Beyond traditional CI

Machine Mgt

Manage config

Manage binaries

golden vm does not exist

VM Appliance is start

stemcell vm's (cfr. puppet guy)

Or rebuild from scratch ?

scope of working software = works in ops

possibly including monitoring

examples for different environments

webapp in opensource environment

cruise/hudson?

cargomaven/capistrano/vlad the deployer?

cfengine/puppet/chef?

backoffice software in proprietary environment

MS SMS

sysprep/WDS

virtual and physical

xen VM + deployment tools (EC2, openqrm)

fai+pxe booting+expect script+cfengine

traditional CI stops at software building and testing

add deployment, performance testing, integration

audience

integrators

qa people

ops people

dev interested in "what happens next"

System Source definition = 2 parts

OS packages + Other software

Configuration of each component

different pipelines need to be integrated

application

environment

Server

Network

Desktop

data

Migration of data to new instance

Monitoring

Integration Monitoring

Unit Monitoring

Regression Monitoring

Changes

Patches

WSUS

RPM updates

Config changes

Software changes

Any ideas for applying the same for Network?

Virtual switch?...

either virtual or real will work, scripting needed

adding network tests (ping, routing?)?

evolving system administration

Started with manual installation of everything

Then we automated the installation and kept doing maintenance on the installation

Next step rebuilding the whole environment

Special attention to data migration