Well just to put some more on new feature that could be useful...

I'm currently getting some more data to graph for netapp filers via the stats command.

From that command I'm getting a big amount of data: interfaces, per volume/aggregate/qtree statistics, single disk utilization statistics. From all those data I'll end up with a really big number of possible graphs and putting them all in the trends column will result in an almost unreadable screen (it's difficult to search for a single graph in more then 200).

One simple to implement solution (at least to find the graph) could be to create link a the start of the page for each section and, a header before the graphs saying what they are and a footer under the graph to come back to the start of the page. For example:

CPU -> CPU stat graph Vmstat -> vmstat graph Interface -> interface graph Disk -> disk graph Snapshot -> snapshot graph

Cpu statistics: Graph1 Back to start

VM Statistics: Graph1 Graph2 Graph3 Back to start

Interface statistics Graph1 Graph2 Graph3 Back to start ... And so on

Another nice addiction can be to have more trend columns where user can define what data column will be displayed. In this case we can differentiate graphs by data type. We can have a trend column for disk spaces (disk space, inode, snapshot and so on), a trend column for machine performance (cpu, vmstat, iostat), a trend column for network performance (traffic and tcp statistics) and a trend column for environment status (power, temp, etc).

Having the 2 addiction at the same time will get probably the most out the graphs.

Francesco