[dev.icinga.com #1120] don't show passive checks (active disabled) as critical in host/service problems and tac #514
Comments
Updated by mfriedrich on 2011-01-21 08:38:19 +00:00
|
Updated by mfriedrich on 2011-02-07 15:52:30 +00:00
http://www.mail-archive.com/nagios-users@lists.sourceforge.net/msg24407.html |
Updated by mfriedrich on 2011-02-07 16:18:26 +00:00 thruk adds a cfg option to that... http://www.thruk.org/documentation.html
|
Updated by mfriedrich on 2011-02-08 07:28:43 +00:00
can you please have a look at that? |
Updated by ricardo on 2011-03-11 13:06:18 +00:00 would be nice if someone could point out which conditions exactly should apply.
And if service/host has a problem then it is shown as real problem and not disabled? |
Updated by mfriedrich on 2011-03-14 21:13:48 +00:00 if the passive check is !OK it should be counted as this. but if the passive check is in the OK state, it shouldn't show up as disabled check. i'd propose a config option, and leave that by default disabled. i'm not sure if the check_freshness must be set, since some passive checks only depend on passive check results, but won't change their state because of freshness disabled. |
Updated by mfriedrich on 2011-03-24 23:24:16 +00:00 i think the idea of thruk fits best. |
Updated by mfriedrich on 2011-04-28 20:33:54 +00:00
let's postpone that since it's a bigger issue to find a satisfying solution. |
Updated by mfriedrich on 2011-05-19 09:28:01 +00:00 possibly this can be done the same way as the new status header proposes - just to split between active and passive checks and show them accordingly. so the main idea is to mark
as passive instead of 'disabled'. would this be a possible implementation then in the first place? |
Updated by ricardo on 2011-05-19 21:36:34 +00:00 Just started working on that. Fells like I'm in a status counting abyss. Just to let you know. This can't be switched back to old behavior with a config option. The counting of states will be faster and completely changed. Maybe I have to add some more filters for status.cgi to reflect the changes. But I hope, when it's finished, you will know exactly what's going on and where. |
Updated by wols on 2011-05-23 18:30:45 +00:00 A replacement 'status.cgi?...[host|service]props=42' by '10' lists passive Hosts/Services to "Unhandled hosts/services" (http://www.nagios-portal.org/wbb/index.php?page=Thread&postID=85222#post85222). |
Updated by ricardo on 2011-05-24 10:51:21 +00:00 hey Wolfgang, nice to see you here :D will have a look into it. Well guys. This sounds bad but I have to restructure the tac header again! I know It's not nice but after you have seen it, you know why. Instead of splitting it up in Active/passive/acknowledged we have to make it in "Unhandled/handled" Problems. So Putting active and passive back together. Still not sure where to to put the disabled (no active and no passive checks) Stuff. Should it be represented by an extra number or under handled stuff? At the current tac header, disabled stuff isn't shown at all. |
Updated by mjbrooks on 2011-05-24 11:07:41 +00:00 This is not exactly the thread for discussing the tac header, but here's my $0.02. Disabled stuff isn't intended to be shown. It's disabled and thus deemed unimportant. The tac header is meant to show just the important stuff. There are plenty of other ways to find and see all of the unimportant stuff already. |
Updated by mfriedrich on 2011-05-24 11:19:10 +00:00 please open a new issue for changing the tac header again. 1.4.x introduced a new look, you can't just change everything else from active/passive counting to handled/unhandled. the tac.cgi itsself is meant to show that, and the proposed change of the filter links would make sense on the global tac.cgi but not the tac_header. even though, you should discuss that in a seperated issue as said. or even ask on the blog then if people like that idea. |
Updated by ricardo on 2011-05-26 21:42:13 +00:00
I'm sorry for the inconvenience this will cause to change the tac header again. But I hope this is a solution which will be adopted rather quickly. here the changes as written in the commit:
|
Updated by mjbrooks on 2011-05-27 00:52:12 +00:00 Nice and detailed commit message, but missing refs #. Also, as I mentioned to you on IRC, don't forget to add an entry in the Changelog. Looks good so far from my quick look at it. Great work Ricardo! |
Updated by mfriedrich on 2011-06-24 10:29:46 +00:00
looks very nice indeed. only the width of the colums is hardcoded, because on larger monitors this could be stretched a bit. but for default 1024 resolution it fits. |
Updated by mfriedrich on 2011-07-12 06:49:59 +00:00
|
Updated by mfriedrich on 2014-12-08 09:39:53 +00:00
|
This issue has been migrated from Redmine: https://dev.icinga.com/issues/1120
Created by mfriedrich on 2011-01-15 13:14:27 +00:00
Assignee: ricardo
Status: Resolved (closed on 2011-07-12 06:49:59 +00:00)
Target Version: 1.5
Last Update: 2014-12-08 09:39:53 +00:00 (in Redmine)
currently, passive checks require active checks disabled. the cgis handle disabled active checks as host|service problem, and show that in
since passive checks are rather normal in configuration, this behavior should be fixed, or at least the fix being made the default, and adding a config option to restore the old behavior optionally.
Changesets
2011-06-15 21:55:57 +00:00 by ricardo b6fd633
The text was updated successfully, but these errors were encountered: