About Bad, Warning statuses and Actions.

[Dubolomov]

Hi.
I have one problem. I have test that check remote gateway by ping test method. Sometime i have no reply (some problems with connections). In this case i have in alert profile action that e-mails me about lost communication. And all dependent tests stop testing for this time. All that's good.
But i would like to have a Warning status when reply of this gateway is more then 200 ms. In this case i don't want to check dependent tests (so "Thread warning status as bad" is checked) or no actions must be executed for dependent tests.
But i would like execute any actions for the warning state and if gateway was in a warning state and then it is in a Bad state.

Actions must be executed when gateway is:
- Warning after Good state for a 2 times
- Bad after Good state for a 2 times
- Bad after Warning state for a 2 times
- Good after Warning state for a 1 time
- Good after Bad state for a 1 time

Actions for dependent tests must be only if gateway is in a good state.

[KS-Soft]

Do you need 5 different actions? Or you need single "send e-mail" action?

Regards
Alex

[Dubolomov]

First of all i would like for example try to send e-mail on bad status, on warning status and on good status. And don't execute any alert actions of dependent tests when status of this test warning or bad.

[KS-Soft]

First of all i would like for example try to send e-mail on bad status, on warning status and on good status

Then use 2 standard "Send e-mail" actions:
1) bad action
- start when 2 cosecutive bad results occur
- repeat 1 time
2) good action
- start when 1 cosecutive bad result occurs
- repeat 1 time
- enable "action depends on bad one" option

These 2 actions will fit 4/5 of your need
- Warning after Good state for a 2 times
- Bad after Good state for a 2 times
- Good after Warning state for a 1 time
- Good after Bad state for a 1 time

Plus you need 1 "advanced" action for "Bad after Warning state for a 2 times" condition... H'm, I don't see how this can be done.
What about "Bad after Warning state for a 1 time"? Then you may use "advanced" action with expression like ('%Status%'=='No answer') and ('%LastStatus%'=='Warning')

In this case i don't want to check dependent tests (so "Thread warning status as bad" is checked) or no actions must be executed for dependent tests

This option has nothing to do with dependand tests.
Quote from the manual
===========
"Treat Unknown status as Bad" and "Treat Warning status as Bad" do NOT have an influence on the following features:
...
Master-dependant test relations: HostMonitor treat Warning status just like other "Bad" statuses (Bad, Bad Content, No answer)
===========

Regards
Alex

[Dubolomov]

I found the solution.
I checked "Thread warning status as Bad". In field "Use Warning status if":

Code: Select all

((%FailureIteration% > 0) and (%FailureIteration% < 2)) or ('%SuggestedReply%'>'200')

In action profile i have:
- Bad status Action
Advanced action with

Code: Select all

(('%SimpleStatus%'=='DOWN') and ('%Recurrences%'=='2')) or (('%Status%'=='No answer') and ('%LastStatus%'=='Warning')) or (('%Status%'=='Warning') and ('%LastStatus%'=='No answer'))

- Good status Action
Standard mode with 1 result, 1 time.

So.

Code: Select all

((%FailureIteration% > 0) and (%FailureIteration% < 2))

don't marks test red when one ping test lost.

Code: Select all

('%SuggestedReply%'>'200')

helps to find out when connection is not good.

Code: Select all

(('%SimpleStatus%'=='DOWN') and ('%Recurrences%'=='2'))

allerts me when connection is lost for a 2 times.

Code: Select all

(('%Status%'=='No answer') and ('%LastStatus%'=='Warning')) or (('%Status%'=='Warning') and ('%LastStatus%'=='No answer'))

alerts when status was warning or bad and now it's bad or warning accordingly.

[KS-Soft]

If you need Warning status for 1st lost ping, then yes - such settings should do the trick

Regards
Alex