recovery messages

Given the state of the alerts module for this type of case (reference my
earlier reply), I've had to live with it.  Only other solution beside a
rewrite of the alert module I've imagined, has been to just use a single
rule to send all alerts via script to an external script/process, that
does a more complete management of sending alerts.  My belief is that
the recovery event does happen - there's just no way to write the
matching recovery rule.
-Alan

Larry Barber wrote:

The alert rules are very simple:

PAGE=%^APHIS$
        SCRIPT /usr/local/hobbit/server/ext/SaveAlert.py DUMMY
COLORS=red DURATION>0 REPEAT=30d RECOVERED

... and so on for each page of the Hobbit installation. The script is
the start of a process that helps our NOC to manage alerts and notify
responsible parties, not being able to automatically remove things
that go red->yellow->green will cause problems.

Thanks,
Larry Barber


On 9/4/09, *Josh Luthman* <user-4c45a83f15cb@xymon.invalid
<mailto:user-4c45a83f15cb@xymon.invalid>> wrote:

    What are your alert rules?  I believe the same step that "turns it
    red" also goes through alerting.


    On 9/4/09, Larry Barber <user-6ef9c2864140@xymon.invalid
    <mailto:user-6ef9c2864140@xymon.invalid>> wrote:

I've noticed that if I have a situation where I have red alert
    that turns
yellow before finally returning to green that no recovery
    messages are being
sent. Alert notifications are only being sent on reds, not
    yellow. I suspect
that when the red turns to a yellow that Hobbit is "forgetting"
    that it sent
out the notification when it turned red since it is not sending
    them out on
yellows. Does anybody out there know how to fix this?

Thanks,
Larry Barber

    --
    Josh Luthman
    Office: XXX-XXX-XXXX
    Direct: XXX-XXX-XXXX
    XXXX Wayne St
    Suite XXXX
    Troy, OH XXXXX

    "When you have eliminated the impossible, that which remains, however
    improbable, must be the truth."
    --- Sir Arthur Conan Doyle