Initial checkin

[klazarsk]

.
├── agents
│   ├── alert_smtp_debug.sh__________Debug wrapper to aid user in customizing smtp alert agent
│   ├── alert_smtp.sh__________________original smtp agent extended with scoping/filtering capability via options= or direct edit
│   ├── alert_syslog_debug.sh_________Debug wrapper script to aid user in customizing syslog alert agent
│   ├── alert_syslog.sh_________________new syslog agent with scoping/filtering and user-definable
│   └── README.md___________________ Initial ocumentation for the above items
├── ansible__________________________These are placeholders for deployment, and will be updated week ending June 20
│   ├── debug_alert_smtp.yml
│   ├── debug_alert_syslog.yml
│   ├── deploy_alert_smtp.yml
│   └── deploy_alert_syslog.yml
└── README.md

[wenningerk]

Sry for coming back so late and for making you extra effort.
Wanted to start this off with the exact examples from the pacemaker repo.
For reference and to better cope with licensing (still some things missing there) I wanted to have that as well with the history. Turned out not to be trivial (at least for me not having done that before) and then I took it personally ;-)

[wenningerk]

And I'm not so sure if I like the .._debug.sh approach. This doesn't change much if the alert-agents aren't packaged anyway and just copied to their target. But once we are going for a package we'd probably rather like a clean filesystem that has just files coming from an rpm. Also an sosreport won't cover this kind of changes to the filesystem.
What would speak against something along the lines of how this is done in resource-agents (trace_ra)?