I Know with SNMP Trap test you can set it to be ok only if the user acknowledges the bad status. Is it possible to do this for other tests? I want some tests to be marked as BAD and stay that way and not change to OK automatically, instead they need to be acknowledged by the user once they have investigated and resolved the issue.
This is useful for problems that might happen only at certain times or intermittently but we don't want the test to go back to OK without user intervention.
Only set OK when acknowledged
Text Log and NT Event Log test methods offers the same option.
Other tests do not have such option
Probably you can use "Change test interval" action. E.g. this action can set 2 hours test interval when test status changes to "bad" so HostMonitor will not "refresh" test status for 2 hours. Operator may check the problem and "refresh" test item manually, then "good" "Change test interval" action may restore test interval.
Regards
Alex
Other tests do not have such option
Probably you can use "Change test interval" action. E.g. this action can set 2 hours test interval when test status changes to "bad" so HostMonitor will not "refresh" test status for 2 hours. Operator may check the problem and "refresh" test item manually, then "good" "Change test interval" action may restore test interval.
Regards
Alex
Not sure we will implement such option 
Unless other admins need it as well...
May be we can add some new status to show such tests... on the other hand you always can check log file, "quick log" pane, or create SNMP Trap test items and send "bad" trap messages when such "special" test fails(HostMonitor can send message to itselfa)...
Regards
Alex
Unless other admins need it as well...
Actually it sounds pretty strange to me. If remote host restore "good" status, HostMonitor should execute assigned "good" actions, write "good" record into log and so on but at the same time keep "bad" test status?? No, we will not do this. I think this may cause a lot of problems and confuse people.I would want the test to keep running and log the results but like those other tests not return to OK status without a manual acknowledgement.
May be we can add some new status to show such tests... on the other hand you always can check log file, "quick log" pane, or create SNMP Trap test items and send "bad" trap messages when such "special" test fails(HostMonitor can send message to itselfa)...
Regards
Alex
Yes you are probably right and I guess it would be confusing for some. Setting an SNMP TRAP idea is a good one and probably will work exactly how I need.
My thought was for monitoring things where bad and good are not always closely related, e.g. once a threshold is reached, even if it goes back below the threshold then this might mean that something is still wrong with the system until someone takes some action. I wouldn't expect the good actions to kick off until the person set the status to OK, so it would stay bad but keep checking and logging the status or value (E.g. a temperature).
I guess I am thinking more along the lines of process, where someone has to do something if a threshold is reached and it cannot just be ignored because it goes back below the threshold and the status goes back to Good/ok automatically. Will play with traps and secondary tests/alerts as I think that will do the trick.
My thought was for monitoring things where bad and good are not always closely related, e.g. once a threshold is reached, even if it goes back below the threshold then this might mean that something is still wrong with the system until someone takes some action. I wouldn't expect the good actions to kick off until the person set the status to OK, so it would stay bad but keep checking and logging the status or value (E.g. a temperature).
I guess I am thinking more along the lines of process, where someone has to do something if a threshold is reached and it cannot just be ignored because it goes back below the threshold and the status goes back to Good/ok automatically. Will play with traps and secondary tests/alerts as I think that will do the trick.