I suggest exposing some configuration items to be modified via the WebUI.
- Correlation rules
- WebUI session options
- Hosts and their properties
- Services and their properties
Any changes need to be checked for validity prior to commiting to the configuration.
Any changes done to the configuration need to be consigned to an audit log.
In the nagios configuration files all the password are in clear text. It would be great to secure it especially when you use it via a web interface.
SourceForge interface is far from being intuitive. This can limit the popularity of the project.
Like the module for VMWare, it would be a good idea to generate the dependencies between hypervisors and virtual machines for Open Source solutions like Xen / KVM.
https://sourceforge.net/apps/trac/shinken/ticket/67 For users, they cannot open a port from the Arbiter to the satellites, because they are in the customer site, and the customer do not want someone to take the control of it's process. So the satellites must know how to get conf from the Arbiter if configured like it. Implementation : Instead of put_conf to socket, the Arbiter "put the conf" in memory, in standby mode, ...more »
The basic: frontend configuration and mysql database store.
centreon isn't a good thing.
Log Event Correlation: using 'ESPER' opensource for this purpose.
Global command timeout can be override by a specific one.
Fusion inventory bridge
It can be good to get hosts/service info from a real inventory tools. And raise "problems" in a real helpdesk tool too.
I would suggest having user profiles that are stored across sessions. This user profile can have a default setting and also remember things like: Dashboard entity groups selections, list of links in main dashboard.
Roles should be assigned to users.
HMAC message digests for signing pyro messages. (Pyro is the messageing framework that Shinken runs on). This feature is available starting at Pyro 4.04.
This would be a good first step in securing default Shinken installations.
With only warning/critical, we can't really know what to fix first. We need to haev a way to "tag" hosts and/or services with a criticity indicator.
Just an int between 0 (I nearly don't care about it) to 5 (if it fall, I lose my job!).
So UI can print "problems" by taking this into account (order by impacts criticity).
Show maintenance periods can hemp admins to see what the others have planned.