Ad Widget

Collapse

Triggers unknown mystery

Collapse
X
 
  • Time
  • Show
Clear All
new posts
  • limo
    Senior Member
    • Dec 2004
    • 192

    #1

    Triggers unknown mystery

    Hi to all,

    we have problems with 1.3.6 and triggers. There are still unknown triggers, even if we cleaned up db (history*, trends). We thought it is due to our database that it is slow. But now we know it is not due to database. Data for triggers are OK, when I look into latest values, I see right dates, times and values.

    Please, how evaluation if trigger works? When and how it is evaluated?
    If there is trapper item, trigger is evaluated only when next data are sent or continously from history?

    Some of triggers which are still unknown:

    Pocitac A406-1 mimo provoz A406-1:errstate.str(OK)=0
    Host consuelo down consuelo:icmpping.max(#4)=0
    WEB (HTTP) server is down on consuelo consuelo:http.last(0)<1

    I think it is only in simple checks and trapper items.

    Next, when I will drop ALL icmp traffic to zabbix server, all triggers are showing state "ON" which is good. When I enable icmp pings again, after some time there is unknown state..

    Thanx!
  • Farzad FARID
    Member
    • Apr 2007
    • 79

    #2
    [1.3.6] Some triggers still count as &quot;Unknown&quot; when they should be &quot;False&quot;

    Hi,

    I have a similar problem with a trigger that is "false" today, but the various monitoring screens show it as "unknown". There is something suspicious with the dates and times, that the sample below illustrates.

    The faulty trigger was "unknown" for a moment, as the trigger information screen below shows, but the last value is really "false":

    Code:
    Time	Status	Acknowledged	Duration	Sum	%
    2007.May.03 12:13:09	FALSE 	-	22.1 hours	22.1 hours	100%
    2007.May.03 12:12:09	TRUE 	-	60 secs	60 secs	0.08%
    2007.May.03 11:02:09	FALSE 	-	1.2 hours	23.3 hours	99.93%
    2007.May.03 11:01:09	TRUE 	-	60 secs	2 mins	0.14%
    2007.May.03 10:40:09	FALSE 	-	21 mins	23.6 hours	99.86%
    2007.May.03 10:39:09	TRUE 	-	60 secs	3 mins	0.21%
    2007.Apr.26 22:32:09	FALSE 	-	6.5 days	7.5 days	99.97%
    2007.Apr.26 22:31:55	TRUE 	-	14 secs	3.2 mins	0.03%
    [B]2007.Apr.26 22:31:54	UNKNOWN 	-	1 secs	1 secs	0%[/B]
    2007.Apr.26 22:29:09	FALSE 	-	2.8 mins	7.5 days	99.97%
    2007.Apr.26 22:28:09	TRUE 	-	60 secs	4.2 mins	0.04%
    [B]2007.Apr.26 22:27:10	UNKNOWN 	-	59 secs	60 secs	0.01%[/B]
    2007.Apr.26 22:19:09	FALSE 	-	8 mins	7.5 days	99.95%
    2007.Apr.26 22:18:09	TRUE 	-	60 secs	5.2 mins	0.05%
    2007.Apr.26 21:05:09	FALSE 	-	1.2 hours	7.5 days	99.94%
    2007.Apr.26 21:04:15	TRUE 	-	54 secs	6.1 mins	0.06%
    The following screens show the trigger as "unknown":
    • Reports / Status of Zabbix: The "number of triggers" line shows 1 trigger as unknown.
    • Monitoring / Overview: The box is greyed, instead of being green.
    • Monitoring / Triggers: After clicking on "Show all triggers", I have the following line in the list:

    pragmatic-old /etc/inetd.conf has been changed on server pragmatic-old UNKNOWN Warning 04 May 00:13:47 .....

    The date and time don't correspond to any date of the previous list... I hope this helps solving the problem, otherwise I can provide more information on demand.

    Regards

    Comment

    • limo
      Senior Member
      • Dec 2004
      • 192

      #3
      Yes, right. I did not see other screens but you are right. Same problem here.

      Comment

      • Alexei
        Founder, CEO
        Zabbix Certified Trainer
        Zabbix Certified SpecialistZabbix Certified Professional
        • Sep 2004
        • 5654

        #4
        Thanks for the details, we are working on this currently.
        Alexei Vladishev
        Creator of Zabbix, Product manager
        New York | Tokyo | Riga
        My Twitter

        Comment

        • Farzad FARID
          Member
          • Apr 2007
          • 79

          #5
          Originally posted by Alexei
          Thanks for the details, we are working on this currently.
          Hi,

          The problems seems to be gone in revision 4092. Thanks for your quick patch!

          Comment

          Working...