- (a) One of the first things I noticed using the product is that there isn't even a host column to sort by - Odd. It is, afterall, called Advanced Host Monitor.
--- In fact, a host may be called different things (Address, URL, computer, etc.) or only abstractly referenced depending upon the type of test being used.
(b) Looking through a list of, potentially, 20 millions tests is unwieldly at best.
(c) Replication of that many tests and the maintenance of those replicated tests could turn into a full time job.
Now with hosts as the focal object, we could:
- (1) start with a host profile, including an attribute for host type (SQL Server, Web Server, Router, etc.) and the ability to turn inheritance on/off for that host.
(2) hosts could then be grouped in folders/views and tests, or even a group of tests, applied to the folder/view, the host type attribute, or both.
--- We would need to have the ability to define additional uninherited tests per host.
--- On second thought, since some servers perform multiple roles, the ability to apply multiple profiles to a host would be nice.
Advantages to this approach:
- (1) The necessity for Replicator is sharply reduced, if not eliminated. (Honestly it's a handy tool, but a bit of a stopgap solution when compared to an inheritance model).
(2) This also resolves the issue of re-replicating changes. Say, for instance, I decide to change my threshold on a certain type of test, or I decide to use a different action profile for a group of hosts.
(3) Simply moving a host to a different folder could optionally change the inherited test list (allow user to turn off inheritance with a move, copying existing tests for the host to the new folder as uninherited tests.)
(Advanced Feature) Taking into account another user's issue, in a redundant or clustered server environment the functionality of the group (folder/view) could be good (green) even if one of the individual hosts is down. --- I know this is something that can be accomplished with a complex dependency list, but to simplify the process a relationship tree could be 'drawn' in order to define and visualize the releationships based on the host profiles, similar to the tables relationship view in M$Access

In this way you could then claim asset management/inventory as a feature of HM, especially with user defined fields (Purchase Date, Warranty status, etc.)
In fact, I'm not sure I want to pay more but, it could be sold as a separate, integratable product.
So, does a new potential revenue stream kick this up in priority?

-----(sorry for being so long winded
