Update "Chapter 1. Introduction to OpenStack High Availability in OpenStack High Availability Guide" to explain HA at VM/ Applications level

[CloudDon]

Introduce HA concepts at the Instance/ Application level not just the infrastructure level
-----------------------------------
Built: 2015-01-23T07:37:19 00:00
git SHA: a6b516ae73d6671894a5cd3246e5db026b0e6505
URL: http://docs.openstack.org/high-availability-guide/content/ch-intro.html
source File: file:/home/jenkins/workspace/operations-guide-tox-doc-publishdocs/doc/high-availability-guide/ch_intro.xml
xml:id: ch-intro

[Lana (loquacity)]

Some example content to get our writers started would be great!

[Ben Silverman (tersian)]

It was my understanding that the HA guide is not for workload HA, only infra HA. The only time workloads would be mentioned would be in the case of a host failure, maintenance, or other planned/unplanned infra outage (evacuate, migrate, etc?)

[Lana (loquacity)]

Ben makes a very good point. Marking as Opinion.

[Adam Spiers (adam.spiers)]

I think there *might* be some confusion around terminology here. Infrastructure HA does also include HA of the hypervisors and VMs, as covered by this user story:

http://specs.openstack.org/openstack/openstack-user-stories/user-stories/proposed/ha-vm.html

and indeed by the HA guide itself:

https://docs.openstack.org/ha-guide/compute-node-ha.html

Everything outside and including the hypervisor boundary is still considered infrastructure, but it still directly affects the availability of VMs. So references to "instance HA" or "HA at the instance level" are often considered within the scope of this aspect of infrastructure HA. But references to "workload HA" or similar are a bit more ambiguous, and probably refer to HA *inside* the instances. And Ben is of course 100% right that that should remain outside the scope of the HA guide.