Download a 30-Day free trial

HomeUser SupportServer Groups Feature

Server Groups Feature


I'm a potential buyer. Will the following 3 features be available in v3, or maybe added in the future?

Put servers into groups, with a "Master" server/service. i.e. I'm monitoring a Router/switch via ping. There are 10 servers connected to that switch, and I'm monitoring 30 total services. If that switch fails, it will send me an e-mail for the switch failing only, and I won't get 30+ emails for each individual service.

Same thing above but a "Master Service" for a particular server. (not as important)

Scheduler of time(s) when alerts are not sent or monitoring is paused. For like scheduled server reboots/maintenance.

Thanks.
Marc | email
Sep 5 2008, 01:03 PM
I'm also a potential buyer and I think you have a very nice, clean, and simple to use and install product.

The "Master" service feature Marc describes is one of the most basic features needed for your product.

It could be implemented by including a "dependencies" or "Parent/Child" pull-down menu for each rule. If a rule is dependent on another rule (i.e., if a rule is the child of a parent rule), then the user would use the pull-down menu to select a previously defined "parent" rule.

Without this feature, a customer's email and cellphone/pager would be flooded with incorrect error messages, which makes it difficult to track down the actual error.

I hope to see this feature soon!
Thanks, Brian
Brian | email
2 days, 5 hours since original post
I want to add some additional detail to what Marc and I are suggesting regarding "Master" or "dependent" services.

If you implement this feature, be careful regarding timing of the testing.

For example, lets say you have 1 switch and 30 servers connected to the switch. The switch is the "Master" service. If it fails, then the servers will not fail, because they depend on the master.

To implement this in your software logic, you could first test the master switch, and then test the 30 servers. If the master switch is failed, then you could disable testing of the 30 servers.

However, in the real world, the master switch is NOT going to fail immediately before you test it. What if the switch fails immediately after you test it? In this case, all 30 server tests would fail because the master switch was OK last time it was tested.

My recommendation is to add one additional logic step to take care of this issue: if any device, for example a server PING, fails the first test, go back and re-test any "Master" devices that the server PING is dependent on. In this case, before marking the server PING as fail, go back and test the switch. If the switch also fails, then disable the server PING test and mark only the switch as fail.
Brian | email
3 weeks, 5 days since original post
Thank you for the suggestions. A Master service is a feature that is likely to appear in a future release.
Tartanleaf Support | email
3 weeks, 6 days since original post
This topic is closed