9df
Submitter: Jenkins
Branch: master

commit 267a81625f46ad9dd6d0f2ad6fe516c86820c9df
Author: Morgan Jones <email address hidden>
Date: Wed Apr 1 15:16:35 2015 -0400

    Avoid unnecessary restart of replication master

    Eliminates the unnecessary restart of the master site
    in a replication set when a new slave is added to a
    previously existing master site.

    Also removed some dead code.

    Change-Id: Ia76cca5cb084270c45ce782b7cecc3e17f8402d8
    Closes-bug: #1368419