Ok Normis, let me explain.
We have large distributed branch network based on mikrotik and cisco devices.
Sample topology is star, with vpn terminators in few Regional DataCenter (RD).
The connection to the RD through gre tunnels over ipsec in transport mode (point-to-point).
For each RDs we have different administrative groups of management.
Sample of our Network Map:
Where:
1. RB 951/1100AHx2 Routers
2,3. GRE Tunnels with OSPF inside
4. Cisco catalysts
5,6. RD core cisco devices
Inside RD we have lot of regional instances divided by geographical location (RU).
For each RU we have own MAP in monitoring server and own group of network admins.
What did we need:
1. Privilege separation for admin groups to access by maps (admin in group A can`t access to map B etc)
2. Syslog separation by devices and maps (different containers for each map includes each device on this map(named by IP/perfect hostname))
3. Popup window with alerts availability of GRE (probably check tunnel status or ping remote IP on tunnel) tunnels separated by maps
4. Fix critical bugs (i.e. sometimes after reboot we get empty dude.db, perfect - change type from SQLITE to MySQL)
5. Alerts to e-mail separated for each map/user group, (perfect alerts by trigger i.e. TunnelDown)
6. More clear scaling and resizing (maybe additional navigation window with approximation, now when i zoom out the map to see all devices, i get over 80% unused space)
7. Automated backup
8. Maybe something more....
As you can see, it`s not major changes of monitoring architecture but it extremely necessary for us.
If we can accelerate of development for this features it will be great.