Compute restart causes period of network ‘blackout’

[David McNally]

When a compute service is restarted each instance running on the host has its iptables rules built and applied sequentially during the host init stage. The impact of this, especially on a host running many instances, can be observed as a period where some instances are not accessible as the existing iptables rules have been torn down and not yet re-applied.

A suggested work-around for this would be a configurable/flagged deferred mode that would prevent the application of the iptables rules until all instances on the host had been initialised then the rules for all instances would be applied at once preventing a ‘blackout’ period.

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.openstack.org/11029

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/11029
Committed: http://github.com/openstack/nova/commit/8f1c54ce98fed9cb7384be9cbb9f28eba2f12c2d
Submitter: Jenkins
Branch: master

commit 8f1c54ce98fed9cb7384be9cbb9f28eba2f12c2d
Author: David McNally <email address hidden>
Date: Wed Aug 8 16:20:23 2012 +0100

    Compute restart causes period of network 'blackout'

    Fixes bug 1034401

    When a compute service is restarted each instance running on the
    host has its iptables rules built and applied sequentially during
    the host init stage. The impact of this, especially on a host
    running many instances, can be observed as a period where some
    instances are not accessible as the existing iptables rules have
    been torn down and not yet re-applied.

    The presented work-around for this is a configurable/flagged deferred
    mode that prevents the application of the iptables rules until all
    instances on the host had been initialised then the rules for all
    instances are applied all at once preventing a 'blackout' period.

    Change-Id: I0da90d07e54225fb63f3884897fb00a6027cd537