Hi,

When managing a large scale environment in Zabbix (specifically network monitoring), there are several impractical implementations in the frontend. These limitations are mostly seen when hosts with a large number (500) of items/triggers/graphs and screens are added. This isn't an exceptional situation for just service provider class switches, but also applicable for standard 48 port enterprise switches.

• Problematic over view of host latest data. Even with per port applications it still ends up in > 100 port applications.
• Difficulty selecting a specific item for using in a graph
• Difficulty managing triggers for a single host
• Selecting the correct graph for a host with 1500 graphs in a small drop-down box isn't really practical.

I'm suggesting the following changes:

• Implementing sub-applications/categories. This would give us the ability to have an overview in latest data like: $Linecard --> $port --> $item
• Search/filter for triggers/graphs per host.
• Search/filter on pages where you have to select items/triggers/graphs. On several places in the interface you have to select items/triggers/graphs you want to use (e.g. graph items, screenedit, mapedit).
• Search for screens. You mostlikely want to search on screens that only have graphs for a specific host so you can select the appropriate screen.